Crying babies once required vocalization for calming down, but the modern world has found a quicker method. Better be careful not to burn the baby.
The most interesting and important contribution in Dean Falk‘s new book Finding Our Tongues: Mothers, Infants, and the Origins of Language is its discussion of pre-verbal vocalizations that go on between mother and infant during the first year. Thanks to her, I have a deeper appreciation of the complexity of vocalization. It is not just a matter of somehow increasing the human tendency to make noises.
Meanwhile, concerning another post, a visitor to the blog last week (“Uzza”) commented, “You seem to imply a progression, babbling --> words -->language, but since a language must exist as input for babbling, the patterned repetition of phonemes, that would be a circular argument.” I confess to having been startled by this comment as I had never before seen it put so starkly that babbling depends on the prior existence of language. I think Falk would be surprised as well since that argument, if successful, would send her back to the drawing boards. However, thanks to Falk I have a better sense of what’s wrong with Uzza’s bold position.
In a nutshell, the progression from directly from imitative vocalizations to spoken words goes through two steps at once. Humans can sometimes deliberatelycreate and follow a multi-step procedure, but evolution never has a goal or strategy. Each step in a process must first justify itself before a second step can be added to the thing.
Speech requires two things: (1) a knowledge of words, and (2) a voluntary ability to make the sounds of the words. I might be able to learn to recognize some words in a click language, but since I cannot make the click sounds will never speak those words. Thus, even if we accept Uzza’s comment that infants need a language as input before they babble we can see that the process depends on something else—a capacity to imitate the sounds spoken around us.
Can infants make any sounds without first hearing them? Obviously, yes. They do not need any models for them to cry. Falk takes crying much more seriously than most students of language origin and argues:
Crying… may be the missing link in theories about how language is learned and how it emerged in early hominins. [p. 82]
Crying is so natural and, at least originally, so independent of feedback and the need to imitate that deaf babies cry. An article published in 2007 found that deaf infants make a range of cries so that the “hungry cry, sleepy cry, and cry of discomfort or annoyance” are distinguishable. (Reference here; no abstract available.) Deaf children continue to make noises, coos, and babbling sounds, although as Uzza pointed out, they do not make “canonical” sounds. Now that I have read Falk this progression from purely hereditary behavior to mixing innate vocalizing with cultural inputs is in no way surprising. As infants develop, their vocalizations are evermore influenced by the sounds spoken around them, but even if they cannot be influenced by their environment they will continue to make sounds in response to a genetic push.
Notoriously, this mixing of push and culture is not found among chimpanzees, not even those raised from birth in human homes by “parents” who speak to them just like the parents of humans speak to their babies. Why don’t chimpanzees master the sounds? One reason Falk offers startled me.
From a neurophysiological perspective, the production of sound waves from the vocal cords (phonation) and their articulation into speech sounds by the vocal tract are independently controlled, so these processes must become tuned to each other during development. … the two systems begin to synchronize by the time infants are about three months old 
Taking control of the ability to vocalize is natural in human infants and unknown among chimpanzees. It is this absence of control among apes that led Michael Tomasello to argue for the primacy of gesture in the evolution of language. After all, apes do have voluntary control over their hands. (See: Speakers, Listeners, and their Common Ground.)
It seems, however, that the influence of culture on the vocalizations of hearing infants comes well before they babble. Falk reports work that found language influences crying rhythms so much that eventually German-born infants make distinctive German-like sounds in their cries. Crying is not just a “bio-siren.” It begins that way, but turns into a system for making vocalizations match those of the mother and other caregivers.
German-sounding cries do not take on the accent of a movie actor, stamping, “I vant mein mama.” Instead they take on the “melody arcs” of German. A melody arc is “a single rise, then fall of pitch in a sound an infant produces in one expiration of air” . Over time, these melody arcs increase in number during a single breath. This capacity leads to the ability to make longer coos and other soft sounds. Thus, by the age of seven months, when babbling proper—the repetition of syllabic sounds—comes on the scene, hearing babies have been learning to make culturally-matching sounds for over half their lives. And still no words. So why are infants vocalizing like this without using language? Falk argues that
Strong mother-infant bonds were vitally important for prehistoric infants to survive, and as a result, visual, vocal, and physical mechanisms evolved that enabled mothers and infants [quoting Ellen Dissanayake http://www.ellendissanayake.com/] “to enter the temporal world and feeling state of the other.” 
She goes on to say she “pondered … for some years” why these especially strong bonds became necessary and eventually hit upon her putting-the-baby-down hypothesis (discussed last week). The odds of her being right strike me as small. More plausible, though far from certain, is Robin Dunbar’s thesis that vocalizations replaced grooming. But frankly I don’t think we need to guess exactly what problem (or set of problems) vocalization overcame. It is clear that at some point the human lineage became much more cooperative and interdependent than is typical of ape society. Emotional ties were and remain necessary to hold the community together, and the method we hit upon to build and maintain such ties included extensive vocalization.
That solution was not automatic. It required synchronizing the two elements of the vocal system, and an ability to develop matching sounds, or at least sound patterns (“melodies”). Why matching sound patterns? Because the point of the vocalization is to create emotional bonds, and similarity strengthens that sense of connection while difference creates a sense of gap.
Thus, there may never have been a time when human vocalizations were not cultural (i.e., included a learned and shared portion), but there was probably a long time when vocalizations were emotional without being meaningful, that is without piloting attention to some neutral topic.
… Dear me, I’ve still not gotten around to another very important part of Falk’s book, the role of “motherese,” vocalizations that, Falk says, are critical in steering babies from emotional and melodic vocalizations to true speech. I will consider that subject next time.