Is the theory of evolution true?

Dale:

It seems that we have a disagreement on terms. I always thought that retrodictions were just a special sort of prediction.

 

Log in or Sign up to hide all adverts.

I think it is dangerous to use that approach. It opens the door to the ID nutjobs. Their "intelligence did it" so-called-theory has 100% explanatory power and 0% predictive power. If you put the two on equal footing then ID gains some credibility that it doesn't deserve IMO.

I would be glad to listen if you have an argument for your use of the terms.

-Dale

 

Log in or Sign up to hide all adverts.

Not really. By definition, a retrodiction IS a prediction. What separates ID and evolutionary theory is that ID does not make useful predictions/retrodictions.

No. I suggest you think about my following claim carefully: ID has ZERO explanatory power, and ZERO predictive power. The GODDIDIT theory cannot make useful predictions and/or retrodictions (assuming that retrodictions are not predictions, which I strongly disagree with). Evolutionary theory can make useful retrodictions and predictions.

I'm not going into detail at the moment, as I have a lecture in 10 min. However, I suggest you read my chromosome challenge thread (thank you Spurious for reminding me about this).
http://www.sciforums.com/showthread.php?t=53938
It demonstrates the difference between a useful retrodiction, and an ad hoc justification, which has no explainatory or predictive value.

Just a quick question. Assume you're a forensics investigator, and you suspect that a particular individual is guilty of murder. You then assume that if the suspect is guilty, their fingerprints will be on the murder weapon.
Is this a prediction?

 

Log in or Sign up to hide all adverts.

James has already address your "falsifiable" concern, I want to comment on the "predictability" part. It is generally true what you say from an operational POV because in most case, to make a prediction and then experimentally test it takes a very long time for any organism that interests anyone. NSF does not fund experiments with results 20,000 years in the future.

Thus, one must turn to small, short-lived organisms. You may not realize it, but I am sure you have tested and confirmed Darwinian natural selection at least 100 times. - I am referring to your immune system. You have among the 10^11 or so T-cells in your body a few that will fit, but poorly, onto the new antigens of the germs that is making you sick - a totally inadequate defense against their rapidly growing number. Fortunately, the next generation of T-cells (many such every hour) will fit better. Tomorrow, many generations later a significant number will fit the antigens well and their children will have expanded to perhaps 10^8 strong, all by "selection of the fittest" - quite literally the fitest - This is because those that fit well and bind to the germ antigens have a “changed environment,” which facilities them being preferentially copied, but not perfectly. The variant in their “children” which fits even better will likewise be “preferentially selected” because, fitting better, more will bind to the antigens and thus be in the environment for increased representation in the next generation, etc. Read up a little on how the immune system evolves in hours the nearly correct fit T-cells the germ antigen and makes you well again. (AIDs does it evil job by blocking this natural selection process, so you remain with too few defenders.)

I admit that has nothing to do with prediction, so let me drag from memory an old experiment that predicted and then tested. - If memory serves, it took place in South America, perhaps Brazil, but it was years before I moved to Brazil. I must "make up" some of numerical details to explain the prediction, but only the numbers I will use, not the dynamics or results. The experiment ran for about 15 years.

Many small fish living below a water fall were netted and transferred above the falls. Some were more carefully studied tanks with the same water flow thru. The size, 3 inches, and age (2.5 years, by scale rings) at first sexual maturity (developed eggs in females) and life expectancy (6 years when with adults 6 inches) were determined from the tank fish. After the transfer above the falls, predator fish were introduced below the falls, but not above and the years passed. Later fish from both above and below the falls were netted for study in separate tanks. As expected no change above the falls. AS PREDICTED, below the falls the size and age at first sexual maturity below the falls was reduced as was the egg production per fish. When you your chance of being eating is relatively constant, if your genes favor later than average sexual maturity, the percentage of your genes in next generation is decreasing compared to the fish that reproduces at a younger age. In the short period of the experiment, the average onset of sexual maturity was reduced from 2.5 to 1.5 years with the sexual mature fish only 2 instead of 3 inches long and only 2/3 as many eggs per smaller fish.

Fish were chosen with most continuing to live naturally (and cheaply) but one group with a decidedly changed environment (predator fish) which selected out of the gene pool those “foolish” enough to delay sexual maturity, such as had been predicted in the original funding proposal. Fish were also chosen because many fish do adapt their sexual behavior rapidly (some can even convert their sex - a drastic change in less than one generation but this change of course has nothing to do with evolution.)

Many creatures, including man, adjust their sex at birth ratios to provide equal number of both sexes at sexual maturity. (Pre-puberty boys have several percent higher death rates, at least prior to modern medicine, than girls. That is why in all of the world there are more boys than girls conceived*, but the excess is not as great in some Asian countries, where female infanticide has been widely practiced from prehistoric times.) this is more evidence, but not of consciously planed experiment as the “fish experiment” was. The fish experiment was not what however made me believe evolution is true (one confirmation is hardly proof of anything) What convinces me, is I see no way in which it can be avoided. I am confident it applies on all planets to all life forms that encodes and transfer characteristics (imperfectly) to the next generation. If you can conceive of how Darwin’s basic idea could be false for such a life form, I would very much like you to describe it.

I.e. Can you conceive of any such system that would not evolve, at least slowly, if the environment is not static? I think even very conscious effort (careful genetic measurement with quick death to all failing to “measure up”) by very advanced society, might come close, static for thousand of generations, but with such a stable gene pool some “germs” would evolve to seriously threaten it, so, being very advanced, that society would then also consciously change its own gene pool to resist them. Etc. I.e. they would exhibit “selected Darwinian selection” to cope with any serious environmental change. There are few “laws” that seem to be “absolutely true.”. For example I can easily imaging black body radiation might go as T^5, but I can not imagine how Darwin’s basic idea can be false. Some of the details, yes; but not the basic idea for any life form that encodes and hands down the characteristic that make it the life form it is.
---------------------------------------
*With modern sonograms, in some now the boy BIRTHs is significantly higher than in western countries - big problem ahead for some of these boys

 

Since I have never heard the word "retrodiction" until you used it I won't argue definitions. What I will argue is the value of a retrodiction vs a prediction in evaluating a theory. This type of thing always happens in physics:

Say you have a theory that is attempting to describe some phenomenon. You have some historical observation you wish to explain using the theory along with some set of initial conditions and perhaps some free parameters in your model. All of these have some error ranges associated with them. You throw in your best guess for all of the parameters and initial conditions and you find that you are way off. Then you look closer, if you tweak this initial condition that way and that parameter this way (all within the error ranges) then suddenly you have your observation exactly.

Was it reasonable to tweak the parameters that way? Well, perhaps your best guess wasn't particularly accurate. This is a particularly attractive claim if your initial conditions were long ago, far away, or otherwise difficult to observe. Unfortunately, even if your initial guess really was erroneous this approach is pretty obviously biased in favor of your theory succeeding in explaining the effect. It is also possible that several other, substantially different theories, could agree on a given historical observation, particularly with each being allowed to tweak parameters.

So how do we overcome these limitations in physics? We take the theory and we find some place where it makes a substantially different prediction and then we try to set up the corresponding experiment. We know the initial conditions that are essential (according to the theory) so we control those as tightly as possible, we therefore have much less "wiggle room" than in the retrospective approach and therefore less opportunity for our own bias to interfere. We also are purposefully selecting an observation that is difficult to explain and therefore strongly challenges the theory. If the theory is successful at predicting the result then it is much stronger, more credible, and more widely accepted than if it had simply succeeded in explaining historical observations.

I hope this explains why predictive power is more valuable than explanatory power. A correct prediction is a much higher bar for a theory to pass than a correct explanation for the reasons explained above. The best method of evaluating a theory is through its predictions, not its explanations.

-Dale

 

BillyT,

First, I know about the use of short-generation organisms in the study of evolution. I explicitly mentioned it in my first post on this thread.

Second, I have already said that the predictive power of evolution is not zero, it is just low. It is particularly low when considered relative to the kinds of theories that I am used to working with on a daily basis in engineering.

Third, I also already said that I think that evolutionary theory is correct. I am not even looking for an alternative theory.

I think that the pro-evolution crowd is so sensitized to the crackpots on this forum that you are misunderstanding what I am saying. I understand this sensitization based on my experience with the anti-relativity crowd, but I really wish you would read what I am saying more carefully. Evolution is the only real theory in biology, so you can't compare it to other biological theories. Therefore my point of reference is physics theories and, relative to physics theories, it is much weaker in terms of predictions.

I don't see why that is controversial at all. Its weakness doesn't make it wrong, it just means that I hope for a better theory in the future. I certainly think that my favorite physics theories will eventually be surpassed by superior physics theories (obviously the new theories must always agree with old ones where the old ones were known to be correct). This is part of normal scientific progress. Why should biological theories be exempt from such scientific progress?

-Dale

 

I'm sorry, but it seems that you are arguing my point. Given what you have said here, why would anyone assert that it has as much predictive power as gravitational theory? It is patently absurd. Again, I am not saying it is wrong, I am just saying that this lack of predictive power makes it weak.

-Dale

 

Sorry. I have not read much of thread, but not much harm done by making some aware of the predictive fish experiment; I think you will agree. I agree that Darwin's theory needed and has received some minor corrections, perhaps more yet to come. Biology typically has 100s of variables in all its theories, and physic only a dozen or so, at most. (Well not completely true, the position of Jupiter did influence the fall of my pencil a few minutes ago), but there are many more very significant factors in almost any biological theory. - So do not demand too much of their predictive ability until we know at least the first 100 most significant factors.

Please Register or Log in to view the hidden image!

PS there are lots of theories in biology. for example on 24th a surgen will cut into my wife's neck. Her parathyroid homrmone, PTH, is twice as high as normal and as is the blood Ca level and a lot of other less important indicators / factors associated with the theory that PTH high and Ca high implies one of the four parathyroid glans has a tumor. (They are normaly about size of grain of rice and her Brazilian radiological scan is officially negative, but I noted a few darker pixels in the over-exposed first-exposure set that are consistent (same location) in the two of the three views that placed the corresponding tissue closest to the gamma ray detector, so I think I know which of the four it is. - Your work may make you every knowledgable about all this. We are going to Tampa general hospital for the proceedure as she has had two prior ceverical operations, years ago and we want a very experienced doctor. Have you seen this surgery done with minature sterile giger counter as part of the operation? Watch this surgery at link on www.parathyroid.com

 

Evolution can't predict the future for a particular animal, but If you think about it, neither can physics. If our understanding of physics were complete, shouldn't we be to? Why can't we predict the weather, or wether someone will fall in love? The point is, physics predicts things artifically placed within the boundry of problems we say physics should be able to solve. It's like comparing knowing the properties of ice, and being able to predict the path of an iceberg. Due to the butterfly effect, such things may never be knowable with certainty.

But scientific theories are judged by their power to explain the past and present, which evolution accomplishes elegantly.

 

My God this is depressing.

What do you think your made of Mr PetriFB. Take some DNA (made from a few basic ingredients) stick it a cell give it the right chemicals at the right time and the right place (originally derived by plants from dust and air and water) and hey presto you've got man or anything else alive for that matter.

Consider genesis as an analogy of the events, a simple fairytale made for simple children of little understanding but with an eye for a good story.

If you can't see that we're related to chimps then you aren't accepting what your eyes are telling you, and God gave you those eyes so pity you, poor blasphemor. God must weep at your blindness to the scale and beauty of the world God created for you, amoungst all the creatures to behold.

 

That’s such a load of crap. I draw your attention to this response:

Theories in physics have no more power to predict random outcomes in massively complex chaotic systems than the theory of evolution has. I'll wager that the ToE is the most well-supported theory in all of science. Every branch and sub-branch of biology lends support as does numerous non-biological fields of science.

 

I never claimed that evolution should do so, that would be ridiculous. I understand very well the idea of the domain of applicability of a theory and I do not expect evolution to predict the future of a particular animal. In physics, the same can be said of quantum mechanics, which cannot predict the behaviour of a given particle. That does not prevent quantum mechanics from being an incredibly accurate and predictivly powerful theory.

I would no more expect evolution to predict the future for a particular animal than I would expect QM to predict the future for a particular particle. Please don't try to imply that I am introducing such a challenge that is clearly outside the domain of applicability of evolution. I am not being specious in this discussion.

-Dale

 

Can you all honestly claim that there is no more room for improvement in current biological theory? I don't think that is a reasonable claim at all, and I suspect that neither do you.

If there are problems with the current theory then our disagreement is not substantive but only a matter of degree. I see important limitations that lead me to call the theory "weak". You probably see the same limitations but do not consider them as important as I do (and therefore do not consider the theory weak).

-Dale

 

What sort of predictive value are you talking about then?

 

Again, Evolution is a fact, not a theory. The theories of evolution are what explain the "how" it occurs. For example, "Punctuated Equilibrium," "Population Genetics," "Genetic Drift," "Geographical Isolation" (as in Darwin's Finches), "Natural Selection," "Horizontal Gene Transfer," "Mutations." What is the question?

These are the "theories" of the mechanisms of evolution, but evolution itself is a fact.

 

Predictive power does not have to be projected on to the future. It can also be done to the past.

For instance, evolution predicts the existence of an intermediate between a lineage.

One day this intermediate is found.

Needless to say we can't predict exactly what this intermediate looks like in endless detail. We can predict more general details. That it will have intermediate characteristics of the two points that connect the lineage.

For instance, people took a drosophila gene and started searching for the equivalent in vertebrates. Nobody knew the gene would be there for sure, but the theory of evolution by means of natural selection predicts that it could very well be there. And imagine what happens? It is there!

This kind of predictive power has given a huge momentum to biological research and everybody thinks nothing of it nowadays. The theory of evolution has in fact been a huge catalyst for the field of biology, based partly on its predictive power.

 

When it comes to human opinions and hypothesis something is either less or more probable….it is never true or false.

Can a theory adequately explain a phenomenon using logic (experience based predictability and consistency) and reason (analysis devoid of emotion and bias) or can it not?