All classificatory terms are impossible of exact definition. Their use always has and always will depend upon the consensus of opinion of those best qualified by wisdom, experience and natural good sense. They will never become stable; we shall never cease to amend, to change, to repudiate old and propose new, because we shall never reach the final summation of science. [S. W. Williston, “What is a species?”, American Naturalist, 495, 184–194, 1908]
So we have come a distance. Categories about the natural (i.e., the non constructed) world are not based on some naive notion of phenomena or observations, but upon experienced (that is, expert) observations. Nevertheless they can be modulated and biased by cultural and sociological factors. Let me talk now about an example close to my heart: species (and other taxonomic categories).
As is well-known (Wilkins 2009, 2018 forthcoming, Zachos 2016), scientists do not at all agree on what species are. Nor, it must be said, do they agree on what counts as sufficient evidence that two organisms are in different species. At least, not all of the time. And yet, the standard view is that species are fundamental units of evolution, ecology, and other ways we deal with the biological world. In my forthcoming book (the new edition of my 2009, now known as Species: The evolution of the idea) I argue that the very notion of a “level” of biological taxa such as species is itself the outcome of sociocultural factors – to wit, the need to work out what “kinds” meant in the Noah’s Ark story so the logistics could be rationalised [also here]. If that is the origin, why does species persist among scientists as a category?
Once, when I was still invited to attend such things, I went to a workshop that included the late John Maynard Smith, and of course I took the opportunity to ask him about species (this was very early on in my studies on the topic, so I had nothing profound to offer him), and to eat half his sandwich. I asked him what he thought they were, and his response, which I dismissed at the time, was that they were merely communicative terms, for the convenience of biologists. I now know this was (in 2000) a common response to the species problem by geneticists, but it shocked me. How could “species” help communication, if they were not referring to anything? If the category was unreal, as he suggested, then what were we talking about?
Other biologists, particularly my now colleague and coauthor Brent Mishler, had been saying this for some time: “species” was an unnecessary rank in an otherwise rankless evolutionary process. Brent, by the way, has a book out shortly after mine on this very subject. He, and other species deniers, considered the very category nonscientific. I had suggested something similar in 2003, but it was muddled. Against the deniers are the species realists, who argue that species not only exist, but that there is a “level of organisation” in biology that answers to “species”. I have had papers rejected because reviewers took it as an article of faith that there was such a level of organisation not only in this or that group, but across the entirety of biology (except, perhaps, single-celled microbes), and the editors did not challenge this.
There is a distinction between category realism and entity realism about species that needs to be attended to. Category species realism is the view that the category is real as well as the individual entities that fall within it, and it is that which I take aim at here. Entity realism is the view that, individually, the entities (as populations, lineages, etc.) that get called species are real, but the category is not natural. It occurs to me now that Maynard Smith may have been right. We can refer to individual species like Homo sapiens, Mesoneura opaca, and Alchorena ilicifolia [Ereshefsky 1998], even if they are not all of the same category or type. I can have diverse things in my pocket: a coin, a lighter, and a ticket. They do not need to be a natural kind to get referred to as “pocketed items”.
It looks very much as though species as a category is not natural in the “natural kind” sense that philosophers have been talking about since Mill. Instead it looks like the outcome of a series of more or less frozen accidents in theology, philosophy, and science, for which retrospective justifications have been given as the needs arise. In short, species is a cassowary. A purely cultural concept. Or is it? Several things mitigate this rabid reductionism. One is that species are often named and retained for centuries. Many of the species identified in the medieval period remain “good” species, for example, although their nomenclature and arrangement in relation to other species have changed a lot. It seems that good observers can identify things in the unconstructed world even in the absence of theory and method held to be essential to good science. Another is that as I noted in the first post in this series, humans tend to find levels of categorisation of the unconstructed world fairly consistently and across cultures and times.
Hence my argument, given in my coauthored (with Malte Ebach) book The nature of classification, and expanded on in the forthcoming Species, that species are phenomena. Given that natural phenomena are a reciprocal arrangement between the unconstructed world and experienced observers, and that we have a tendency to find patterns in the world, this explains why it is that we so often find species in this or that group so obvious, even though there are multiple boundary cases where they are not obvious at all. Species, and by extension to the rest of science, natural categories, are patterns we match in observational data, based on our prior experience of uncontested cases. They need not be, and often rarely are, theory-based; but they are based on the practicalities of doing natural science. In some disciplines, such as physics, one needs to be very theoretical to even observe microscale and even mesoscale phenomena; but that doesn’t mean we are forced to be theoretical in the observation of biological, geological or even astronomical phenomena; at least, not all the time.
This goes to the main question of this post series: are the categories of science in any way objective? To answer this, I need to discuss the role explanation plays in science, or at least a rough sketch of it. Next post…