There are other ways to discard the idea. For example, societies which have no stigma attached to homosexuality therefore shouldn’t have paranoia.
Another method would be to look at the chain of reasoning that lead Freud to this conclusion. If the reasoning is invalid, we can discard the conclusion as a meaningful theory. It may be an accurate statement, but if it’s only reached by an invalid argument, holding it is invalid.
I suspect that the homosexual element isn’t necessary for paranoia. We’d end up eliminating one causative factor after another until we were left with the sole claim that something causes people to believe others are against them.
It would seem, then, that you are likewise implying that should it be shown that cigarette smoking does not cause cancer, it does nothing to your logical deduction. It may be the case that disproving a premise does not necessarily show the conclusion to be false, but I think it is similarly erroneous to state that in no case can it be done.
In the case of natural selection, it cannot operate any other way. There may well be other ways in which Swedish women can have a higher incidence of cancer than because they smoke more. There are no other ways in which natural selection can function. If there is no variation, there is nothing to select from. If there is no competition, then there can be no advantage bestowed upon any individual.
I will also grant that the logical form of my earlier argument may be fallacious, even though in the specific case I do not feel it is. But, not to worry, as we can play it your way.
Just a few examples, off the top of my head:
[ul][li]We should not see a prevalance within a population of fatal mutations which persist for several generations. [/li][li]We should not see populations of organisms which are adapted to a given environmental extreme outcompete and displace a population of organisms adapted to an opposite extreme (e.g., displaced desert animals cannot outcompete native tundra dwellers, and vice-versa). [/li][li]We should not see large populations evolving more quickly than small ones. [/li][li]We should not see populations adapt to conditions not present as the norm (Some individuals may possess mutations which bestow such adaptations in the absence of their need, but the frequency of those mutant alleles within the population will vary more in accordance with stochastic models, rather than a selective model, from one generation to the next. This does not rule out possible fixation of those alleles, but on average, does imply that it ought to require more generations before fixation in the absence of selective pressures than in their presence). [/li][li]We should not see identical solutions to evolutionary problems in widely divergent lineages, even if those lineages inhabit the same general environment. [/li][li]We should not find examples of “use and disuse” in nature, or, at most, such examples should be very rare.[/li][/ul]
You will, I trust, note that each is phrased as “we should not find”, rather than “we should expect to find”. Each of those is a falsfiable statement, directly or indirectly associated with evolution via natural selection.
Mind you, I have no real stake in whether evolution via natural selection is scientific by Popperian standards. I do find it odd that you are so willing to attempt to show that it is not, despite the fact that you seem to disgaree with Popper anyway: if you find that Popper insufficiently accounts for historical sciences, why go through the effort to show that one such historical science is not Popperian? It strikes me as a pointless exercise for all concerned.
The short answer to your question is that I often find myself in debate with people who believe that Popper does sufficiently account for the historical/natural sciences. Abe and I debated this question a year or so ago, for example, but never resolved it (at least to my satisfaction); and I’ve debated the issue with a number of others here as well. It’s just a bit of a pet peeve, I guess.
The pattern of threads like this tends to go something like: is creation science true? True? It isn’t even a science, since you can’t falsify it. Well, you can’t falsify evolution either. Can too. Can not. Evolution is a science, because its falsifiable. Is not. Is too. Etc. Seems like nobody ever really considers the possibility that the falsifiability of the theory of evolution might just be irrelevant to whether or not one considers it a science. So, in this thread, when the topic came up, I just thought I’d throw a spanner in the works.
In addition, it is topic that I find keenly interesting. For example, I’m really curious to know if, and in what sense, one can claim that evolution is scientific in Popper’s sense. Perhaps I’m completely wrong; I was wrong about the unfalsifiability of evolutionary pathways, maybe natural selection provides more examples of falsification than I’m aware of. Maybe Abe’s right. I’m still just basically trying to figure it out.
Finally, in these discussions I often learn a lot about Popper, natural sciences, logic, and that sort of thing. In particular, though it may sound surprising, I’ve come to respect the depth of Popper’s ideas a great deal, even though I still pretty much disagree with them.
Naturally, if you don’t care one way or the other, or you agree with me that some aspects of evolutionary theory (i.e., natural selection) aren’t falsifiable in Popper’s sense, or just find the discussion pointless, then good on ya. Please don’t let me waste your time.
Since you sound a bit fed up in your last post, truth to tell, I’m not sure whether you would be interested in reading my response to your counter-examples. At a certain point a debate begins to become less fun and more stiff-necked disagreement. I’m not interested in overstepping that line in what should really be a fairly light, pleasant exchange of ideas over a bit of fluff (that just happens to bug me on occasion), especially since I hold both you and Abe in high regard.
You state that you have no real stake in the question of whether evolution via natural selection is scientific by Popperian standards; but it seems to me that you’ve spent the majority of this thread trying to convince me that it is (arguing, essentially, in intricate detail, that evolution is falsifiable). If you don’t really care one way or the other, why bother?
Simply because a) I have not been exposed to an alternative method for determining whether something is “scientific”, and b) because I feel that natural selection is falsifiable (regardless whether that actually means anything).
As for “sounding” fed up, it has more to do with the fact that I am well aware that I have had little formal training in logic or philosophy, and any evolution texts I may have which may shed some light on the matter are currently boxed away (as are all of my other books). As such, I have little to work with when relying on my own fallible memory, and the internet is often not of much help in this particular debate. So, I feel the argument is being lost for lack of my ability to present data, rather than a lack of data, period. I assure you, it has nothing to do with you.
Let’s set the philosophical discussion to one side, then, for a few moments, and look at some of the examples of falsifiable hypotheses generated by the theory of natural selection that have been posted thus far. I’ll start with the examples that I believe are the strongest/ most valid, and move to those that I feel are the weakest /least valid.
Good examples of falsifiability:[ul][li]cajela came up with this one:[/li] Bacteria kept in an antiobiotic-rich medium will evolve resistance.
This statement should be falsifiable in a classic sense – i.e., it’s exactly the sort of observation statements Popper requires of a science. A clear-cut prediction which can be tested in a lab: if bacteria fail to develop resistance in an “anti-biotic rich medium,” the prediction is false, and that failure in turn should be a blow to the theory of natural selection. On the other hand, should the prediction turn out to be true, it lends substantial weight to the theory.
Does anyone know if its been tested?
[li]industrial melanism.[/li] paranormal mentioned this one, which Popper also refers to specifically as a good example of a falsification test:
Before England’s industrial revolution, a certain type of tree was white. During the revolution, the trees turned gray. After several generations, the moths that camoflauged themselves on the trees also turned gray. Then the cleanup act (or whatever it was called) occured in the 1960s. Soon the trees became white again and “voila!” So did the moths.*
Obviously, if Popper thinks it’s a good test, then it probably is. As a side note, it seems to be the only test Popper was aware of that met the standards of his criterion.
[li]I thought I had a couple of other good examples last night, but can’t remember them at the moment.[/ul][/li] Less good examples of falsifiability:[ul][li]The issue of adaptation to certain habitats has been brought up on a number of different occasions. I am of two minds regarding it. On the one hand, at a more general level, one can say that natural selection predicts that animals and plants poorly adapted to their environments should not be found, i.e., we won’t find polar bears in the Sahara Desert, or crocodiles swimming in the Arctic Sea. But, on the other hand, this rule only provides a loose sort of thumbnail sketch of the natural conditions of bio-diversity. For example, as Abe pointed out to me previously, there are crocodiles that have adapted to cold weather conditions. One could argue, on the basis of the prediction above, that the existence of cold water crocodiles should act as a falsifier, technically. [/li]
In fact, one often encounters characteristics in the natural world that would seem to gainsay this claim – such as the peacock’s tail. Don’t misunderstand me: Darwin’s Finch provides a very good (“all that and a bag of chips”) explanation for the tail, but strictly speaking, such examples of what would certainly seem to be “poor” adaptation abound as well. They are not understood as falsifiers of the theory, but rather are explained away in terms of other mechanisms working in combination with natural selection. I find this particular example a tough call, myself, and submit that it could fit into either box: as either a good example of natural selection’s falsifiability, or a quite poor example.[/ul]
Poor examples of falsifiability:[ul][li]cajela suggested:[/li] Artificial selection produces very different creatures than natural selection. For example, plants that cannot propagate without human intervention; animals that are much more docile and small brained than their wild counterparts. This shows that different selection mechanisms produce different things. I boldly predict that species that have been human-selected for >1000 years would be out-competed in the wild by wild-types.
Actually, all sorts of feral animals (in particular dogs and cats) seem quite capable of successfully competing with their wild-counterparts. This alone falsifies cajela’s prediction without impacting on the theory of natural selection in the slightest (as far as I can tell). I suspect that, for example, should a herd of domestic horses manage to escape into the wild, they could quite competently compete with their wild rivals. Either outcome – domestics survive best, or wild rivals survive best – could be handily explained as a result of natural selection.
[li]Abe suggests that:[/li] Further, even Darwin himself predicted on the basis of his theory of evolution the existence of “transitional fossils”. Such fossils were later found and dated, and they fit in with the predictions of evolution theory. The continued absence of such fossils, or the discovery of fossils that did not corroborate but rather contradicted evolutionary predictions (i.e. Australopithecines alive 70 million years ago), would have been more problematic.
To begin with, unless I am mistaken, this example is related to hypothesized pathways, which I’ve already acknowledged as falsifiable. I don’t think it says much one way or the other regarding mechanisms.
In addition, I contend that a lack of transitional fossils would not serve as a particularly good falsifier for the theory anyway, since one can always fall back upon the spotty nature of the geological record to explain their absence. What percentage of animal life on earth actually became fossilized, would you say? During Popper’s time, few transitionals had been located, but evolutionary theory was already scientifically accepted. When the lack of transitional links was pointed out by critics, evolutionary biologists, paleontologists, etc, routinely defended the theory by arguing that the fossil record was incomplete.
Of course, the discovery of such fossils does provide very strong confirmation of the some aspects of the theory of evolution. Unfortunately for Popperians, confirmations don’t mean much. Personally I’m not convinced that the lack of such fossils would present more of an obstacle to evolutionary theory today than it did 50 years ago, all other things being equal.
[li]Abe further suggests:[/li]
Another (deliberately simplistic) example I used was the hypothetical crocodile, the one with eyes and nostrils on its underside competing with its normally built fellows. Were natural selection not applicable to the environment of the crocodiles, there is no reason for such foolishly designed crocodiles not to exist right alongside normal crocodiles (and the same goes for other outrageously poorly adapted animals in general)*.
For reasons I pointed out in my last post, I don’t find this argument particularly convincing. Further, one seldom hears evolutionary theorists justifying the position of the crocodile’s head at the end of its neck as an adaptive advantage.
In order to meet the requirements of the falsification criterion, those specific characteristics that are explained in terms of natural selection and adaptive advantage must also be susceptible to specific falsification tests.
If you are claiming that the lack of a thumb among certain monkeys is the result of natural selection, you cannot falsify this claim by pointing out the fact that no crocodiles have faces on their bellies – at least, not from a strictly Popperian perspective.[/ul]
Examples of falsifiability that we’re still thinking about:[ul][li]cajela again:[/li]
The key point of natural selection is reproductive fitness, so any damage that one takes after reproducing is not important to natural selection and won’t be selected against. So again I boldly predict that genetically related diseases will show a pattern where those expressing in later life have a higher frequency in the human population than those expressing earlier in life.*
I’m not convinced that the opposite state of affairs would necessarily serve to falsify natural selection, and I suspect that the hypothesis involves serious problems regarding operational parameters (how does one define “genetically related disease,” etc).
[li]Abe suggests:[/li]
possibility 1. the human genome study shows no genetic material sufficiently similar to or apparently recycled from genes present in other animals (of course, recycled genetic material is exactly what the human genome shows)*.
I don’t know what you intend with this example, Abe. Not to be pedantic, but perhaps you can put it in the form of 1) hypothesis, 2) derived observation statement, 3) test, 4) result.
The same goes for possibility 2 in your last post. I’m simply not sure in what way you relate the observations to the falsifiability criterion, though I freely admit that they could very well be good support for your argument.[/ul] Darwin’s Finch:
Regarding a: aside from Popper, there have been numerous other attempts to figure out how to determine “scientificity.” Popper’s particular bugbear, actually, is induction: he’s radically skeptical, much like Hume. In fact, Popper’s criterion was specifically designed to help overcome Hume’s critique of inductive thought.
There are many philosophers of science out there who do not find Hume’s critique particularly compelling. They argue that induction is also a reasonable method of investigating nature; in rejecting Hume, they also reject Popper.
There are also those who approach the question from a more sociological angle: that is, who try to frame “science” as a human endeavor within the format of specific institutions, etc. Lakatos, for example, who rejects Popper, proposes that science can be better understood in terms of what he calls “research programmes,” which consist of: 1) a core of theories (that are relatively unfalsifiable); 2) a group of scientists who work specifically with that theoretical core; 3) a set of projects in which experiments or observations are made in relation to the theoretical core. (All that’s a simplification, of course).
Personally, I feel that determining whether or not a given statement is “scientific,” whatever that means, is really a waste of time (which is in part why I oppose Popper so adamantly). I agree with Larry Laudin: the interesting questions are more like, “Does the theory have strong empirical/conceptual credentials?” and “Is it heuristically fertile?” As far as those questions are posed in relation to evolution, the answers seem to me to be resoundingly, “Yes!” – regardless of whether or not you consider it falsifiable, not falsifiable, a science, a “metaphysical research programme,” etc.
*Really? And here I thought you were winning the argument!
P.S. My poor little fingers are typing as fast as they can, and there are only so many hours I can sit in front of the computer a day without risking divorce. I’ll try to get around to responding directly to the items on your list (if you’re still interested) as soon as I get a chance.
I’ll admit I’ve been increasingly confused by your position, Svinlesha, however I have no problem making arguments that could possibly undercut my position and support yours. I’m really not fussy as long as the topic is investigated.
Let’s see if we can straighten this knot out then. You made the claim that evolution is not falsifiable; then, when shown examples of how evolution is falsifiable, you delved into philosophical discussions of falsifiability, and then swung away from the abstract to assess concrete evidence. Unfortunately I have to struggle along as I may on this road you set out, since (like many others, I suspect) my memory of the philosophy of science is murky and often shaky, and I freely admit that I don’t know what you are talking about in some cases. However, I still think that you detected tautology where there is none.
I find it often helps to talk in terms of specifics rather than abstracts when addressing these hugely complicated problems – throw enough philosophical abstracts at someone, and everything or nothing makes sense after a while. Your claim that evolution is not falsifiable segued into your critique of falsification criteria, and you have used the latter to support the former. Or am I mistaken? If I am not, successfully arguing that evolution (or natural selection, or both?) is falsifiable ought to cause a review of your argument.
OK, first of all evolution is falsifiable It may present a challenge, but it is certainly not impossible, at least with existing criteria.
I don’t at any point suggest you deny evolution, however I got the distinct impression that it was part of your thesis to argue, as devil’s advocate, that evolution (natural selection, etc.) is not falsifiable. And I think I’ve been addressing exactly that claim of yours.
See also my correction of your use of the word “impossible” to “difficult”. In both cases I wasn’t being facetious or nitpicky, I was completely serious because the above may be an example of a fault in reasoning I suspect you’ve made on a couple of points. A presupposition is a thing assumed beforehand as a basis for an argument, whereas a good theory will, rather than assume, incorporate existing and observed data and use such as part of its argument. This distinction between incorporating existing data as opposed to presupposing certain factors exists in order to avoid the multiplication of unknowns, something that causes hypotheses like Freud’s psychoanalysis to swerve into the land of fantasy real fast (as I explained near the end of our previous discussion-- I think I counted at least five levels of unknown-multiplication in basic psychoanalysis). This point, or the difference between observation and assumption, is in fact connected to the observation of natural selection in action, I’ll quote it again for simplicity:
A theory is a model that is validated by empirical testing against physical observations, and by its logical compatibility with available data. Natural selection describes and predicts the survival and propagation of organisms best adapted to their environment over those not well adapted. The theory, it seems to me, is logically compatible with the observations cited (points 1, 2, 3).
Your argument is that, because natural selection currently explains the morphology and existence (or non) of all living creatures in an environment, it may not be falsified. That is the same argument used against solipsism (which I think you mentioned), but I don’t see its application here. We know that natural selection is not actually all there is because we have engaged, for thousands of years, in artificial selection of both plants and animals.
In artificial selection, natural selection is removed from the equation. Survival of the fittest is replaced by survival of the most desirable to the controlling agent (farmer, breeder, scientist, etc.). The environment is adjusted in order to provide a setting that will allow the controller to direct patterns of reproduction, modify system resources as needed, and even remove competition entirely from within the system (or any combination of the above).
So here we have a set of environments distinct from natural selection. Because natural selection should not be able to explain observations in a system where artificial selection rules, I would consider that adequate falsifiability of the theory.
In artificially selective environments organisms are observed to become dependent on the controller and susceptible to infection; some hybrid plants or animals lose the ability to reproduce entirely, while lack of genetic diversity (inbreeding) is a common problem: this is not the survival of the best adapted at all, but survival of the most sought-after. Therefore we end up with ears of corn dozens of times as large as their naturally selected ancestors, chickens that lay hundreds of eggs a year instead of a dozen, tiny neurotic lapdogs that bear almost no resemblance to their mighty wolf forefathers, and a monoculture of Irish potato that succumbs to wholesale destruction with just one blight (therefore bringing a fearsome collapse to the society that selected for the tuber in the first place, though this is unimportant to the discussion).
Artificial selection has resulted in a number of breeds that are dependent on us, the controllers. When inserted into a naturally selective environment, many domestic animals fare poorly, and most domestic plants are quickly overrun by weeds and parasites.
Natural selection did not alter the gene frequency, morphology, etc. of the organisms mentioned above, artificial selection did. Natural selection cannot explain why an ear of corn should be so large, or a bulldog have such a squashed face. If I follow you correctly, I have just provided an instance of the falsifiability of natural selection, because we have shown that not all possible morphological or genetic changes are necessarily explained by natural selection, and the theory therefore cannot account for all possible observations.
I think (who knows anymore?) the above has addressed the following comment:
[quote] Originally posted by Mr. Svinlesha
on the one hand, allow us to describe all possible observable outcomes of an assumed evolutionary process in terms of the theory, but, on the other hand, make it impossible to derive specific falsifiable observation statements.**
It seems to me that the confusion may have been my earlier statement that included the atrocious mention of the irremediably immersive environment or some such, the post that apparently inspired you to make this epistemological comeback. Please let that phrase die the ignoble death it deserves. I was exhausted when I typed that, not only was it the end of a long day but my brain had suffered the brutal onslaught of several meetings in only a few hours; there is nothing more stupefying and counterproductive than corporate meetings. Don’t make me resort to English as a third language as defence for the foul phrase.
It’s good that you clarified that, it may be that I am rushing through my posts (usually do, regrettably) but that did not seem to me clear at all before. OK, how do you feel about the testability of natural selection now? It is both a scientific theory, on that we agree, and, as I demonstrated above, it may be tested because there is in fact more data available than simply that already in the domain of natural selection.
No, it seems you misunderstood. Let me repeat the core of my assertion, leaving out the clumsily and hastily (again) worded part that seems to have thrown you: A scientific hypothesis that holds water must rule out some possibilities the actuality of which would render the hypothesis false.
The example I provided to illustrate this assertion was that of experimenting to test the theory of gravitation, which is another of those natural laws we have a hard time getting away from. Let’s take two of the basics of the theory of gravitation, attractive force and the inverse-square law: in each and every case if run correctly, your experiment will yield results showing that gravitation is an attractive force, and that it operates based on the inverse-square law. How do you show the theory to be falsifiable? You have to conceive of other results that the theory cannot predict and explain, such as gravitation exercising a repelling force instead of attractive, or using the inverse-cubed relationship instead of inverse-square, etc. Yet, the actual existence of a repelling force based on the inverse-cubed law when testing gravitation would clearly be utterly ridiculous and could only be an imaginary and fictitious result – just as imaginary and fictitious as a poorly-adapted organism successfully competing against well-adapted organisms in the case of natural selection.
This is quite different from the facile case of astrology or even psychoanalysis. The reason we don’t see poorly-adapted organisms competing successfully against well-adapted ones are that they aren’t able to do so, and they necessarily die out over time (successful competition, we have seen, requires adequate adaptation). This is not a tautology or fallacy as far as I can see, anymore than the fact that objects with mass attract each other exerting a force diluted by distance in accordance with the inverse square law, and they do so in every case that reflects reality. There are no exceptions we know of in reality, just like there are no exceptions (I know of, anyway) to natural selection (other of course, than artificial measures).
On the other hand you didn’t really apply my argument on falsification to astrology, you just took the A and B points I threw in there and plugged them into astrology and psychoanalysis. Inapplicaple:
No, because (as shown immediately after your post) you and Freud have played fast and loose with the logical and scientific processes involved, and in addition to that there are far too many unknowns to make such a hypothesis even remotely worthy – part of the problem one faces with very complex systems.
I agree with the above, but I don’t think I’m advocating such a reduction. Only statements that could conceivably prove a theory wrong in a logical framework should be accepted. Remember, the idea of you letting go of a watermelon from outside your window only to see it shoot upwards into the air is ridiculous, but it does provide falsifiable scope for the theory of gravitation (attractive force as opposed to repellent).
Let’s look at your astrological example:
The astrologers aren’t predicting much of anything, and you are stretching the analogy to fit my statement. I would completely agree with the above example if it were demonstrated that Prague suffers regular deluges of frogs; in that case the prediction of frog-free weather would indeed be a noteworthy falsification criterion, but (of course) nowhere near enough to lend astrology any credibility until they demonstrate that, on top of being falsifiable, astrology “theory” is also able to predict and describe certain observations.
It is once again getting late but I’ll try run the instances of natural selection and your example of astrology (slightly modified) in parallel:
CONDITIONS: Prague is regularly afflicted by rains of frogs // Organisms in the wild compete to survive, feed, reproduce
PREDICTION: There will be no frogs from the sky in Prague on the following dates // It will be extremely unlikely to find populations of organisms that are poorly adapted to an environment competing against well-adapted ones
FALSIFIABLE: The astrologers’ claim is falsifiable because it’s possible there will be a rain of frogs, in contradiction to their prediction // Natural selection is falsifiable because it’s possible for organisms poorly adapted to environment X to exist somewhere, and because we do not observe such organisms to survive long in environment X if they have to compete against better-adapted inhabitants.
SCIENTIFIC OR NOT: the astrologers’ claims are based on arbitrary and entirely variable bodies of opinion without any basis whatsoever in fact or science; they are therefore hopelessly unscientific // The process of Natural selection has been observed, is based on a very large body of other existing reliable data, and is provided in a logical scientific framework with accompanying mechanisms
Agreed in principle, but you realize that it is impossible to find in pre-Cambrian strata a hominid fossil that could not have existed in pre-Cambrian times (unless the fossil was deliberately put there as a hoax, of course). It is as impossible, as conceptual, as fantastical as my suggestion of a poorly adapted crocodile with eyes and nostrils on its underside instead of located on top of its head precisely above water level. It just wouldn’t happen, and if it did it would cause some rethinking of specific models.
I can’t go on, I write my posts from the office during spare time, but it’s the end of the day again and I have to go home. I may address your recent post as soon as I have a chance, unless someone else gets there first. As for Popper, he who is hazardous to health, someone just gave me a copy of “The Logic of Scientific Discovery” for my birthday, I look forward to re-reading that even more after this discussion once I stock up on aspirin, of course.
I screwed up the formatting in my recent post – terrible thing to do in a thorny and lengthy discussion, sorry.
The quote after my single line paragraph:
“I think (who knows anymore?) the above has addressed the following comment:”
The first paragraph of the quote is Svinlesha, the secpond paragraph is my response, and the third paragraph is Svinlesha again. Instead of closing the quote after the end of Svinlesha’s paragraph, I accidentally used “b” as if I were ending a bold type. If a Moderator can fix that it would be fantastic.
Before answering your last post in detail, I’ll wait until I see your response to my list of concrete examples. There are, however, a few points you’ve raised that I would like to briefly respond to.
Concerning the philosophy, I concur with you: I think it may be bogging the discussion down a bit. I probably should have simply launched into my critique of the specific examples of falsifiability offered here first, and left the philosophy to the side for the time being. I have a much easier time of it myself, actually, when I focus on the concrete; and like you, I’m not all that fond of the abstract either, since it can be misleading in its generality.
Looking back now upon both this thread and our last one, I can see that I may have been too categorical in my claims. I realize (now) that I’m mixing unfalsifiable, impossible to falsify, and difficult to falsify. I can see in hindsight that this was rather careless of me, and no doubt confusing to my opponents. On the other hand, there is some justification for equating these phrases: but to explain would require yet another long philosophical digression on Popper’s “critical method” and the problems of theory choice. (Actually, the problem of theory choice is pretty central to this discussion, but let’s leave that to the side for the time being.)
So, instead, let me be short: imagine a continuum between statements that are unfalsifiable, statements that are difficult to falsify, and statements that are easily falsifiable. Basically, my argument builds on two assertions:
Popper’s criterion implies that statements towards the “easily falsifiable” pole are more scientific, while statements towards the “unfalsifiable” pole are less scientific; and
The theory of natural selection is located more towards the “unfalsifiable” pole of the continuum.
I realize that I haven’t been clear about this earlier, and suspect that this lack of clarity is at least partially responsible for our long wandering through the swamps of philosophy. I’ll develop the scheme above in more detail later, if you’re interested, but post it now just so that you understand that I’m not arguing in as categorical a manner as I must have sounded to your ear. Mea culpa!
I’m leaving to the side your discussion of artificial selection for the time being.
Now, now, Abe, you’re being far too hard on yourself here. I think that “the irremediably immersive nature of the environment” is an absolutely lovely phrase, a delightful phrase, a delicate yet windblown phrase, with just the one small drawback: neither of us knows what it means. However, as Salvador Dalí once observed, “Just because I don’t know what my paintings mean doesn’t mean that I’m not a great artist!” (approximately)
After all, posting on the SDMB is an art too. Some might even argue that it is a surrealistic form of art.
And by the way, if English is your third language, let me pause to commend you on your excellent command of it. You’ve practically turned it into an art form!
That’s what I was beginning to suspect. I mentioned to DF that I view the question of “scientificity” as something of a red herring, but anyway, it also seems absurd to me to claim that evolutionary theory is anything other than scientific.
Next, in regard to the watermelon-crocodile debate:
Well, consider it like this. You note that “the idea of you letting go of a watermelon from outside your window only to see it shoot upwards into the air is ridiculous, but it does provide falsifiable scope for the theory of gravitation (attractive force as opposed to repellent).” Indeed, that’s correct. However, the gravitational law you refer to also rules out considerably less absurd possibilities. It rules out, for example, the possibility that larger, heavier objects will fall more quickly than smaller, lighter objects. It rules out the possibility that objects are attracted to each other on the basis of an “inverse cubed” law, or an “inverse x 2” law, and so on.
In fact, your gravitational law would be pretty useless if the only phenomena it ruled out were levitating (or flying) watermelons. Likewise, a theory that only rules out absurdities is also pretty useless, scientifically speaking. The law of gravitation (and to be technically correct, I’m not sure it’s a theory, but anyway) rules out pretty much every other possibility except the attraction of bodies based on the inverse square of their distance from each other. Any other result from an experimental test of that law – any other result at all – would act as a potential falsifier.
In contrast, as far as I can see from the argument you’ve presented regarding this point, your evolutionary theory can only rule out absurdities of the sort equivalent to the flying watermelon (I’m overstating here a bit, as I’ve already mentioned a couple of good falsification tests, but you get my drift). It can’t even rule out the possibility that animals may have mouths located on their bellies, or eyes on their backs, etc. etc. In fact, now that I think about it, if a crocodile with an apparently absurd anatomy were to surface, how do you know for sure that it would rule out natural selection? It might turn out that the apparently absurd characteristics had unforeseen advantages, or were offset by other sorts of advantages, etc. (Pursuant to this, I reminded of a primitive worm discovered by someone (Gould?) that had an apparently incredible anatomy. Gould wrote an entire book about it, only to later discover that what he had thought were legs were actually spines: he had been looking at the creature upside down. The upside-downess of the worm, I hasten to add, didn’t phase Dr. Gould in the least. Does this story sound familiar to you?)
Again, as I see it, the problem with this claim is that it doesn’t really reflect the reality of natural diversity, where in truth we see organisms that are have both adaptive strengths and adaptive weaknesses competing with each other. We can’t predict in advance which advantages will eventually win out over the others, because the system is too complex. This makes natural selection difficult (perhaps impossible) to falsify, at least in terms of observations from the field – no matter how compelling it is as an explanation of natural diversity.
I don’t think the point I’ve been trying to make is disputed, really, among most evolutionary biologists, paleontologists, and so on. For example, tomndebb stated earlier in this thread:
(In contrast, you can predict, quite specifically, the speed with which a watermelon falls using the law of gravitation.)
You may note, as well, the heading of the page you linked to at talk.origins:
If you follow the rather confabulated discussion on page 2, you will also note that Roth concludes:
This is precisely my point, and Roth’s conclusion – that inductions, though problematic, are nevertheless “perfectly useful accepted scientific explanations” – is nothing less than a smack-down on Popper, who’s criterion is expressly designed to rule out the inductive method.
And finally:
HAPPY BIRTHDAY, ABE!!!
Many happy returns from the Svinleshas (Tim, Marita, and Isaac).
That would be Hallucigenia, though Gould neither discovered, nor interpreted (nor re-interpreted) its arrangement - he simply reported the findings in various essays and/or books. From the link:
And to answer an earlier question, yes, I am still interested in hearing your thoughts on my examples of falsifiable predictions of natural selection.
[ nitpick ] Gould threw it in as a couple of paragraphs in Wonderful Life, then corrected the reporting (along with his acceptance of that reporting) in (I think) Bully for Brontosaurus (and the Science article collected into that volume.
(FWIW, I expressed surprise at the “spines-as-stiff-legs” claim when I first read Life, although I will not claim that I atempted to correct Gould or anyone else when I did so, relying on the pros to fix their own goofs.)
[ /nitpick ]
In reference to your list, then:[ul][li]* We should not see a prevalance within a population of fatal mutations which persist for several generations.*[/li]
Correct me if I’m wrong, but many animals are in fact “programmed” to die quickly, often shortly after mating. They have total life spans of mere days, sometimes hours. Human cells, probably all living cells, are “programmed” to commit suicide after a certain period of time, a phenomena researchers believed to be connected to aging. Sickle-cell anemia is a fatal mutation common among Africans. Some insects eat their mates directly after mating – an act which prevents the male from further spreading his genes (obviously), and should run counter your observation prediction (assuming that the behavior is “instinctually programmed” into the genes).
[li]* We should not see populations of organisms which are adapted to a given environmental extreme outcompete and displace a population of organisms adapted to an opposite extreme.*[/li]
This is a variation of the “less good example” of falsification I mentioned in my next-to-last post to Abe. As you’ve formulated it above, it’s obviously a falsifiable prediction, but of course, on the other hand, there are plenty of examples of organisms out-competing others in different environments, so the question becomes, “How do you define ‘extreme’?”
[li]* We should not see populations adapt to conditions not present as the norm…*[/li]
After which you go on to hedge your bet:”This does not rule out possible fixation of those alleles, but on average, does imply that it ought to require more generations before fixation in the absence of selective pressures than in their presence.”
As Abe pointed out earlier:
Technically, falsifiability should rule out a possibility so that should it occur, the observation statement is falsified and the theory from which it is derived threatened. Since I suspect that there exist probably millions of traits that can’t be accounted for strictly in terms of natural selection (if I’ve understood you correctly here), I judge this to be a poor example of falsifiability.
Unless, of course, you are suggesting that we shouldn’t see cold-weather adaptation in hot weather climates, etc. In that case, this claim would be a kind of variation on the one above it.
[li]* We should not see identical solutions to evolutionary problems in widely divergent lineages, even if those lineages inhabit the same general environment.*[/li]
I’m not sure what this means. How about flight? Birds, bats, squirrels (sorta), snakes (sorta), and insects (not to mention fish, but then again, they occupy a completely different environments).
[li]* We should not find examples of “use and disuse” in nature, or, at most, such examples should be very rare.*[/li]
I’ve not idea what “use and disuse” means, but when you write, “at most, such examples should be very rare,” my early-warning system starts ringing. Any example that contradicts the observation statement is a falsification, and a few examples (say, 4 or 5) ought to serve as a complete refutation of the underlying theory, much in the way Popper himself described it:”While no number of observations in conformity with the hypothesis that, say, all planets have elliptical orbits can show that the hypothesis is true or even that tomorrow’s planet will have an elliptical orbit, only one observation of a non-elliptical planetary orbit will refute the hypothesis.” However, since I don’t know what “use and disuse” refers to, I’ll leave to you evaluate my argument on this point.[/ul]So, those are my opinions regarding your list; take them for what they’re worth
Regarding Hallucigenia (what a wonderful name, by the way!) thanks for the correction. However, the problems scientists have had determining her posture also indicate the difficulties of deriving predictable morphologies from the theory, which is a problem, since morphologies are precisely what the theory is designed to explain.
Here you are conflating two separate concepts. One is the lifespan of an organism under normal conditions, and the other is the shortened lifespan that would result from the “fatal mutations” I mentioned. Within the scope of natural selection, a fatal mutation is one that causes an individual to perish before mating. Those are the sorts of mutations which should not persist for multiple generations within a population. That an indiviudal perishes after mating is not indicative of such fatal mutations.
As for the specific examples you give:
Dieing after mating has no bearing on the passage of genes to the next generation, and, in some circumstances, can help to ensure survival of the offspring. Nevertheless, such are not examples of “fatal mutations”
Sickle-Cell Anemia is fatal only if one is homozygous for the allele. It is beneficial if one is heterozygous for the allele. Homozygous instances do not persist for multiple generations, but are “new” creations based on who mates with whom.
Insects being eaten after mating have little do with with this claim, since it is not any specific mutation which results in death (rather, it is the fact of being eaten which results in such!). In many cases, the male does, in fact, attempt an escape, but rarely succeeds as the female is bigger and faster.
I stated it that way specifically as an example of adaptation. Obviously, if one is a generalist, one can compete with other generalists, or even some specialists. The trick is with specialists: if adaptation is to mean anything, then it must be the case that specialists in an environment can (and do, and indeed, must have) outcompete specialists from another, different environment. Thus the example I gave. If a species is adapted to a desert environment, it will not survive in an arctic one, and vice-versa.
In other words, an example of generalist species (like rats, or dogs, or humans) does not necessarily demonstrate that adaptation is a valid concept, preciesly because they can compete successfully (and often become dominant) in a wide range of environments. Specialists do demonstrate adaptation, since they are well-suited to a specific environment, and have become so by outcompeting others in that same environment. Thus, if we transplate one specialist species into the environment of another, the former should fail to gain dominance.
Except you miss the point. Natural selection is supposed to be the creative force of evolution, and adaptation is the result of the action of that force. If it were the case that, as a rule, organisms possessed traits which are most beneficial with respect to conditions that are not present, then adaptation becomes a meaningless concept, and natural selection cannot operate as theorized. Remember that the whole point of natural selection is that those who are best able to survive and compete given the current environment are most likely to pass on their genes, which in turn shapes both the phenotype and genotype of the population. If any old trait can come into being willy-nilly, natural selection is out the window.
This is a prediction arising form a combination of Darwin’s “descent with modification” and the concept of adaptation (both ofwhich are the products of natural selection). On the one hand, if adaptation is simply a matter of “fitting” the environment, then we might expect to see differing lineages solving the same problems the same way. However, the “solution” arrived at is also constrained by history. Thus, while some birds, some bats, and some rhamphorynchid pterosaurs may have all occupied a “fast, flying, insect-eating” niche, they all went about solving the associated problems differently. Wing structure is the most obvious example, in that all three use the forearm as the wing surface, but the form of the wing itself is widely divergent. Even the flight strokes of the three groups are different (with that of the bird and the pterosaur being most similar, because they are more closely related).
Consider also the differences between dolphins and fish, or even different fish, such as tuna and sharks. All live in the same environment, yet all have solved the issues related to underwater, high-speed locomotion some differently. Factor in some extinct creatures like ichthyosaurs, plesiosaurs and mosasaurs, and you see a vast array of solutions to the problem of life underwater, and, even, for a transition from land-bound ancestors to fully aquatic forms.
In essence, what I mean is that while we may expect to see convergence (as a consequence of adaptation), we should also see hints of history underlying any given solution; thus, no two convergences will be exact.
Again, you miss the point. “use and disuse” was a central theme of Lamarckian evolution, wherein the state of characteristics could be passed on based on the extent of their use in a parent individual. Thus, a blacksmith, with big arms, would have sons with big arms. A giraffe which stretches his neck to reach leaves would pass on “longer necks” to his offspring.
However, Darwin proposed that such was not the driving force of evolution - natural selection was. He did not rule out the possibility entirely (as Darwin did believe that some examples existed), but stated that they must be relatively rare – recall that natural selection is a statistical process; it is the major, but not only evolutionary mechanism at work. Thus, any other mechanisms must operate in the minority of cases. If we do find examples of “use and disuse” in nature, that alone would not rule out natural selection as the creative force. If we find a preponderance of such examples, then natural selection cannot be the creative force.
Except here, the problem is not one of evolution at all. It is with interpreting fossil evidence and reconstructing extinct organisms based on that evidence. Hallucigenia, being a soft-bodied wormy-thing, didn’t leave many clues as to which end is up and which is the right way round. Several hundred million years of taphonomy will do that do a body. Thus, the confusion as to what the damned thing was supposed to look like (and why it got the name it did).
Do they? I’d like a cite for that. It’s not always easy to find wild types to compete with since we’ve eliminated most of them. Where do you find wild cats and feral cats competing? (Though that’s hard since no-one is clear on cat ancestry). Easier, where do you find feral dogs outdoing wolves? In Australia, feral cats have done well - but there was no wild-type counterpart present. But feral dogs have not - the dingos were there already. How do you think a toy poodle would do in a wolf pack? Or a seedless watermelon in a field of whatever the watermelon was bred from?
Actually, I’m surprised to find this still going on after so long, but I do have some more general remarks to add.
The hypothesis that natural selection is the major driving force of evolution can be seen to be falsifiable because there are alternative hypotheses that would lead to alternative outcomes. I don’t have time to cover the detail, you know, work to do and lunch is only so long - but this may be food for thought.
Directed selection. This is where agriculture and animal breeding is relevant - it shows that a different type of selection produces different results.
Random selection - often used as he creationists’ straw man. We could try this with bacteria or mice or fruitflies and see what happens… Or take computer models like Dawkins little stick figures, and see the difference betwen directed selection and random selection.
Teleology - genes or organisms have some pre-ordained inherent development path. A “programmed” path. We could test this one in a lab too - try breeding populations of fruitflies or bacteria in different environments for many generations. Then see whether you get the same (pre-programmed) outcome, or something more like adaptation to their environment. (This is in fact done every day in biology.)
Sexual selection - a feedback cycle based ONLY on what is considered sexually desirable being propagated. Again, possible in a lab where creatures are coddled against illness and environmental hardship. Or a simulation.
Selection by birthdate - sort of astrologyish.
Selection by arbitrary trait X without regard to others - say, length of tail or colour of flower.
Lamarckian selection - selection by acquired, not inherited, characteristics (as Darwin’s Finch pointed out).
Some of these sound silly already - but that’s because they are very obviously wrong in our experience. But they do show that natural selection is not a tautology. Other forms of selection are imaginable and have even existed.
By the way, “God did it” isn’t in my list as it doesn’t seem testable. Unless you add that god did it for a specific purpose which we can observe.
Now we don’t find many direct tests of these hypotheses in biology today because it’s too low level. Like physicists rarely spend any time testing Maxwell’s equations or the inverse square law of gravity. it is a foundation to most experiments, and finding its place requires peeling back hugely many layers - but most biology articles in Science or Nature could be deconstructed to get back to an initial premise of evolution by predominantly natural selection.
PS: I’m sorry I made the “antiobiotic” typo. Antibiotic. Feel free to correct it when quoting me!
I’ll admit that some of these last examples are worthy deeper consideration, but come on, now.
By DF:
I had misunderstood this point, you’re right. However, I’m still not sure I understand you here. Your claim now appears to devolve down to the observation that if an organism dies before producing offspring, it won’t produce offspring. Come, come, good sir, I scarcely need an intricate theory of natural selection to convince me of that fact.
Yes. I still think it ought to falsify your claim above, since sickle-cell 1) can be passed from generation to generation, and 2) is fatal.
I will think more about your other examples, especially the last. By the way, “use and disuse” provides Lamarckian evolution with an entire set of imminently falsifiable observation statements: if traits acquired during a single life-time aren’t passed on, then the theory is falsified. Anyway, Darwin, you go on to state:
Okay, listen (I’m playing Popper’s advocate, now): if you claim that natural selection is a scientific theory, then you must be able to derive falsifiable observation statements from it, such that, should your observations fail to conform with the derived statement (prediction), they “automatically” falsify your theory. Therefore, your theory, in order to be scientific, must state categorically what we will, and what we will not, find in nature.
Your claim above, however, seems to represent what Popper called an “immunization.” By drawing back a step, and arguing that, “Sure, there may be counter-examples, but they still don’t falsify my theory” (i.e., “that [examples of use and disuse] would not rule out natural selection as the creative force”), you’ve effectively immunized your theory from what might otherwise be falsifying observations (“use and disuse”). And since, in fact, the theory is formulated much in the way you articulate it, this might lead one to suspect that it is unfalsifiable. At least this example as we’ve discussed it so far seems to be exposing the unfalsifiable aspects of natural selection. You cannot claim that the existence of “use and disuse” traits would function as falsifying examples of evolutionary theory, and at the same time claim that they don’t.
Capiscé?
Finally, Abe and cajela, as far as I can see, the comparisons to other forms of selection are just a non-starter, as far as I can see. Let us say that you discover that artificial selection produces a variety different from that found in nature: you’ve made no step whatsoever towards investigating the laws underlie natural variation. All you can say for sure is that something occurs in nature that does not occur under conditions of artificial selection. Basically, you’re comparing apples and oranges and claiming that, since apples are different from oranges, that proves oranges are a product of natural selection. No soap, sorry. Or, to put it differently, the observation statements you claim to derive from natural selection could be derived from any number of competing explanations as well, such as Lamarckian evolution or even Behe’s ID. They are equally useful (or useless) as falsifiers of those theories, too.
Having written the above, there are other points have been made here that are worth considering, but unfortunately, I’ve got to go to work.
I never said those examples proved natural selection. I said they demonstrated that natural selection was a falsifiable premise, since there were alternative hypotheses possible.
If we’re changing the topic off from falsifiablility to “has been falsified” then I’m just totally not interested.
Also, Darwin’s Finch’s point on the early death mutations seems to have eluded you. If an organism dies before producing offspring, yes, indeedy, it won’t produce offspring. AND that mutation will not be propagated. AND that affects the genetic structure of the population. The mutation doesn’t keep coming back as commonly as non-fatal mutations. is prevalence is reduced to the normal background mutation rates. That distribution of alleles can be checked. That distribution is mathematically predictable. Finding that distribution to be something other than what natural selection predicts would be counter-evidence. We haven’t found any such but that is NOT the point.
And since fatality of most disorders is not 100%, there is even more scope for tests - look at the distribution of alleles that cause fatalities at birth, at 20, at 50 and at 75. (Sickle cell doesn’t kill at birth, BTW.) Look at the more complex distribution of alleles where one is an advantage and two is fatal - the proportion in the population is totally what would be predicted by mathematical modelling as natural selection, until you get to people in places where the environment has changed - for sickle cell that’s no malaria & good health support.
That is also something that my colleagues do every day - distribution of various genes across different populations, usually ones that cause diseases if they go wrong. The incidence varies precisely as by natural selection predictions, not as by predictions of random, telological, directed, astrological, angelic intervention for christians or your preferred religious group, or any other modellable selection force.
If natural selection weren’t in there, say by you coddling the mutated spiral-nostrilled dog with special oxygen treatments because you thought it looked cute; that mutation might get fixed as part of the population. The fact that you don’t see that sort of genetic disorder as a significantly present genotype in natural populations, and you do see it in lab-bred or animal fancier-bred populations is pretty clear evidence that natural selection, not artificial selection or teleology or some other thing is driving their evolution.
You also say:
to which I say, Rubbish!
You have framed a hypothesis (“natural variation is produced by artificial selection”) and discarded it as falsified. Science often proceeds by trying things out and then rejecting them - that is the major point Popper was making about the importance of falsifiability. Witness Lamarckism. Or for that matter, the classical Michelson-Morley experiment on the ether.
And the idea that observed variation could equally derive from competing alternatives is just so much nonsense. That’s exactly how we managed to falsify those competing alternatives - they do NOT predict the same thing as what we see. Different selection methods produce different populations. Predictions. Ergo falsifiability. Furrfu. How obvious is that?
First off, apologies in advance for misspellings and formatting errors: I’m posting from work, on a Mac, and I can’t figure out how to change the spellchecker from Swedish to English.
This has been an inordinately long thread, and since we’re not coming to clarity on the question, maybe would should let it go. But anyway, a couple of very quick responses:
Now stop and think about this. The fact that there exist hypotheses other than ID, for example, does not demonstrate the falsifiability of ID. Not even Behe makes this argument. Consequently, the existence of other hypotheses besides natural selection doesn’t really have a bearing on the falsifiability of natural selection.
Regarding the fatal mutations, I freely admit that the point is eluding me, and that my objections could be wrongheaded; its not computing, quite sincerely. Might be a great example, but I’m just not getting my head around it yet.
Well, why didn’t you say so! If that’s the case – that one can derive a model directly from NS, test it against reality, and demonstrate that it clearly accounts for the resulting observations (such that should any other result occur, NS would be falsified) – then you have a good test (much like your example with bacteria, above).
Very good. You’ve framed a hypothesis that in no sense is derived from NS and shown it to be false. How does that have a bearing on the falsifiability, or unfalsifiability, of NS?
This might have a bearing on the falsifiability of your alternatives, but still isn’t a test of your particular theory; after all, ”God did it,” (also an unfalsifiable hypothesis) is also still available as an explanation after you’ve ruled out your alternatives. The fact that you’ve ruled them out doesn’t make the ”God did it” theory any more falsifiable, does it?
That is not my claim at all. The claim is that if an organism dies before reproducing, it takes its genes along with it. And that means that if there is a genetic cause for that individual’s early death, that genetic cause cannot persist for multiple generations within a population. That is the essence of a mutation being “selected against”. To restate, then: a fatal mutation, which manifests prior to an individual being able to reproduce, should not persist within a population.
Sickle-cell anemia still does not qualify because the gene which causes it is not always fatal, and, indeed, is equally beneficial (in a statistical sense). If you are homozygous for the mutant (SSA) allele, you die from sickle-cell anemia. If you are heterozygous (one SSA allele, the other normal), you live, and are resistant to malaria. If you are homozygous for the normal allele, you die from malaria instead. As such, the case for SSA is more complex than the simple statement I made above. Simply having the sickle cell anemia allele is not an automatic death sentence, thus is not part of the prediction to which I referred.
I would have to disagree with Popper here. Natural selection is not an absolute statement. There are no theories which can be concocted which will tell us what we will see in nature. And, natural selection, being a statistical phenomenon, simply does not apply in all cases. That still does not mean NS cannot be falsified. As I’ve said: show that it does not apply to the majority of cases, and you have falsified it. If one case can be found in which it does not apply, great: so can others, if they exist. As such, there is no logical reason to discount the falsifiable nature of NS simply because it is statistical in nature.
This is the same case we find with such theories as the kinetic theory of gases or the flow of electricity through a wire. Statistically, we cannot predict what any given molecule or electron is doing, but on average we can predict how they will act in general. Finding one errant molecule moving faster than the predicted V[sub]rms[/sub] in a gas does not affect the theory. Finding the majority of molecules doing so will raise more than a few eyebrows.
Again: if you find a majority of such examples, NS is falsified. And, again: if you can find one such example, you can find more, if they exist. Therefore, it cannot be claimed that the theory is unfalsifiable, as the means exist to falsify it. here’s an experiment: go observe multiple generations of multiple organisms. On average, does “use and disuse” appear to be rare, uncommon, common, or absent entirely? If it is common, bingo! Buh-bye NS. If it is uncommon, there are possible issues with NS. If it is rare or absent, NS is safe. One need not even find that the inheritance of acquired traits all by itself constitutes a majority-case mechanism. If the instances of NS operating as theorized are in the minority, then NS, despite being a potentially valid mechanism, cannot explain evolution.
So, on the one hand, you can determine whether or not NS works at all. On the other hand, you can determine whether it applies to the majority of cases. Both aspects of the theory which make it nearly synonymous with “evolution” are testable, and falsifying either aspect will result in serious problems for evolutionary theory as a whole.
I’ve been going over the your last couple of posts with an gradually increasing sense of confusion and frustration, but I think I may have found the point of contention. Basically, it derives from the the fact that while you are both discussing evolution in terms of statistical probabilities, I’ve been discussing it in terms of directly falsifiable single observations. Thus, for example:
The above is, of course, absolutely correct. The only problem here is that Darwin’s original statement contains no reference to mathematical predictions of alleles in a population. It may very well be obvious to a specialist that Darwin is talking about statistical occurrences of that sort, but it sure wasn’t obvious to me.
If Darwin’s point is that organisms which die before producing offspring won’t pass defective genetic material to those offspring, then it’s simply a non-starter. We may have to agree to disagree on that, but I tell you in all sincerity, that as far as I can understand the argument it relies on a very peculiar interpretation of the criterion. If, on the other hand, Darwin’s point is that NS predicts fatal genetic mutations will appear in, say, .01% to .1% of a naturally occurring population, then he has a falsifiable hypothesis: he has but to go out and test the population to see if the mutation falls within the range predicted by his model.
Again, you propose here a statistical understanding of NS. I have no problem with the falsifiability of a statistical models, myself, nor do I think Popper did (he did write a book about what he called the “indeterminant” nature of the universe, but that’s a bit of digression). But, in my defense, this is the first time that the word “statistical” has appeared in the discussion so far.
You must pardon my layman’s understanding of NS, but to me it seems to be generally employed to explain very specific traits – such as, for example, the lack of a thumb on certain monkeys, the color a peacock’s tail, the shape of a nautilus shell, and so on. In order for those particular hypotheses to be scientific (Popper would say), they must also be falsifiable at that level of specificity, if you understand what I mean. Or, to put it differently, we’re talking about two different levels of falsifiable predictions:
If NS predicts that a certain allele will appear within a given population with a specific frequency, that prediction is unarguably falsifiable, statistically speaking. Several such testable predictions would lend great strength to the falsifiability of NS at that level of specificity.
If NS is employed to explain a specific trait as an advantageous adaptation, then (I argue) falsification becomes at this level difficult (perhaps impossible). As a good example of a falsification observation of this sort, I offer cajela’s hypothesis on bacteria – that they will develop resistance in an antibacterial environment. However, I also claim that it is difficult (perhaps impossible) to derive a significant number of those sorts of tests from NS, precisely because – as has been pointed out by so many others in this discussion as well – the real world is too complex to provide us with simple falsifications.
In other words, Darwin, as you note, the kinetic theory of gases doesn’t predict how fast one particular molecule of gas might move; but neither is the theory employed to make predictions about single molecules of gas, or explain its particular behavior. In the same vein, you are arguing that NS cannot be employed to make predictions about the occurrence of a single trait: it can only be employed statistically. As far as that goes, if such statistical predictions can be derived from the theory (as you and cajela claim) then I freely concede that they are falsifiable, and in fact prove your point. However, not to be a stick in the mud, but I would like to see an example of such a test, if you have one handy.
That’s all well and good as far as it goes, but the fact remains that NS is often employed precisely in order to explain a specific trait – which would be the equivalent of employing the kinetic theory of gases to explain the behavior a specific molecule, while simultaneously claiming that if the molecule doesn’t behave that way, it doesn’t falsify the theory.
Finally: Abe? You still out there? Or have you just given up hope on me?
Begging your pardon, but I alluded to its statistical nature in my post of 08/20/03, at 10:09am (page 3?). I made specific reference to the statistical nature of NS in my follow-up post to stick monkey at 2:38pm. I’m afraid you’re not off the hook on this one
The kinetic theory of gases is a statistical model using the motion of molecules, not to predict the motion of molecules, but to model how gases exert pressure on a surface. The theory of NS proposes the mechanisms whereby pathways which generate specific traits arise.
Thus, it appears we have come full circle. You seem to acknowledge that NS is falsifiable as a statistical mechanism, and you have previously acknowledged that individual proposed pathways are themselves falsifiable. If NS is to work in the majority of cases, then we can first begin with the assumption that NS produced this structure, and investigate further to determine if that is, indeed, the case. This may be difficult, but not impossible. If it is determined that NS did not produce this this trait, that does not, however, constitute a falsification of NS as a general mechanism (which, again, because of its statistical nature, requires a preponderance of such cases).
It may present a challenge, but it is certainly not impossible, at least with existing criteria.