ChapTer 3 Development Over the Life Span 87programs simply adjust the connections among
hypothetical neurons in response to incoming
data, such as repetitions of a word in its past-tense
form. The success of these computer models, say
their designers, suggests that children, too, may be
able to acquire linguistic features without getting a
head start from inborn brain modules (Rodriguez,
Wiles, & Elman, 1999; Tomasello, 2003).
Although debate about the relative impor-
tance of innate factors and learning continues,
both sides agree that acquisition of a first language
must require both biological readiness and social
experience. Children who are not exposed to
language during their early years (such as Genie,
whom we mentioned in Chapter 1), rarely speak
normally or catch up grammatically. Such sad
evidence suggests a critical period in language
development during the first few years of life or
possibly the first decade. During these years, chil-
dren need exposure to language and opportuni-
ties to practice their emerging linguistic skills in
conversation with others. Let’s see how these skills
develop.From Cooing to Communicating
Lo 3.6
The acquisition of language may begin in the
womb. Canadian psychologists tested new-
born babies’ preference for hearing English or
Tagalog (a major language of the Philippines)
by measuring the number of times the babies
sucked on a rubber nipple—a measure of ba-
bies’ interest in a stimulus—while hearing each
language alternating during a 10-minute span.
Those whose mothers spoke only English dur-
ing pregnancy showed a clear preference for
English by sucking more during the minutes
when English was spoken. Those whose bi-
lingual mothers spoke both languages showed
equal preference for both languages (Byers-
Heinlein, Burns, & Werker, 2010).
Thus, from birth, infants are already respon-
sive to the pitch, intensity, and sound of language,
and they also react to the emotions and rhythms in
voices. Adults take advantage of these infant abilities
by speaking baby talk, which psycholinguists call
parentese. When most people speak to babies, their
pitch is higher and more varied than usual and their
intonation and emphasis on vowels are exaggerated.
Parents all over the world do this. Adult members
of the Shuar, a nonliterate hunter- gatherer cul-
ture in South America, can accurately distinguish
American mothers’ infant-directed speech from
their adult-directed speech just by tone (Bryant
& Barrett, 2007). Parentese helps babies learn the
melody and rhythm of their native language.wrong (Dunn et al., 2011; Tomasello, 2003; Evans
& Levinson, 2009). Despite commonalities in lan-
guage acquisition around the world, they main-
tain, the world’s 7,000 languages have some major
differences that do not seem explainable by a
universal grammar (Gopnik, Choi, & Bamberger,
1996). The language spoken by the remote Pirahã
in the Amazon and the Wari’ language of Brazil
apparently lack key grammatical features that oc-
cur in other languages (Everett, 2012), though
this discovery is hotly debated (Nevins, Pesetsky,
& Rodrigues, 2009). One team constructed an
evolutionary history for four major language
groups and found that each group followed its
own structural rules, suggesting that human lan-
guage is driven by cultural requirements rather
than an innate grammar (Dunn et al., 2011). The
anti-innate-module school argues that language
is a cultural tool, comparable to the physical tools
that people have invented in adapting to different
physical and cultural environments. Culture, they
say, is the primary determinant of a language’s lin-
guistic structure, not an innate grammar.
Experience and culture certainly play a large
role in language development. Parents may not
go around correcting their children’s speech all
day, but they do recast and expand their children’s
clumsy or ungrammatical sentences (“Monkey
climbing!” “Yes, the monkey is climbing the tree”).
Children, in turn, often imitate those recasts and
expansions, suggesting that they are learning from
them (Bohannon & Symons, 1988).
Some scientists argue that instead of inferring
grammatical rules because of an innate disposi-
tion to do so, children learn the probability that
any given word or syllable will follow another,
something infants as young as 8 months are able
to do (Seidenberg, MacDonald, & Saffran, 2002).
Because so many word combinations are used
repeatedly (“Pick up your socks!” “Come to din-
ner!”), little kids seem able to track short word se-
quences and their frequencies, which in turn plays
a role in teaching them not only vocabulary but
syntax (Arnon & Clark, 2011). Eventually, children
also learn how nonadjacent words co-occur (e.g.,
the and ducky in “the yellow ducky”), and are able
to generalize their knowledge to learn syntactic
categories (Gerken, Wilson, & Lewis, 2005; Lany
& Gómez, 2008).
In this view, infants are more like statisticians
than grammarians, and their “statistics” are based
on experience. Using computers, some theorists
have been able to design mathematical models
of the brain that can acquire some aspects of lan-
guage, such as regular and irregular past-tense
verbs, without the help of a preexisting mental
module or preprogrammed rules. These computer