Human phonemes and genomes are thought to have evolved hand-in-glove out of Africa. Several recent studies have attempted to capture a picture of this global variation in languages and peoples, often supporting (and rejecting) a serial founder model (eg. see Atkinson 2011, Perreault and Mathew 2012, Hunley et al. 2012). In a recent large-scale study of microsatellite loci from 246 human populations and phonemic variation across >3000 languages, Creanza et al. (2015) use a PCA (and tests of correlation) to report several interesting patterns in the co-evolution of phonemes and genomes.
Of note are the observations that (1) there is no universal concordance in genetic and phonemic diversities despite most diversity in both were observed in Africa, (2) there exists a strong correlation in differences among populations and geographic distance (and a reduction in number of phonemes with distance from Eurasia), but (3) phonemic diversity is informative of more recent divergence history, rather than of ancient divergence (also indicated by spatial autocorrelation), and (4) a Eurasian origin to all languages analyzed.
Importantly, their analyses indicate that phonemes, unlike genomes, don’t necessarily reflect vertical evolutionary descent, warranting newer models to study the evolution of languages.
Creanza, Nicole, et al. “A comparison of worldwide phonemic and genetic variation in human populations.” Proceedings of the National Academy of Sciences (2015): 201424033. http://dx.doi.org/10.1073/pnas.1424033112