Species as objects of explanation

Following on from my previous post "Are species theoretical objects", I want now to discuss what the status of species as phenomenal objects is.

Blogging on Peer-Reviewed ResearchSome recent papers by Ingo Brigandt and Paul Griffiths (see refs), a view has been developed for some core concepts of biology - gene and homology - in which the theoretical status of these ideas is challenged. This view treats the concepts as referring to either obervational or operational objects or properties, just as I have suggested that species does. Brigandt has suggested that these objects or concepts are "units of explanation", and he and Griffiths, in an introductory essay to a recent special issue on homology of Biology and Philosophy, say this:

Biological discussions of homology are often conducted in terms of competing homology ‘concepts’. This probably reflects the older tradition of discussing the nature of species in terms of competing ‘species concepts’. In both these debates ‘concept’ is used to mean something like ‘theoretical account’ and does not carry the implication that biologists with different concepts are not thinking about the same thing. But even when it is made clear that ‘different concepts’ is to be interpreted in this fairly mundane way, we do not think it is helpful to treat disputes about homology as disputes about which concept (= definition) to adopt. So-called definitions of homology typically embody substantive commitments about how evolution and development work, and/or embody a particular methodological approach to the study of homology. Even when biologists do seem to be talking past one another as a consequence of adopting different ‘definitions’ of homology, those definitions are the reflection, rather than the cause, of differences in the questions they think are important and the methods they use to approach those questions ...

It is almost unnecessary to refer to "natural kinds" here. Terms like homology and gene are general terms that apply to putatively natural phenomena, just as species does. All such general terms in science are "natural kinds" terms, because they cover a class of phenomena, but that does not mean they are classes in the intensional, or definitional, sense.

There is a tradition in philosophy, including philosophy of science, to treat natural kind terms as being definitional terms, either in terms of necessary and sufficient properties that all, and only, members of the kind have, or as theoretical terms, defined by the role they play in the models and laws of the theory. The latter are often referred to as Ramseyfied terms - a logical sentence known as a "[Carnap-]Ramsey sentence" quantifies over all terms of the theory, and those that are existential rather than functions are considered to exist relative to the theory. In short, if the theory quantifies over variables, they are held to be factual objects.

This may (arguably) work in physics and the physical sciences that are sometimes called "general sciences". The "special sciences", which include biology, are harder to do this with, in part, I think, because there are no general exceptionless laws in those sciences that are not just the laws of physics. Many of the objects of explanation are not in their own right theoretical objects. Instead, as Brigandt notes, they are the explicanda, the things that must be accounted for. I used the mountains analogy in the previous post: mountains are not objects of geological theory; but of course they are objects of geology. Species, genes, and homologies are likewise not bounded variables of their respective theories (I may accede on genes here). If you Ramseyfy the best theories of anatomy, development, evolution and ecology, you end up with "organ", "module", "stage", "character" and "niche" etc, but not "homologue" or "species".

Taking Ingo and Paul's lead here, I dub this the phenomenal turn in biology. We have been so misled by the Ramsey approach to theories pushed by Carnap as part of the deductive axiomatic approach to theories that we have tended to leave out the phenomenal aspect of science. The parallel desire to identify "units" of biology leads us to think that all kind objects must be in a single equivalence class. But, as I have argued before, ranking objects as units is unnecessary unless there is a strong reason to do so. It is an onerous duty, not a default assumption.

So, I take slight issue with Ingo's term these "units of explanation", but the message is the same: there are phenomena of biology that call for explanation that need not be held to be objects of any theory. Ingo refers to "taxic" homologues and developmental homologues, the latter being units of independent developmental trajectories (if I understand him correctly in his 2007: 716). The former are identified as the same characters across phylogenetic and taxonomic boundaries; and are thus the outcome of a phylogenetic analysis, while the latter are objects of observation and empirical manipulation. But a homologue can be anything from a shared sequence of DNA to a complete behavioural repertoire (as Ereshefsky 2007 argues), so what is gained by calling it a "unit"? "Unit" implies that there is a rank or level of analysis; the phenomenal turn however is rankless.

I conclude that species is a phenomenal class of objects that I have elsewhere argued (Wilkins 2003) are (i) evolved modes of existence (and are therefore homologies when these modes are shared across related taxa), and (ii) are clusters of genomes and heritable traits (Wilkins 2007a, b).

References

Brigandt, Ingo. 2003. Species pluralism does not imply species eliminativism. Philosophy of Science 70 (5):1305-1316.

Brigandt, Ingo. 2007. Typology now: homology and developmental constraints explain evolvability. Biology and Philosophy 22 (5):709-725.

Brigandt, Ingo, and Paul Griffiths. 2007. The importance of homology for biology and philosophy. Biology and Philosophy 22 (5):633-641.

Ereshefsky, Marc. 2007. Psychological categories as homologies: lessons from ethology. Biology and Philosophy 22 (5):659-674.

Griffiths, Paul. 2007. The phenomena of homology. Biology and Philosophy 22 (5):643-658.

Wilkins, John S. 2003. How to be a chaste species pluralist-realist: The origins of species modes and the Synapomorphic Species Concept. Biology and Philosophy 18:621-638.

———. 2007a. The dimensions, modes and definitions of species and speciation. Biology and Philosophy 22 (2):247 - 266.

———. 2007b. The Concept and Causes of Microbial Species. Studies in History and Philosophy of the Life Sciences 28 (3):389-408.

Categories

More like this

I've been so busy reading and assimilating the latest issue of Biology and Philosophy I forgot to let you all know about it. It's a special issue on Homology, edited by Paul Griffiths and Ingo Brigandt. A discussion group has now been set up at Matt Haber's blog The Philosophy of Biology Café.…
Continuing on from my last post, let's consider the modes of speciation that are called into account for the existence of species. Here is a list taken from Sergey Gavrilets, which I put in my most recent paper in Biology and Philosophy (2007). Vicariant – divergent selection and stochastic…
There have been several attempts to produce an ontology of biology and the life sciences in general. One of the more outstanding was Joseph Woodger's 1937 The Axiomatic Method in Biology, which was based on Russell's and Whitehead's Principia and the theory of types. In this, Woodger attempted to…
So, in the last episode, you'll recall that the dastardly villain "theory" has relinquished its grip on species in a cliffhanger. But that raises a few questions. What, for instance, is it to be a theoretical object? Traditionally, something was a theoretical object, that is, an object that was…

Before you continue taking "slight issue with Ingo's term 'units of explanation'", can you please tell me where I use this term or idea, as it is not clear to me what you refer to.

I agree with your objections to the Carnap-Ramsey-Lewis approach to theoretical terms. (Technically speaking this is not an argument against the formal framework: if you take the formal 'theory' defining homology to consist in stable homology criteria rather than a contentious explanatory account of homology, then you can ramsify. The disagreement is about which of the ideas pertaining to a scientific phenomenon define the concept.)

On my reservations regarding such labels as "the phenomenal turn", see my comment on your homology special issue post.

Well I have gone through my material and notes, and you are right. I can't find that term. I'll keep looking, but I wonder where it came from. Perhaps it was in the work of a student. I know one has cited your papers.

I will modify the post accordingly.

As far as 'species' not being a theoretical term (or referring to theoretical objects if you like), the different kinds of species you mention in the post from 8/31, species caused by hybridization, natural selection, sexual selection etc., are picked out as such by our theories. I will need to read some more of your stuff, but I just don't see any species out in nature. I see a bunch of birds and plants and stuff, but no "species formed by hybridization". I only see those when I understand the theory of speciation by hybridization, without a theory those observations don't exist in some sense.

Similarly with most terms in population genetics and evolutionary theory (whatever the distinction between these two things is), I think of them as theoretical. We pick things out in the world only in virtue of our theories about them. I'm not denying they don't exist, but we carve up the world with our theories. Yes, theories do explain phenomena, but they also pick out the phenomena to be explained.

Have you read Berent Enc's (with the squiggly under the c) paper on the reference of theoretical terms from 1976 in Nous? I really think he makes a strong case that a Kripke-Putnam account of reference, what I take to be something similar to your phenomenal account(?), fails to do justice to non-ostensible terms like 'electron' and 'gene'. But I'm trying to argue these days that his objections also apply to putatively ostensible terms like 'drift' and 'species', namely because these terms are indeterminate and also depend on our theories about them. Of course I'm still stuck in the 1970s and all of this might be a bit obsolete now. If you have any more recent papers relevant to this I'd love to hear about them.

Matt, that is a very perspicacious comment, but I think I can deal with it. We do not see a "species by hybridism" we see organisms that are hybrids of two phenomenal objects. If they persist and form their own lineage, then we see a phenomenon (the species), that we explain by recourse to hybridisation.

Also, what picks out the phenomena are the theoretical ideas apart from the explanans, or the whole thing is viciously circular. We use molecular genetics as, for example, the baseline of the phenomena in order to identify groups in need of explanation: but these can be haplotype groups, populations or species. Genetics is not the theory that defines species, because it can identify clustering at a number of scales. But it does give us the phenomena, at least in part. But so too does general observation: we have (biologically constrained) similarity metrics inbuilt. We also have a host of social and cultural identification protocols, mostly implicit. Nobody needed a theory of "species" before, say, 1900, in order to identify them. It is too theory-centric to claim that theories define what we see - I reject the radical interpretation of theory-dependence of observation (for Hackingesque reasons).

I have read (and have a physical preprint somewhere in the mess) Enç's paper. Thanks for reminding me of it.

Being a systematic zoologist, it is unfortunate that I do not understand much of this discussion. I do think the comment that homologies are the result of phylogenetic analysis is just backwards. Is it not rather that phylogenetic analysis is based on recognition of homologies?

I get the impression that much of the conversation here is biased toward botanical systematics. I am not real clear on how botanists go about erecting their taxa. I suspect this is because there is a fair amount of non-overlap in the biological natures of plants and animals. For example, I recall reading that perhaps 25% of the higher plant species are of hybrid origin; whereas there is a small percentage (surely less than 1%) of animal species known to be of hybrid origin.

By Jim Thomerson (not verified) on 13 Nov 2007 #permalink

While you are right that plants hybridise much more than animal species do, I think that the difference is one of degree rather than kind (forgive the pun!). There are numerous hybrid species of animals, mostly avian or reptilian. And there are fewer than one might expect amongst plants. Ferns and fern allies have, if memory serves, around 5% hybrid origins, although if you include the ancestors, more like 95% have hybrid events in their lineage.

Phylogenetic analysis, as I understand it, is geared towards eliminating homoplasies from the data set. If homologies were unobjectionable or easy to obtain, this would not be the case. So phylogenetics and homology are linked inextricably. But what gets it started is a recognition of homology based on past experience and a knowledge of the organisms. This then leads to the recognition of other homologies, especially at the molecular level.

But I am not a practicing systematist, so I may have it wrong.

I agree with your last paragraph. The recognition of homologies is complexicated by convergent and parallel evolutionary events. So the the process is back and forth as you suggest. Darwinian parsimony; that similartiies are evidence of relationships, is only the first hypothesis, and is often wrong. I well remember my major professor correcting me when I said I had shown that two populations were more closely related to each other than to a third. No, I had only shown they were more similar; not that the similarity was due to relationship.

I'll have to look into animal hybrid origin species a little more. The only ones I am aware of right off are the all-female species of fishes and lizards. I don't doubt that there are a few more I don't know about. But I do not think hybridization is a major contribution to origin of species in animals.

By Jim Thomerson (not verified) on 14 Nov 2007 #permalink

I just read de Waal's Our Inner Ape, where he notes hybrids between a bonobo male and chimp females have been observed. No idea if the progeny were fertile, though.