A comment on truth, science and evolution

At ResearchGate—a debate on whether evolutionary theory is scientific or true, and the criteria for scientificity:

I think truth comes in all kinds of fuzzy packets - e.g. what is true for the members of a Muslim community may not be true elsewhere— and science, too, is less of a sharp-edged body than one might think at first. Even within the Popperian paradigm, while the theory of falsification is clear in principle, when you examine specific instances there are all kinds of diverging views as to whether a given theory has been falsified or not, or whether it is or not falsifiable at all. So, there are more or less central assumptions about what is scientific or not, in communities of scientists which are themselves fuzzy—and other assumptions which are more questionable if only because the evidence is less widely known in the relevant discipline. Broadly speaking, the theory of evolution is of course scientific—but the details of such and such specific theory or specific case may not be as unquestionably scientific. Not everything that Darwin said would pass muster today as solid science, which of course is only to be expected in the process of scientific developments. This doesn't make his theory the less important for the history of biology.

(Ponen por cierto un link interesante a un artículo de Douglas L. Theobald, "A Formal Test of the Theory of Universal Common Ancestry", un modelo sobre cómo la probabilidad estadística puede servir como criterio de falsación de la teoría de Darwin sobre el origen común de todos los seres vivos).

Me contesta muy airado y despectivo uno de los interlocutores, Federico Calvoli, del Imperial College London—aquí su respuesta y mi instructiva réplica.

#
Federico Calboli · Imperial College London
"I think truth comes in all kinds of fuzzy packets"

This statement is utterly pointless and unhelpful. The word 'truth', as used in day to day basis is often no more than a statement of opinion than anything else ('what my ridiculous religious tenets say is true') has nothing to do with the issue of science.

Point one, as Popper pointed out, scientific theories *are not false*, they are never 'true', because they are always open to falsification.

Point two it confuses the process of falsification with the fact people have opinions and biases, neither of which is rooted in empirical evidence. The diverging view about any scientific theory are part and parcel of the falsification process, as long as they are rooted in the empirical falsification of a theory. Fabio might disagree on one of my theories because his data do not support my view, or he can disagree on my analysis of the data or my interpretation of my results. All these disagreements are perfectly fine (and as Popper pointed out, necessary part of science) because they are based on *empirical data*. Data can be tested for soundness and correctness, analysis of data can also be tested for soundness and correctness, and disagreements about interpretation of results can also be resolved by empirically producing and analysing more data. The constant update of scientific theories does not invalidate a Popperian view of science, because it is part of it. The theory of evolution is a perfect example of a theory that is empirically falsifiable and has undergone multiple updates as more empirical evidence has come to the fore, and these properties make it scientific.

If you, on the other hand, disagree with my theory 'because there are different points of view', and fail to address the disagreement *empirically* your opinion is utterly pointless and irrelevant.


José Angel García Landa · Universidad de Zaragoza:

Oh well, don't let's be so "utterly". My opinion may be utterly pointless and irrelevant to you, other people may see in it some kind of point. It's not "pointless and irrelevant" in itself, but "for someone". Your opinion is partly pointless to me, but not wholly so. When I spoke of truth I was not speaking of science, but of all kinds of "truth effects". Science is only one provider of such truth effects. You may disagree there; however perhaps we're on more common ground if we agree that science is the only provider of scientific truth effects. Still that's a panoramic statement of the issue, and (I insist) whenever we get down to the details things get fuzzier, once again. Much scientific discussion, for instance, is on a "nowhere land" (or a somewhere land, if you prefer) between empirical backing and falsification. Why? Because not everyone whose views or experiments are potentially relevant to the issue agrees on the interpretation of the data, or even has all the available data. There's too much going on at once. Your clear-cut view of what counts as scientific presupposes a self-evidencing falsification taking place "in itself", as against "for someone", or a sole context of relevant evaluation, which is an idealistic artifact. Sorry if I'm pointless again.

#
Les Kaufman · Boston University ×
The theory of evolution is science because it is based on evidence, not faith, and generate testables hypotheses.

#
Federico Calboli · Imperial College London ×
Josè, you are charging though a wide open door, when you state that "not everyone whose views or experiments are potentially relevant to the issue agrees on the interpretation of the data, or even has all the available data". Nobody here is saying otherwise, in fact I did say that the differences in opinion on how to generate and interpret the data is integral part of the falsification process, which is normally not a fast and clean cut process.

What is clear cut is that, scientific praxis is not just pontificating and spitting out opinions. It is a debate based on empirical data generated to falsify a prediction arising from a theory. The fact that falsification is not a clear cut process does not make it in any way similar to people debating religion, social norms or the relevance of comic strips in how architectural design evolved through time.

#
Federico Calboli · Imperial College London ×
The difference between science and other stuff:
http://store.xkcd.com/xkcd/#Science
#

José Angel García Landa · Universidad de Zaragoza ×
Lots of things work, besides science - in non-scientific ways, of course. Even falsified science works in the right context (e.g. Newton's gravity). And, what's more, every scientific discipline plays data and theories against each other following different protocols. Which is not to say that I completely disagree with what you say, not in the least.
#

Julie Hebert · University of Maryland ×
I think part of the problem, particularly regarding evolution, is the difference between common definitions and biological/scientific definitions. A theory to a layman is an idea, perhaps based on some logic, but not a proven fact. A theory to a biologist is a testable hypothesis that has been found to be true in many cases, and therefore is as close to a law as we tend to get in biology. (The "problem" with biology as a science is that there are no laws - because of the stochasticity of the biological world, there aren't broad predictions we can make for something to happen everywhere and always, at least not unless we reduce the circumstances to something so specific that it doesn't make sense for it to exist as a "law".)
As for evolution, many of the common definitions you can find on the web or in dictionaries include "gradual change" and from "a simple to a complex form" however those do not necessarily apply to a biological definition. In its most basic definition, evolution is change through time, or changes in allele frequencies (or phenotypic frequencies) in populations from generation to generation. Using that definition, it is hard for someone to show that evolution does NOT occur. The underlying mechanisms may differ in different cases (gradual vs. rapid, drift vs. natural selection, etc.) but evolution itself allows for all of those mechanisms to exist, and they are not necessarily mutually exclusive. So, in terms of science, I think it evolution is actually one of the more straight-forward and easily testable scientific concepts.

José Angel García Landa · Universidad de Zaragoza ×
I'm not one to disagree, Julie. A phenomenon such as genetic drift can be tested in laboratory conditions and, using statistical models, in actual populations as well. But as Stephen Jay Gould liked to say, biology is also a historical science. When it comes to case studies, the variables become too many to handle, too many disciplines have to cooperate to suggest and interpret, and the fuzziness I mention reappears. Take a spectacular case, dinosaur extinction. Part of the evidence to be used in a theory accounting for it will be testable in terms of "hard science", take e.g. experimental ways of determining the levels of iridium in a layer of rock—but there are just too many sides to the question in the final account, and partial rock-solid evidence added to probabilistic interpretations results in a whole panoply of different kinds of evidence. And if the result is science, it is also a kind of narrative resulting from many different kinds of science, none of which alone would be able to give a scientific account of the fact under study, i.e. the disappearance of dinosaurs. Lots of things are testable on the way, but the global account as such is not testable.

Fabio Machado · Universidade de São Paulo ×
Jerry Coyne started a similar discussion at his blog. Maybe something in his comment section could be of some use.

http://whyevolutionistrue.wordpress.com/2012/04/29/is-falsifiability-a-good-criterion-for-a-scientific-theory/

I think that José's "unfalsifiable" historical narrative fits the idea of best-explanation inference, but others could argue that, since this narrative has some logical testable implications, the narrative itself is testable.

Los orígenes de las especies

 

 

                                                            —oOo—