Two dredged leg bones now define Denisovan height. Two bones cannot define a species.

A team led by Yousuke Kaifu of the University of Tokyo has put numbers on Denisovan body size: about 1.8 metres (5 feet 11 inches) and 83 kg for one individual, and about 1.9 metres (6 feet 3 inches) and 91 kg for another, using bones dredged from the seafloor of the Penghu Channel in the Taiwan Strait ^[1][5][6]. That turns a species known mainly from genetic sequences taken from fragments such as a small finger bone at Denisova Cave in Siberia's Altai Mountains into one with a body plan, and it does so from two incomplete leg bones whose find-spots were never recorded ^[2][4][9].

The specimens are an incomplete femur and a tibia, attributed to Denisovans through their preserved proteomes rather than DNA, since proteins can retain evidence of evolutionary relationships after DNA is gone ^[4][3]. The proteome signals differed slightly, which is why the authors treat them as two individuals rather than one ^[7]. Two described bones divided between two individuals leaves one bone behind each stature estimate ^[19].

The measurement technique is not the weak link. Long bone length scales well with overall height on average, given a large sample of bones ^[8]. Skilled anatomists can also be reasonably confident about reconstructing the full length of an incomplete bone ^[15]. Neither of those capabilities addresses sample size, and the phys.org analysis makes the point with a crowd in London's Trafalgar Square: picking two people to stand for the whole species would be a tall order ^[16].

The provenance gaps compound it. The bones came up among the fossilised remains of terrestrial ice age mammals, probably deposited when lower sea levels exposed the channel as dry land near the height of the ice age, and their precise discovery locations and geological contexts were not recorded ^[9][10]. Neither bone is directly dated, so both can only be placed broadly in the Middle to Late Pleistocene on the basis of associated animal remains and preservation evidence ^[11]; the Pleistocene as a whole ran from roughly 2.58 million years ago to 11,700 years ago ^[12]. Without sediment context, we do not know whether the two individuals lived at the same time or even in a similar interval ^[13].

That matters for the comparison the study draws. Kaifu and colleagues write that the estimates "are comparable to the largest known Pleistocene Homo from Africa and Europe" ^[14]. The two Penghu figures differ from each other by about 0.1 metres and 8 kg, which is an ordinary gap between two adults ^[20], and they may be separated by an unknown stretch of time. Setting them against the top end of a much larger and better dated fossil record is a comparison of a maximum to a sample of two.

None of this makes the fossils uninformative. Their anatomy and preserved proteins carry real evidence; what has been lost is the contextual information an excavation normally supplies ^[17]. The phys.org piece is blunt that inferring tallness as a species-level trait from two individuals is challenging ^[18], and the Harbin or Dragon Man skull, recovered from a river in northeast China and linked to Denisovans by DNA and protein analysis, shows how quickly this picture is being assembled from isolated, poorly provenanced finds ^[21].

Watch for Denisovan limb bones from stratified excavations, which would let a species-level stature claim rest on something other than dredge spoil, and for any attempt to date the Penghu bones directly.