Jürgens, Uwe. 2000. [German Primate Center, Göttingen] A comparison of the neural systems underlying speech and non-speech vocal utterances. in Becoming Loquens ed. by Bichakjian BH, Chernigovskaya T, Kendon A, Müller A. Peter Lang: Frankfurt am Main.

"Abstract
Neuroanatomically, the step from genetically determined controlled vocal patterns is associated with the emergence of a direct connection between the motor cortex and the laryngeal motoneurons, a connection lacking in subhuman primates." p. 1

"Learned against innate motor patterns
From this the question arises of why the motor cortex, ventrolateral thalamus, cerebellum, putamen and pyramidal tract are necessary for speech production but dispensable for monkey vocalization. The only explanation we have, at the moment, is that speech consists of learned motor patterns while monkey calls are genetically determined in their acoustic structure, and that the motor cortex, ventrolateral thalamus, cerebellum and putamen are dispensable for the production of innate motor patterns but indispensable for the production of learned motor patterns. Speech clearly represents a learned motor behavior - otherwise the diversity of languages would be difficult to understand." p. 5

"Cortico-motoneural connections
The critical question then is: What is the difference in the neuroanatomical and neurophysiological organization of the motor cortex between monkey and man that could explain why man has voluntary control of phonation and monkey has not. Unfortunately our knowledge about such differences is very poor. One of the very few things we know, however, is that in humans, there is a direct connection between the motor cortex and the nucleus ambiguus, that is, the site where the motoneurons innervating the laryngeal muscles are located. Such a direct connection between motor cortex and laryngeal motoneurons is lacking in the monkey." p. 6