Concepts, methods and categories
So after much rambling (I know, I should have edited this series, and I will), here are my takeaways.
Taxa and explanations
So given the Deductive-Nomological explanatory template of the first post,
it is very difficult to make out that a taxon is an explanation, for there are few if any laws or generalisations (L), the background conditions (C) are rarely known before the taxa are named based on the phenomena (P) no matter how theoretically grounded the techniques used to acquire phenomenal information, nor what the adjacent theories of reproduction, ecology, biochemistry are.
On the broader scale of explanations, the point remains. We get the data, and we analyse it to get the phenomena. Then we explain it. The taxa are not the explanations, nor the phylogenies (a point that epidemiology could be reminded of from time to time). They are the summaries of information that call for explanations, of which such things as common descent, shared ancestral developmental systems, ecological niches and trophic webs give the explanantia.
I have, though, noted that scientific disciplines gain, through historical accident and misunderstandings often, new meanings to core concepts like explanation and hypothesis. Moreover, the explanations and hypotheses of one science (like Newtonian physics) routinely fail to properly capture what is going on in another (such as biology, social science, geology, etc.). This is licit, even in philosophy.
But novel uses of a term should not appeal naively to the cachet of the traditional or technical meanings (think of “scientific socialism” as used by Lenin); and they really ought to have a consistent meaning. The Farris-style meaning of explanation seems to me to be a garbled interpretation of the use and mandate of the term by analytic philosophers of science such as Reichenbach, Popper, and Hempel, et al. Likewise, “hypothesis” is something that wants eventually to become all or part of a theory. It’s an educated guess.[1]
So what, in my view, is the relation between taxa and explanations? The short version is that an explanation explains regularities in the world. Taxa are formed from data and analysis of data.[2] Whether or not they are constructed by abduction, or IBE, or simply by pattern recognition, as I think, There is no credible meaning to “taxa are hypotheses”. Perhaps this is because the notion of a hypothesis is itself at fault. Taxa are revisable in the light of further investigation. That is something philosophers have always held to be true of hypotheses and theories. Maybe we need a new category of scientific activity – I’m tempted to call it adduction[3] just to be mischievous – but I think pattern recognition is clear enough. And pattern recognition is what neural networks do, without a lot of theory, but with an incremental growth in past experiences and generalities. It is not a separate rational action or inference in science or much anything else. Patterns are recognised, and then we try to ascertain whether they are natural or artifactual, whether we are sacrificing accuracy for caution (Type I errors), or caution for accuracy (Type II errors).
A similar objection can be made about the theory-driven concepts of taxa, such as genetic explanations of species. What they explain is why genomes cluster in gene space, or why phenotypes are found mostly in environments of a certain kind. But what is required first of all, is an idea that there even are such phenomena as genetic regularity, regular interbreeding, etc. As a friend of mine who shall remain blameless often says, hypotheses are not evidence of the hypothesis.
Conclusions and Musings
I do not think that biology even has an ontology. It has things, sure, and these things are almost always processes with a degree of homeostasis. Hence, if species do exist, they are homeostatic processes (see the essays in Duprè and Nicholson 2018). Philosophy of biology has been overly influenced by a physics-down notion of science.
Moreover, I do not think that species, as a category, is a natural kind concept. In Tolstoyan fashion, each species is a species in its own way. Indeed, all taxa. Generalisations of properties are either the result of much observation (in effect broad summaries), or they are based (too often) on prototypical properties of some familiar taxon, or maybe even the first one studied. This is why the “fur and feathers” effect in biodiversity and conservation biology occurs. Birds and beasts, firs and ferns, are the organisms were encountered first, historically and individually. They get the priority in attention, funding and public communication.
I hold, and have argued on paper, that kind making is what humans do when faced with some domain, and this is because we do not have the working memory as persons or as institutions to record and retain every piece of data we get, or even to decide on what data to acquire. Instead, science has uncertainty and limited information for at least the special sciences, and so we always start a process of investigation from prior knowledge, however wrong it may be, and expectations.
It’s complicated, and there are no rules in biology, including that one.[4]
Thanks
Jay Odenbaugh and Mark Thompson started me on this. Neither are to blame.
Reference
Dupré, John, and Daniel J. Nicholson. Everything Flows: Towards a Processual Philosophy of Biology. Oxford: Oxford university press, 2018.
[1] I never believed Popper’s claim that hypotheses could be dreams, or Feyerabend’s similar view that they could originate in the I Ching or the Bible.
[2] I have been called a “naive empiricist”; I am not. I do think empirical work is the grounding of science, if not the basis of any particular development of theory, because otherwise we end up with theoretical hallucinations. This is roughly what van Fraassen 1980 called “constructive empiricism”. I am unsure if this is a critique of Bayesianism, but my friend Jason Grossman assures me that Bayesian reasoning can include new data in revised priors. Like I said, it’s complicated; who knew?
[3] From the Latin term for bringing forward, used in anatomy and elsewhere. In law, adducing is the act of bringing forward evidence in trial.
[4] Stolen from the late great David Hull, who was a great encouragement to me, personally and professionally, at the beginning of my philosophy of biology work.
Thanks for the bite size!
I am glad that you have corrected that (it's SNaRC, by the way). However, I am still bemused as to how to read Kirk's hypothesis. An explanation needs some sort of explicit or implict generalisation to motivate it. I for the life of me cannot see how species are explanatory hypotheses. I do not understand what that even means, and nothing you or Kirk have ever said makes a difference to my comprehension of it. This may be my limitation, but I suspect that philosophy takes these things in a different sense.
As to the D-N model not applying to historical objects that is not an obvious conclusion. As far back as William Dray's _Laws and Explanation in History_ (1957), which I read as an undergrad in 1980 [!], people have been supposing that historical generalisations are not "laws", but still function in explanations. In fact, since Cartwright's 1999 The Dappled World, even the classical laws notion of explanation has receded.
Finally, I do not "start" for the position that species are things to be discovered. That is the background assumption *of systematics* (just today I received a scan of the Proc Royal Linn Soc for 1935 where among others, J S Gilmour made that statement). I propose that speciesd are groups, not kinds (and Fitzhugh is not the target of that). But you then assert your conclusion as a starting point. This neither helps me nor is good argumentation (sorry).
Peircean abduction is not, in itself, a reason for adopting a hypothesis as correct or even viable. It is a process of inference to be sure, but not everything, even by Peirce's admission, is abduction, any more than Whewell's method of consilience is all there is, etc.
Enjoy Quine. He is one of my faves too.
Mark, thanks for posting these here.
However, this is a bit of a fire hose, so I will address some bite sized chunks. In coming days.
My one question for now is why it is you think I say you cannot look at organisms? In my Understanding Species book, I have the following passage (p84):
> There are, says Professor Julia Sigwart, an American mollusc specialist (malacologist), species *makers* and species *users*. The former are the taxonomists, and they identify, name and record species in technical journals and store the *type specimens* (the original specimen that ‘bears’ the name) in museums and other collections. There are way too few of these. The latter – well, that includes everybody, according to Sigwart. She notes in her 2019 book *What Species Mean* (chapter 3) that looking out of her window she sees species of tree, animal, bird and other living things, and that this knowledge involves two main steps: knowing that something is different from other similar (or related) things; and giving it a unique name to communicate and identify it to other users, for the taxonomists are also users of species. Knowing and naming species are related activities, but not the same.
> Sigwart makes what, as a philosopher, I think is a minor mistake, although not one that causes science that much difficulty. She doesn’t see a species of hummingbird. She sees a bird, an organism, a *specimen*. A specimen is an example of a broader group, the species, but the relation between species and specimen is fraught. On the one hand, you cannot identify a species without the use of specimens, usually by looking at many of them to see a pattern, which is then called a species. On the other hand, without knowing the broader groups that specimen is a part of (Aves, or birds, the family Trochilidae, the genus *Calypte*) you cannot identify it as a specimen of an existing species (*Calypte anna*), nor as a specimen of a newly discovered species. In short, there is a reciprocal illumination from general knowledge to particular knowledge and back again.