It’s interesting that one of the San genomes had a non-standard FOXP2 SNP. I am wondering if this can be a sequencing error? Because if it’s not, it’s huge. For it is not just some non-standard SNP but a full blown missense mutation, a potentially drastic one, too – L558P. That position is in the forkhead DNA binding domain, close to protein-DNA interface, likely to be critical for maintaining local structure. It seems to be absolutely conserved in mammals (I checked: monkeys, elephants, hamsters). All that in a protein that has only three amino acid residue differences between mouse and human and it’s either a sequence error or a mutation that ought to have huge functional consequences.