Tautology Wiki

back to http://scratchpad.wikia.com/wiki/TauTology

Susuexp was a poster on the dakwins.net forum before he closed it down.

Fodor issue[]


It´s worth noting that the criticism G&L faced for the spandrel paper came in two forms:

a) That´s of course correct, but hardly worth fussing over, we´ve known that selection works that way for decades and nobody in their right mind ever argued otherwise (Maynard-Smith, Dawkins)

b) That´s completely wrong. Selection doesn´t work that way at all. (Barrash) Staddon blows in the second horn (and cites Maynard-Smith in support, no less), confuses spandrels with function change (the same issue also came up in a debate between Gould and Pinker).

What NS really is[]


Natural selection is a bias in reproduction. If you have two organisms which are identical save for one genes alleles, they will not necessarily differ in the actual number of their first generation osspring. But they may differ in the probability of reproduction and their likely contributions to future generations will be influenced by them.

Differential survival and reproduction is not selection. CROSS CHECK THIS WITH RichardDawkins saying "differential survival"

Differential probabilities of survival and reproduction produce selection. Selection is what makes alleles that provide a benefit, i.e. increase the rate of reproduction or decrease the rate of death, more likely to become fixed than those that don´t provide that benefit.

NS as stochastic process[]

Natural selection is a stochastic process. Even though it´s been repeatedly stated, let´s do this again: When we´re talking about stochasticity, we´re not talking about mutation in the fist place. We talk about drift.

fitness tautology[]

http://www.richarddawkins.net/forum/viewtopic.php?f=4&t=7428&p=1192388&hilit=exponential This is not a tautology, because fitness was derived from assumtions on what morphological features would confer an advantage and these turn out to hold in reality.

defines natural selection[]


What do you mean: made? Natural selection is a causal process. If a mutation increases the expected number of offspring of carriers, then it has an increased probability of being fixed in the population (i.e. of ending up in all the individuals at some point). There´s no entity that makes a selection. There´s only the expected number of offpring.


IF the expected number of offspring increases, then the mutation has an increased probability of being fixed in the population.

There´s no entity that makes a selection. There´s only the expected number of offpring.


his is not a tautology, because fitness was derived from assumtions on what morphological features would confer an advantage and these turn out to hold in reality.



´d disagree, it is a standard technique aimed at our unprotected balls. What creationists have started to do is undermining technical terms, though so far they´ve only taken on ones that are not nearly as central as selection. They´ve managed to almost kill the distinction between micro- and macroevolution, by inventing their own alternative BS meaning for the later term. They did kill "irreducible complexity" a term that actually had a scientific meaning and discussions about it prior to Behe had merit, but after Behe redefined the term to mean something completely different it has become unused. They´ve almost killed the reasons why some fossils are transitional ones and others are not (the same is true for transitional forms). I´ve said it before, they don´t render the people they reach scientifically illiterate, but anti-literate. No Chemist could convince a person that water freezes at 0°C, no matter how good his evidence is, if that person had been brainwashed to think that water refers to what chemists call Magnesium. And I´ve noticed that many of us don´t realize this is going on and even use the terms in the creationist sense. And that´s something that galls me about articuletts statement that mjpam uses "the language of Behe". Behe doesn´t own these words, he has ursurped them.

Selection has been described as a stochstic process for more than 70 years and stating that selection is random is nothing more and nothing less than agreeing with the basic work that went into the modern synthesis, especially that by the original 3 - Wright, Fisher and Haldane. 

That IDiots like Behe have started to abuse probability theory, to reject the scientific meaning of terms like probability, not to mention the travesty one can observe every time they start doing what they call "mathematics" (let´s have a grotesquely inflated omega, then use a uniform distribution), is not reason not to use the terms in their proper way and explain how they don´t.


http://www.richarddawkins.net/forum/viewtopic.php?f=4&t=53068&st=0&sk=t&sd=a&start=25 NOT WORKING


I´m using selection in the sense used in evolutionary biology, as a process (a stochastic one at that), that produces a pattern (which is distinct from the process leading to the formation of the pattern).

REPHRASE1: I´m using selection in the sense as a random process that produces a pattern (which is distinct from the process leading to the formation of the pattern).

REPHRASE2: I´m using selection in the sense as a random process that produces a pattern, such a pattern is the effect of this random process.

REPHRASE3: I´m using "selection" in the random sense and whatever will be the result of such a random event will be a pattern.

REPHRASE4: I´m using "selection" in the pattern sense and whatever happens will be just a pattern.

REPHRASE5: I´m using "selection" in the pattern sense and thus what will be will be.

But why are you using "selectus" in the pattern sense if Augustus, Nero and Aurelius used it 99% of the time in the design sense ?

1) After a great storm knocked over a heap of rocks, a "selection" of rocks settled in a circle on the mountain. 2) A spy satellite saw the climbers arrange a "selection" of rocks in a circle on the mountain.

Point [b]1[/b] "selection" is used in the pattern sense and [b]2[/b] "selection" is used in the design sense. A better word for "selection" in [b] 1 [/b] would be "accumulation". If your intent is in the pattern sense then the word "selection" is simply not available to you given your premises. This is a point of logic that Dernavich wrote an article about on Infidels and I don't thing he is a creationist.

NS defined precise technical meaning[]

Anything younger than that? Note that I´m not using "selectus" but "selection" and I´m doing so in the context of a discussion of evolution, where the term has a precise technical meaning. A lot of terms have technical meanings in a field, that are totally different from the vernacular. I come from germany and Quark is the word for curd cheese, it has been used in this sense since the 14th century (according to my ethymological dictionary). Still I know that physicists are using the term in a different sense and I wouldn´t even think about criticizing physicists for using a term that Goethe, Kant and Schiller used for a tasty agricultural product for something completely different.

What is natural selection[]

Articulett: Evolution is NOT about randomness unless you mean "unpredictable"-- but that doesn't help you understand it more. Evolution is about what gets passed on and why-- that is "natural selection"-- and calling it "random" has no explanatory power and confuses more than it clarifies. Which is why creationists use it.

SuseExp replies:

I repeat, this is incorrect.

Natural selection is about organisms with higher fitness having a higher propbability of reproducing.

This is not a deterministic process, if it was, we´d find that the two cells resulting from a bacterium splitting split again at exactly the same time (because they have the same genes and thus the same fitness). We don´t observe this, rather we find differential distributions of durations between splits depending on genes. Non-adaptive alleles have a low, but still existant probability of becoming fixed and this effect does play a part in speciation models. There are a lot of observable features of selection that non-random models do not replicate, but which stochstic models do replicate. So in a lot of cases the stochstic models do have significantly more explanatory power than non-stochastic ones. One of my favorite examples are the MBL papers:


Naming Conventions rin