Vertebral numbers and human evolution.

Ever since Tyson (1699), anatomists have noted and compared differences in the regional numbers of vertebrae among humans and other hominoids. Subsequent workers interpreted these differences in phylogenetic, functional, and behavioral frameworks and speculated on the history of vertebral numbers during human evolution. Even in a modern phylogenetic framework and with greatly expanded sample sizes of hominoid species, researchers' conclusions vary drastically, positing that hominins evolved from either a "long-backed" (numerically long lumbar column) or a "short-backed" (numerically short lumbar column) ancestor. We show that these disparate interpretations are due in part to the use of different criteria for what defines a lumbar vertebra, but argue that, regardless of which lumbar definition is used, hominins are similar to their great ape relatives in possessing a short trunk, a rare occurrence in mammals and one that defines the clade Hominoidea. Furthermore, we address the recent claim that the early hominin thoracolumbar configuration is not distinct from that of modern humans and conclude that early hominins show evidence of "cranial shifting," which might explain the anomalous morphology of several early hominin fossils. Finally, we evaluate the competing hypotheses on numbers of vertebrae and argue that the current data support a hominin ancestor with an African ape-like short trunk and lower back.