The origin of articulate language revisited:
the potential of a semi-aquatic past of human ancestors to explain the origin of human musicality and articulate language
Mario Vaneechoutte 2014 Hum.Evol.29:1-33

Articulate language depends on very different abilities, e.g.
- vocal dexterity,
- vocal mimicking,
- the acquirement by children of the very different & arbitrary phonology & grammatical structure of any language.

A vast array of experiments confirm:
children acquire grammar by
- the use of prosodic clues (intonation & pitch),
- facial expression & gesture.
Prosodic clues, provided by speech, are exaggerated in infant-directed speech (motherese).

Strong overlap between musical & linguistic syntactic abilities in the temporal lobes of the brain has been established.
A musical origin of language at the evolutionary level (for H.sapiens) & at the ontogenic level (for each newborn) is parsimonious, and no longer refutable.

Why was song (vocal dexterity & vocal learning) evolved in our species?
Why is it largely absent from other ‘terrestrial’ animals (incl. other primates),
but present in disjoint groups: Cetacea, seals, bats, 3 orders of birds?

We argue:
this enigma (+ a long list of other specifically human characteristics) is best understood by assuming that our recent ancestors (from 3 Ma onwards) adopted a shallow-water diving lifestyle.
The swimming & diving adaptations of the upper airway (& vocal) tract lead to
- increased vocal dexterity & song,
- increased fine-tuning of motoric & mimicking abilities.
These are shared by creatures that can
- freely move in 3 dimensions (swimming, flying) &
- respond instantaneously to behavioural changes of other animals.

Increased bodily mimicking + increased vocal dexterity (both a consequence of a semi-aquatic lifestyle) lead to integrated song & dance, which predisposed to
- producing & mimicking speech & gesture,
- the ability to use prosodic clues to learn the grammar of whichever language.