[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index][Subject Index][Author Index]

Re: Popper and Palaeontology (was: Re: "running" elephants - locomotary analoges)

> It's not at all clear that Hutchinson et al., or any other biologists 
> for that matter, are working within a popperian framework. Cladists 
> often claim that they are taking a "relative falsificationist" 
> approach, but that is somewhat different to Popper's brand of 
> falsificationism. Cladistics, at least, seems to me to be more of a 
> neo-positivistic method (that is; it is more intuitive to say that 
> synapomorphies support clades, rather than homoplasies semi-falsify 
> hypothesis). 
And to me, it's not at all clear what holds this paragraph together. :-)  
The mentioned biomechanical analyses are not about phylogeny, and while  
cladistics is always about phylogeny, it is not always about biology, it  
can also be -- and sometimes is -- applied to languages.  
        Back to part of the topic: Cladobabble. The way I see it at least,  
every cladogram represents the following 2 hypotheses:  
1. The cladogram is the most parsimonious (or, for molecules, often most  
likely) way to represent the data.  
2. Evolution is most parsimonious (respectively, what we call "maximum  
likelihood" is really that), or at least it's parsimonious enough that the  
few extra reversals and convergences would not change the cladogram, 
except if we had lots more data and would use those to make the cladogram. 
The first can be tested very easily (and therefore rarely is): run the 
matrix through the program again (exhaustive search), run it through 
another 1, 2 or 3 programs, if you still don't believe it dig through the 
source code (if Farris gives you that of Hennig86) and look for bugs. 
Respectively look for typos and other mistakes in the matrix -- but don't 
add information that wasn't genuinely overlooked or dismissed by the 
authors, because then you're making another cladogram. 
        Which in turn is a good idea, because the fewer you take the more 
likely you are to choose a subset that will not give you the signal that 
there is in the totality of the data. In short, add characters and (a bit 
more important, according to theoretical studies) add taxa. That's what's 
being done all the time, and why ever-growing data matrices are being used 
to examine the same old problems (the phylogeny of certain groups) again 
and again. 
The second is much harder to test (though that shouldn't be completely 
impossible... breed fruit flies for a few decades, so you know their true 
phylogeny, examine their characters, make a cladogram with those and 
compare...). Nevertheless, we must operate close to it, or cladistics is 
completely futile. All I can say is that most parsimonious and most likely 
trees tend to look similar, both tend to look similar to the results of 
traditional, more intuitive approaches, and both tend to look similar for 
different data sets (e. g. morphological vs. molecular) and the same 
groups. Besides, I can't see a reason to assume that evolution is extra 
> Popper's ideas certainly do provide a robust philosophical framework, 
> but they are notoriously difficult to apply to biological sciences. 
Why do you think so? 
> They are even harder to apply to biological sciences such as 
> palaeontology, where testing is so limited. 
The famous time machine problem is not so horrible IMHO. Because we can't 
prove anything anyway (and having a time machine would come very close to 
proving that something was so and not different), we need only find some 
prediction that our hypotheses make and that can be falsified without a 
time machine. Old example: The K-T catastrophe could have been produced by 
a supernova. This would leave plutonium-244 respectively its decay 
products in K-T sediments. There aren't any => hypothesis falsified. 
> > However, it is my personal observation that Popperism also appears to 
> > impose a certain idealised historical view regarding how science 
> > happens or has happened. [...] science does 
> > not necessarily flow as this Popperic ideal fluid. 
> (Maybe Popper would have said: "if those old astronomers and physicists  
> had used my method, they wouldn't have taken so long to get where they 
> did!" ;-)) 
Very likely. Because usually (not my idea, some philosopher's) science 
behaves more like a punctuated equilibrium: a certain view is in place 
(maybe because certain professors are in their positions), and for a long 
time nobody notices the contradicting evidence. When someone discovers 
some, he (have been few women so far... related to the fact that IIRC 58 % 
of Austrian students, but only 7 % of Austrian "ordinary" professors are 
women) isn't taken serious. Decades (or, today, maybe months :-) ) later, 
when some more evidence crops up, the establishment begins to take it 
serious, but tries to explain it away. Then sometime or other the smoking 
gun comes, everyone thinks "bingo, that's how it is", hagiographies are 
written about the 3 people who held the view before everyone else, and it 
becomes the new paradigma. Until the next guy comes and suggests a 
> > So trying to make it flow within these constraints, while calling   
> > everything else as pseudoscience, 
The classic way to tell science and pseudoscience apart is: A scientist 
takes his favorite idea, tries to destroy it with all means, and, if he 
fails, publishes it so that pillaging hordes of colleagues may have more 
success in this. A pseudoscientist takes his favorite idea, collects a bit 
of evidence that supports it, and publishes it in order to convince the 
world. It's a little strict, though. Real genuine scientists are rare by 
this definition. :-) 
> Popper was quite clear that he thought the creation of a 
> hypothesis was not methodological, 
What does this mean? 
> I don't see why you would think that Dinogeorge and Greg Paul are   
> working in a less popperian way, [...] 
I completely agree. 
> > However, in evolutionary biology, where reconstruction of the  
> > ancestor is the central issue, 
It is an important issue, but the central one? -- In any case, once you 
have a cladogram, you can simply read out many properties of the MRCA of 
any node-based taxon, and you test the presence of these properties by 
testing the cladogram. 
> > there are clear definitions of what consitutes   
> > the test of a hypothesis.  
> Hmm... parsimony, or relative falsification, or maximum genetic   
> explanation, or synapomorphic support, or evolutionary scenarios, or   
> intuition, depending on your chosen philosophy of science. 
Yeah. The remaining problem is how close any of these is to the real 
phylogeny... parsimony is at least repeatable, and see above. 
> > The only problem is that these definitions stem from certain axioms   
> > which are not necessarily as clear as Euclid' 5th postulate or its   
> > alternatives. Hence, I believe evolutionary biologists need to lay out  

> > their axioms more clearly if they want a broader audience to grasp   
> > their doing meaningfully (not the superficial understanding seen in   
> > the press)  
> I think the reason evolutionary biologists don't state their axioms   
> more clearly is because 
there aren't many, if any. :-) What could be such axioms? 
> John Conway, Palaeoartist & Protophilosopher  
David Marjanovic, who misses the accent on his c and thinks that there are 
3 kinds of philosophy: science theory, logic, and useless dilettantism by 
outsiders in biology, sociology, psychology and so on. :-) 

+++ GMX - Mail, Messaging & more  http://www.gmx.net +++
Bitte lächeln! Fotogalerie online mit GMX ohne eigene Homepage!