Does anyone remember Terra Nova? It was a television show that started when Steven Spielberg asked, "What if the monster in Lost was really a bunch of dinosaurs?" and ended with me telling a friend halfway through the premiere, "I think I'd rather go to sleep." For all its faults, the show did have one interesting conceit: the dinosaurs appearing on the show were, for the most part, unknown species. "Unknown" here doesn't mean "unpopular" or even "undiscovered." It means "lived in environmental conditions that would not fossilize or preserve remains."
I was thinking about Terra Nova as the Data Analysis class discussed inference from samples to populations. In particular, our professor suggested that knowledge about an entire population would be "god-like." I've read enough science fiction and fantasy to know the dangers of god-like power, but I'm totally in for god-like knowledge.
And so I found myself wondering: when I finish this Masters degree and finally have god-like knowledge of extinct species, what will I have knowledge of? As it turns out, not much turns up in a four-minute search of the philosophical literature about populations. So here are some half-baked thoughts on the issue of extinct populations.
First: populations (in the data analytical sense of the word) are made up on individuals. One has to be able to observe some subset of the population's individuals in order to infer something about the whole population. This might sound trivial, but maybe stating the obvious is the part of philosophy that separates the dabblers from the professionals. (That looked better in my head.)
Second: I think it's clear that a population (in the data analytical sense of the word) is an object of scientific study only if we can draw inferences about it. No statistic can help us to know more about the number of angels that will fit on the head of a pin and so, as far as science is concerned, angels don't form a population. As a corollary: the inferences we make about the population help us to make further inferences about individuals in the population.
Third: there are very likely (say) dinosaur species that for which no individuals have been preserved. Mountain- or rainforest-dwellers, for example, are unlikely to die and have their bodies settle in depositional environments conducive to permineralization. What if you were a dinosaur that made its living on a small island that's been submerged in the last few tens of millions of years? Or what if you were a dinosaur that made its living acting on Terra Nova? In that case, certainly no one would ever find you.
Fourth: even in the species for which we do have preserved fossils, individuality is a problem. I already wrote about that. (See? I have two drums that I keep banging.)
And so: I don't think that the study of fossils can yield any god-like knowledge about biological populations, even in principle. If the normal chain of scientific inference would be [information about observed individuals in sample]→[information about target population], and we acknowledge that there is inaccessible information about dinosaur populations and that fossils don't necessarily represent biological individuals, then inferences from fossils to dinosaur species don't follow the scientific pattern.
And, returning to the other drum I keep banging: fossils do give us information about fossil species, and so I'm just growing more and more convinced that we need to draw a hard distinction between (say) the fossil species named T. rex and the as-yet-unnamed population of organisms that left those fossils behind.