Kenyanthropus platyops makes an interesting case study of species in the fossil hominin record. The name formally applies to only two fossils, which are the holotype KNM-WT 40000 skull, and a maxilla KNM-WT 38350 named as a paratype. Both of these fossils come from the Lomekwi area and are between 3.6 and 3.2 million years old. This is a time interval in which scientists have identified Australopithecus afarensis at several sites in Ethiopia to the north, and at Laetoli to the south of Turkana, and it might seem reasonable that Au. afarensis probably also was between these northern and southern extremes at some times, possibly at Lomekwi. Indeed, mandibular fossils including KNM-WT 8556, discovered in 1982, were long identified as Au. afarensis.
This leaves a problem: If the fossils formally identified as K. platyops have no mandibles, then how can we test whether a mandible from Lomekwi or other sites belongs to K. platyops?
Anthropologists have taken a number of different tacks on this question. One extreme has been to deny that K. platyops is different from Au. afarensis at all, either because the holotype is too crushed and distorted to test, or because Au. afarensis is so variable that it can encompass these fossils easily. Another extreme is to provisionally accept every hominin from the Turkana Basin of this age as K. platyops. So for example, research on stable isotopes in Turkana Basin teeth lists dozens of K. platyops fossils, although none of these have been formally identified as K. platyops in any publication.
At one level this is all academic. The evidence really is not strong enough to test hypotheses about how and why these hominin populations may have been different from each other – or indeed, even whether the Turkana hominins from 3.5 million years ago are in biogeographic terms the same population as those from 3.3 million years ago. It is an honest and valid opinion to say that the fossils of hominins from Ethiopia, Kenya, and Tanzania from this era all share some features that reflect a shared ancestry, that they exhibit both within-sample and between-sample variability, and both of those aspects of variation may reflect regional population differentiation, time-successive occupation of the same region by different populations, or speciation.
That doesn’t mean that we need to be agnostic about whether such paleospecies are or were “real”. As Dobzhansky put it, “Classification in biology and in anthropology is as indispensible, and for the same reason, as in a large library: A book misplaced may be as useless as a book lost.” It does matter how we fit a fossil like KNM-WT 8556 into our taxonomy, and the remaining uncertainty is an important cue for us about how far we are from understanding these matters for this period of our evolutionary story.