Pre-Darwinian Empiricism Read Through Peirce
On one hand, we're always nervous about the "science doesn't work the way scientists think it works" narrative - it sounds too much like a setup for the inevitable "also, science is just another narrative" punchline, and is kind of trivial to everyone who's not a scientist anyway. That said, the recent work on pre-Darwinian species classification is a little tangled. Here's the most recent update on work in the field:
Historian Mary P. Winsor published recently (2006b, in the December 2006 edition, but it just came out) a paper discussing how the Essentialism Story was constructed by Arthur Cain, Ernst Mayr, and David Hull. The Essentialism Story is the claim that before Darwin systematists and biologists in general treated natural kinds such as species as being defined by necessary and sufficient conditions. That is, to be a member of a species, an organism has to have all the right properties. After Darwin, goes the story, "population thinking", which denies that there are such necessary properties. Polly (she prefers to be so called) argues that this was based on Mayr's hatred of Plato, Hull's reading of Popper, and Cain's dismissal of Aristotle. In fact, she (and I) think that systematics practice was not at all based on metaphysics, but on good empiricism and she terms her replacement view "the empiricism story".
There are a couple more references, but I'm going to make you click over to Evolving Thoughts to get them if you're interested.
Obviously this is in some ways an aggressively pro-science spin: "rather than being guided by metaphysics, good scientists were (of necessity?) sound empiricists". But there is a slightly different take that's a knock against the triumphalist "we're led only by the facts no matter how radical they are" self-image of the average scientist on the bench. The idea in this triumphalist narrative is along the lines of "what's the big deal in scientific revolutions - we're not changing how we do science, they are good science". More technically, the triumphalist narrative eschews the influence of categories of thought - it's the "science is driven entirely by data and results" narrative.
After the jump, why that narrative is incomplete. Plus some Peirce.
First of all, that's obviously not a complete picture of how science works. And it's not just trivially incomplete - it actively obfuscates, at the very least, part of what science is and a lot of how scientists think. In social science, Keynesian economics was being taught well before all of the equations had been rigorously worked out. Einstein's equations were accepted long before they were fully vetted and verified not only because of faith in his genius - they were almost instantly accepted because, though they were revolutionary, they obviously fit within the vague intuitions - and rigorous knowledge - that scientists had about the universe. They were insight-ful.
Which is not to give credence to any insipid claims like "science is all intuition" or "scientific rigor is a myth". Quite the opposite - sense and sensibility for object and method are the result of either extraordinary (almost singular) genius or decades of rigorous work.
So one spin on pre-Darwinian empiricism might be that it emphasizes continuity in science. This can again be read as a kind of tempered pro-science view: science proceeds both by slowly hammering out rough edges in its methodological approach to certain objects and by slowly accumulating tenable knowledge about those objects. But it also calls into question the everyday, triumphalist "scientists are always guided by hypothesize-test-accept-test" narrative by insisting on the evolution of method.
The Peircean idea here is that certain objects in the world lend themselves better or worse to certain ways of approaching them - so (and this is obviously idealized, but usefully so) a scientist will try out one method, see if it squeezes out anything useful about the object, tweak the method, apply it again, and so on. This explanation is at the very least a nuancing of the much-vaunted scientific method: along with figuring out whether hypothesis are true, scientists are feeling out the most appropriate ways of testing those hypothesis in the first place. They adapt method to object as much as they examine object through method.
This give and take gets missed both by the triumphalist error "scientists are complete servants of data" (the nominalist error) and by the science-is-all-narrative error "scientists totally construct data" (the idealist error). The Peircean (realist) view insists upon both the influence of stratified beliefs and methods and upon different facets of nature's recalcitrance (in this language, existent Seconds refuse to resonate equally with all interpretations, they reveal themselves better when approached in some ways rather than others, etc).
The increasing evidence for pre-Darwinian empiricism can be straightforwardly but compelling read in these terms: species simply are not the kind of things that lend themselves well to metaphysical "necessary and sufficient" classification. There's only so much you can say about them that way without straining the bounds of credulity. And so people who want to say something about species - and who have a sense for how to do it - gravitate towards empiricism (or, they partially have a sense of how to do it because they have an empiricist sensibility).
The really great theorists are the ones who, despite being immersed in a methodological tradition, have such an innate sense for their objects that they can just feel out that methods' limitations. Copernicus, Newton, Einstein, of course - but also the leaders of a given field in any generation who slowly push the method outward. The degree of continuity between these two kinds of scientists is a matter of emphasis as much as historical sensibility, but it increasingly looks like Darwin was closer to the latter than the former. Which isn't to take anything away from Darwin or from big-S-Science - but it is to emphasize the reflexivity of method and the role of recalcitrance in shaping it.
References:
* The Essentialism Story [Evolving Thoughts]
Previously:
* Welcome to IIS
* Habermas's Peircean Fallibilism
* Prof. Nietzsche Frowns Upon Your Glib Globalized Paradigms
Comments (2)
I fully agree with you. The narrative is neither all data nor all metaphysics, but in the course of actually looking at the source material (and I have a sourcebook on definitions of species in press) I found that metaphysics played effectively no part in the development of the biological applications of the term "species" in natural history. Kaspar Bauhin may have appealed to medieval logic for his use of it, but that's about it. It was a word in vernacular use, and the Essentialism Story (Myth, really) is based on equivocation between the logical/metaphysical uses, the vernacular uses, and the uses in natural histry and later, in biology.
Scientific concepts track a lot of things: practice, nature, social need, and so on. But what anchors science in nature is the fact that it ultimately has to deal with the realities of investigation. In this case, the use of "species" tracks phenomenal realities. Things really do cluster into groups. The myth was devised to donwplay the originality of those who came before the "hero" - Darwin - and to increase the apparent originality of the moderns who devised the modern synthesis. The use of history to do this is a recurring theme amongst scientists - Kuhnian textbook history uses narratives in a presentist and triumphalist manner. But my slogan is that nobody was stupider before Darwin than they were after him, and the modern conceptions of species are heirs to conceptual traditions that go back a very long way.
Posted by John Wilkins | August 27, 2007 4:33 PM
Hi John,
Absolutely. My worry about the kind of unnuanced science on the bench view is not so much that it's wrong as that it's dangerous - and dangerous not in the traditional science studies "oh, it makes science hegemonic and colonialist way", but in the "it hurts science" kind of way.
Science as a fallibilistic process is (a) more accurate and (b) better suited to answer anti-science forces from the right ("there are uncertainties in evolution so we can't know if it's true") and the left ("there are cultural influences in science so all claims to knowledge must be impositions of power"). Both the rightist and the leftist mistakes are (ironically) forms of the claim "anything less than absolute knowledge is not knowledge" - which holds science up to an impossible standard in order to dismiss it.
Scientists and there defenders, sadly, often play into this conception either naively or in the name of a certain kind of overenthusiastic "we're THIS superior to the other side" machismo.
Posted by Omri Ceren | August 27, 2007 9:12 PM