The Neuroscience of Speech, Language & Music | Dr. Erich Jarvis

KEYPOINTS

• 01

  There is probably no separate language module in the brain

  Jarvis argues speech and language are psychology-derived terms that do not map neatly onto brain function. Rather than a language center instructing other areas, the complex algorithms for spoken language are built into the speech production pathway that controls the larynx and jaw, and the algorithms for understanding speech live in the auditory pathway.

• 02

  Understanding speech is common; producing learned sounds is rare

  The auditory perception pathway is widespread in animals, which is why dogs can learn several hundred human words and great apes several thousand, yet cannot say one. The speech production pathway is specialized in humans, parrots, songbirds and hummingbirds, only three of roughly 40-plus bird orders. Koko the gorilla, raised with humans for 39 years or more, learned sign language but could not produce the sounds.

• 03

  Innate vocalizations and learned speech use different circuits

  Most vertebrates vocalize, but with innate sounds such as crying or barking controlled by brainstem and hypothalamic circuits. Learned vocalization uses forebrain circuits. In humans and parrots, the forebrain has taken direct control of brainstem vocal circuits, enabling imitation.

• 04

  Songbirds and humans converged on the same circuits and genes

  Parallels include critical periods, deterioration of vocalizations after deafness, and homologous areas, such as the robust nucleus of the arcopallium and laryngeal motor cortex. The specialized genes, and even similar mutations, overlap despite a common ancestor 300 million years ago. FoxP2 mutations that cause human speech deficits cause similar deficits when induced in vocal-learning birds. Jarvis thinks spoken language arose in our ancestors within the last 500,000 to 1 million years and that Neanderthals likely had some form of it.

• 05

  Speech circuits switch off repulsion genes and turn on neuroprotection

  Jarvis's lab predicted connectivity genes would differ in speech circuits, which have a direct cortical connection to the motor neurons of the larynx or syrinx. Surprisingly, several axon-guidance genes that normally repel connections were turned off, allowing connections that would not otherwise form. Calcium-buffering and neuroprotective genes such as parvalbumin and heat shock proteins were turned up, probably because laryngeal muscles are the fastest-firing in the body, so their control neurons need protection from toxicity.

• 06

  Critical periods are strongest for speech

  The whole brain passes through critical periods, but speech has an especially strong one; feral children who pass puberty without language struggle to acquire it. Jarvis's view is that the brain has limited capacity and stability aids survival. Humans carry an extra copy of SRGAP2, which keeps speech and other circuits more juvenile and plastic throughout life than in other animals.

• 07

  Multilingual children keep more phonemes, easing later languages

  We are born able to produce many phonemes and narrow them to those our language uses. People who learn several languages early retain more of them, so a third or fourth language is easier, not because of greater plasticity but because the sounds are already available.

• 08

  Only vocal learners dance

  Aniruddh Patel and others found that synchronizing body movement to a musical beat is limited to vocal-learning species. Jarvis's motor theory of vocal learning origin holds that speech pathways evolved by duplicating surrounding motor-learning circuits. The tight auditory-motor integration needed for speech then spread to neighboring regions, letting us coordinate whole-body movement to sound. Speech carries more semantic meaning; dance carries mostly affective, bonding communication.

• 09

  Semantic and emotional communication share circuits but differ by hemisphere

  Courtship song and meaningful words use the same song or speech circuits, with emotional centers in areas such as the hypothalamus and cingulate adding tone. The left hemisphere dominates speech, while the right is more balanced for singing and music. All vocal learners use learned sounds emotionally; only a few, including humans, some parrots and dolphins, use them semantically.

• 10

  Reading and writing run through silent speech

  When you read, visual signals travel to Broca's area and you silently speak the words, with tiny electrical signals reaching the vocal muscles. Writing works the same way, so speech bridges thought and the written word. Writing flows best when your hand or typing keeps pace with your inner speech.