during interactions with caregivers. In the third stage
(at four to six months), vocal play or expansion, sylla-
ble-like productions with long vowels appear.
Squeals, growls, yells, bilabial or labiodental trills,
and friction noises demonstrate infants’ playful ex-
ploration of their vocal tract capabilities during this
stage.
In the extremely important canonical babbling
stage (at seven to ten months), two types of produc-
tions emerge: reduplicated babbling—identical, re-
petitive sequences of CV syllables (e.g., \ma\ma\,
\da\da); and variegated babbling—sequences of dif-
ferent consonants and vowels (e.g., CV, V, VC, VCV
= \ga\e\im\ada). Such productions are not true
words, as they lack meaning. Canonical babbling is
syllabic, containing mainly frontal stops, nasals, and
glides coupled with lax vowels (e.g., \a\, \e\, \o). The
emergence of canonical babbling is highly important,
holding predictive value for future linguistic develop-
ments. Oller and her colleagues in 1999 argued that
babies who do not produce canonical babbling on
time are at high risk for future speech and language
pathology, and should be carefully evaluated by a lan-
guage clinician.
In the fifth stage (at twelve to thirteen months),
jargon or intonated babble, infants produce long
strings of syllables having varied stress and intonation
patterns. Jargon sounds like whole sentences convey-
ing the contents of statements or questions, and often
co-occurs with real words. Yet, it lacks linguistic con-
tent or grammatical structure.
Pre-Speech Vocalizations in Different
Target Languages
The early interpretation of similarities in the
phonetic structure of babbling among infants who ac-
quire different languages (e.g., Japanese, Hebrew)
was that pre-speech vocalizations are universal. This
observation was explained by the strong constraints
of the mouth’s anatomical characteristics and by phys-
iological mechanisms controlling movements of the
tongue and palate. Cross-linguistic research in the
1990s revealed, however, that clear influences of seg-
mental and suprasegmental patterns (i.e., intonation
and stress) of the input are recognizable in pre-
speech vocalizations. This is particularly true during
the second half of the first year of life. In a longitudi-
nal comparative study by Bénédicte de Boysson-
Baradis (1999) of ten-month-old Spanish, English,
Japanese, and Swedish infants, the relative distribu-
tion of consonants in their canonical babbling resem-
bled the distribution of these segments in their
language. As babies grow, the segmental similarity be-
tween their babbling and early words increases. Sever-
al studies by Peter Jusezyk and colleagues on speech
perception indicate that infants’ sensitivity to the
acoustics and phonetics of languages increases with
age, influencing their ability to discriminate the se-
quences of sounds and syllable structures typical to
their own language. Indirect evidence for the role of
audition in the development of pre-speech vocaliza-
tions derives from studies on deaf children, who show
significant delays in the emergence of canonical bab-
bling and also a decreased variety of consonants ut-
tered from age eight months onward.Mutual Imitation within Mother-Child
Interaction
In 1989 Metchthild and Hanus Papouˇsek were
among the first researchers to point out that more
than 50 percent of two- to five-month-olds’ noncrying
vocalizations are either infant imitations of mothers’
previous vocalizations or mothers’ imitations of in-
fants’ previous vocalizations. They suggested that this
mutual vocal matching mechanism relates to the
emotional regulation of communication in the begin-
ning of life. Joanna Blake and Bénédicte de Boysson-
Bardies found in 1992 that infants tend to vocalize
more while manipulating small objects and especially
when adults are present. Edy Veneziano in 1988 ana-
lyzed vocal turn taking in pairs of nine- to seventeen-
month-old babies and their mothers. She reported
that, as children advance toward conventional lan-
guage, mothers’ imitations of what babies say be-
comes selective. Mothers imitate only those infant
vocalizations resembling conventional words, thus
signaling to the child what constitutes a linguistic
symbol with meaning.Pre-Symbolic Productions in Hearing and
in Deaf Infants
Cumulative research on pre-speech vocalizations
clearly indicates that babbling is in fact structurally
and functionally related to early speech. Locke ar-
gued in 1996 that when variegated babbling emerges,
a consistent relation is identified between vocaliza-
tions and specific communicative functions (i.e., pro-
test, question, and statement). At around age
eighteen months, the child’s phonological system is
clearly shaped by the target language’s phonetic char-
acteristics, and at that time conventional words
emerge.
Indirect evidence for the developmental signifi-
cance of babbling was published in a revolutionary
1991 paper by Laura Petitto and Paula Marentette on
hand babbling in two deaf infants of signing mothers.
The argument was that these two infants (who were
recorded at ages ten, twelve, and fourteen months)46 BABBLING AND EARLY WORDS