Eh, if I were a mad scientists with a unaligned AI and a wetlab I'd keep jamming more and more language related genes in animals to see if anything interesting happens. I know there's been some work around the NOVA1 gene that is showing interesting results.