Typical Development of Prelinguistic Vocalization
Patterns Across Languages
Bilinguistics, Inc.
This course is offered by Bilinguistics, an ASHA Approved CE Provider
Content Area: Basic Communication Processes
Instructional Level: Intermediate
Continuing Education Units: .1
Objectives:
Participants will be able to demonstrate knowledge in and identify:
•
Basic acquisition of speech sounds at the pre-linguistic level
•
Sound acquisition norms that are universal to all languages
•
Common consonant and vowel patterns that are produced during babbling
•
An understanding of suggested stages of sound acquisition
•
Frequency of initial consonant use and consonant-vowel co-occurrence
Copyright © 2007 Bilinguistics All Rights Reserved
1
Introduction
The development of the sound system leading up to the production of words has
been well documented (Oller, 1980; Stark, 1980; Vihman, Macken, Miller, Simmons, &
Miller, 1985; Vihman, Ferguson, & Elbert 1986). Data exist concerning the order in
which speech sounds and sequences emerge (Davis, & MacNeilage, 1995; Elbers, 1982),
the make-up of sound segments (Nathani & Oller, 2001; Mitchell & Kent, 1990), as well
as the function of the speech mechanism when producing sounds (Davis & MacNeilage,
1995; MacNeilage, Davis, Kinney, & Matyear, 2000). An understanding of typical
speech development is essential in order to identify the differences in early vocal
development when a disorder is present (e.g. loss of hearing), and to understand when
developmental processes are influenced by a second language.
For the purpose of this paper, speech development will be categorized and defined
as, pre-linguistic vocalizations and canonical babbling. Pre-linguistic vocalizations refer
to vegetative and non-vegetative sounds. Canonical babbling refers to patterned,
minimally consonant-vowel (C-V) combinations such as CV, VC, or CVCV that meet the
rhythmic requirements for sounding speech-like (Oller, 1980). Canonical babbling will
include all vocalizations traditionally defined as reduplicated and variegated sequences
meeting the timing requirements for being perceived as speech-like.
The Emergence of the Sounds of Speech
The complexities of speech find their roots in the simplified sounds produced by
infants as young as six months of age (Nathani & Oller, 2001). Previously, it was
accepted that infant vocalizations that are recognizable as speech were not connected to
sound patterns in first words. However, this idea of ‘discontinuity,’ as first suggested by
Copyright © 2007 Bilinguistics All Rights Reserved
2
Jakobson in 1941, has been refuted by recent research showing continual development
from babbling through to speech (Vihman et al., 1985). In both babbling and early
words, stop consonants (e.g. /b/,/d/) are frequent and there is an absence of consonant
clusters (Nathani & Oller, 2001).
With research on the emergence of sounds in speech in infancy, Oller (1980)
created a hierarchy of pre-linguistic speech development that continues to be widely
accepted. His goal was to include the many characteristics of spoken languages and
provide a way for any infant sound to be described in terms of crucial speech parameters
(Oller, 1980). He realized this by analyzing aspects of speech such as pitch, tone, quality,
and timing instead of relying heavily on phonological transcription (Oller, 1980). Earlier,
empirical research in the 1940s (e.g. Jakobson) failed in its attempt to find relations
between adult and infant speech patterns due to strict use of phonetic transcription
(Nathani & Oller, 2001). Previously, speech-like (i.e. ba) and non-speech-like (i.e. a
burp) sounds were collapsed into the same category using the International Phonetic
Alphabet (IPA), which made any difference or similarity indistinguishable. Therefore,
Oller analyzed speech based on larger properties and defined five stages of speech
development occurring in the first year of life.
Copyright © 2007 Bilinguistics All Rights Reserved
3
Table 1: The Emergence of the Sounds of Speech in Infancy.
Stage
Stage 1:
The Phonation Stage
Age (in months)
Type
0-1
Vowels, consonants, throaty sounds, and
phonation with a closed mouth
Stage 2:
The Goo Stage
Stage 3:
The Expansion Stage
2-3
Velar (/k/ and /g/) consonants
4-6
Strong vowels, bilabial trills, squeals,
growling
Stage 4:
The Canonical Stage
7-10
Patterned, consonant-vowel combinations,
reduplicated babble (mama), and nonreduplicated (ada)
Stage 5:
The Variegated Babble
Stage
11-12
Diverse babble (bada), gibberish, intonation
Note: Adapted from Oller (1980).
Speech development begins when babbling takes on adult-like prosody and form
(Nathani & Oller, 2001). Babbling, as defined by Oller (1980), begins at a stage where
consonant and vowel sequences conform to rhythmic patterns perceived as speech-like in
timing. Prior to babbling, the precanonical sound qualities such as grunts, squeals, and
gooing (Stages 1,2, & 3), are most likely reflexive or accompany distress. These sounds
do not follow any linguistic rules and do not make use of the full potential of the vocal
tract (Nathani & Oller, 2001).
En route to producing discernable words, canonical babbling begins at around
seven months (Oller, 1980). The timing characteristics of a child’s speech become
relatively rigid and syllable structure begins to resemble mature language. Productions
are either non-reduplicated (/abid/) or reduplicated (/mama/) and may be perceived by
adults as words (Oller, 1980).
The progression from canonical to variegated babble marks a giant shift from
prelinguistic to linguistic behavior (Mitchell & Kent, 1990). Repetitive sound clusters
Copyright © 2007 Bilinguistics All Rights Reserved
4
diminish and there is an increase in the variety of consonants and vowels within
utterances (Mitchell & Kent, 1990). Perceivable changes in intonation (Elbers, 1982 )
are accompanied by phonetically-varied reduplicated (dudududu) and non-reduplicated
(dudi) forms (Mitchell & Kent, 1990). Diphthongs emerge as well as jargon that
resembles the ambient language (Elbers, 1982).
While Oller’s explanation of emerging sounds is widely accepted, some
researchers feel that speech development cannot easily be divided and categorized. Stark
(1980) suggested that so much development is occurring simultaneously during the
canonical and expansion stages (Oller Stages 3 and 4) that linguistic development around
twelve months cannot be divided. In her assessment of speech development within the
first year (See following table), Stark (1980) demonstrated that speech development
could be illustrated logically using four stages. Stages 1 through 3 mirror the work done
by Oller (1980) but stage 4 is a combination of the canonical and variegated babble stage.
In Stage 4, the infant is making use of CV patterns, reduplicated and non-reduplicated
babble, gibberish and intonation. There is not one fixed point in time where one process
ends and another replaces it. Stark argues that there is a transition between vocalizations
and employing sounds in a speech-like manner (Stark, 1980).
Copyright © 2007 Bilinguistics All Rights Reserved
5
Table 2: Speech Development in the First Year.
Stage
1
Class
Vegetative function
Type
Crying, fussing, nasal (closed mouth)
phonation
2
Cooing sounds
First voicing, pleasure, laughter, nasal
sounds, velar sounds
3
Vocal play
Vocalic and consonant elements,
frication, nasal murmurs
4
Mastery of speech-like activity
Babbling, controlled CV transitions,
reduplicated and non-reduplicated sounds
Note: Adapted from Stark (1980).
Babbling is the culmination of development that precedes meaningful speech
(Vihman et al. 1985). In a typically developing child, babbling begins around 7 to 10
months and speech emerges from this initial use of consonant and vowel structures.
However, in a child who is hearing-impaired, development is not only delayed but may
have a unique pattern. When investigating these differences in a child with hearing
impairment, one must first understand why speech develops in typically developing
children.
Theories on Speech Development
Theories vary widely accounting for how and why the speech mechanism
develops to eventually produce meaningful speech. While it is now generally accepted
that sound qualities in canonical babbling are continuous with sound qualities in the early
word period (Vihman et al. 1985), there are varying theories as to why pre-verbal
vocalizations and canonical babbling take place.
Copyright © 2007 Bilinguistics All Rights Reserved
6
In order to account for babbling’s influence on speech, a theory must take into
account the reduplicated, dependable nature of babbling, the maturation of the sound
system in proportion to growth, the adult-like speech it imitates and then becomes, and
the universal nature of babbling in its initial stages. Davis et al. (2002) address these
issues biologically in terms of the frame/content theory. The ‘content’ (consonants and
vowels) of an utterance is placed within a ‘frame’ (oscillation of the mandible). The
movement of the frame is believed to have emerged from reflexive motions of ingestion
such as eating and sucking. Vowels (V) evolve from the open aspect of mandibular
movements where as the closing phase creates a consonant (C) (Davis et al. 2002).
Canonical babbling, once again, is a sequence of successive syllables, and is
characterized by both reduplication and variegation in vocal sequences (Stark, 1980;
Elbers, 1982; Oller, 1980). In tying frame/content to babbling’s relation to speech,
rhythmic open and closed states (CV or VC) create a temporal regularity that is perceived
by adults as “speech-like” (MacNeilage & Davis, 1990).
As maturation changes the physical articulatory structures, babbling is enhanced.
In canonical babbling’s initial stage, the tongue and lips are seldom employed, relying on
the mandible for the full range of motion (Davis & MacNeilage, 1995). “Changes in
amplitude of the mandibular cycle would result in height changes for vowels (e.g. [ae]
verses [i]) and manner changes for consonants (e.g., [d] versus [j]) (Davis & MacNeilage;
3, 1995).” Alternations between oral and nasal sounds appear when control is gained
over soft palate closure. The basis of babbling exists within the timing and development
of the action and not in descriptions of phonological representations (Davis &
MacNeilage, 1995).
Copyright © 2007 Bilinguistics All Rights Reserved
7
Biological assertions relating physical movements to speech cannot be language
specific or dependent. That is to say, results should be universally challengeable and
provide similar outcomes across an array of ambient languages. MacNeilage et al. (2000)
suggest that in early development an infant babbles in similar patterns regardless of
ambient language. In their study on serial organization in infants across languages,
MacNeilage and colleagues (2000) explore consistencies in babbling productions and
compare data across languages. Six language populations (English, French, Swedish,
Japanese, Brazilian Portuguese, & Quichua) and one infant with deafness who had
received a cochlear implant were studied, yielding babbling patterns that contain some
universal consistencies (MacNeilage et al. 2000).
The frame/content theory suggests that it is initially important to understand the
frame (mandibular oscillation) that establishes the rhythmic alteration between
consonants and vowels. The content (vowel and consonant) of each oscillation can then
be scrutinized separately and quantified in syllables (CV, VC) that each content pair
forms (MacNeilage et al. 2000). Consistencies in babbling have been reported across
different language populations and one implanted child (MacNeilage et al. 2000). These
consistencies are also present when hearing-impairment and language differences are
combined (Moore et al. 2007).
Speech Content
The fundamental questions for interventionists relate to what vocalization patterns
occur at what age relative to typical developmental expectations, as well as what might
be the cause if a child is not producing these vocalization patterns. These questions are
addressed by identifying the type, number, and developmental sequence of vowels and
Copyright © 2007 Bilinguistics All Rights Reserved
8
consonants within a child’s repertoire. Independent research studies have described the
content and framework of early speech and collectively offer a greater understanding of
babbling and its developmental sequence.
Vowels
Vowels exceed the number of consonants produced during babbling for most
children (Davis & MacNeilage, 1995). While no vowel pattern characterizes every child
(Davis & MacNeilage, 1995), certain tendencies exist universally. Children who are
babbling have a propensity to use mid-front (head), low-front (had), and central vowels
(hut)(Davis & MacNeilage, 1995). As illustrated in the vowel chart below, a vowel’s
designation is derived from the manner (height) in which the jaw is open and the position
of the tongue (front to back) when the vowel is produced. There is a tendency for
children to begin producing front vowels and shift back in the mouth as maturation
occurs (Oller, 1980). Low back vowels are not often produced until later in development
(Oller, 1980). Although vowels represent a greater portion of an infant’s utterances, it is
the consonant that is usually interpreted as carrying the speech message.
Table 3: Vowel Chart
Indicates direction of
acquisition over time.
Indicates most common
vowels found in babbling.
Copyright © 2007 Bilinguistics All Rights Reserved
9
Consonants
Like vowels, the first consonants that are produced are guided by the constraints
of physical development. Stop consonants (/b, p, t, d, g, k/) and nasal consonants (/m, n,
ŋ/) are the most common that are produced during babbling across languages (Vihman,
1985; Locke, 1983; Davis & MacNeilage, 1995). The production of these sounds is
physiologically similar in that they are all produced with a total occlusion of the oral
cavity. The nasal consonants differ in that the nasal passage remains open during the
sound’s production (Davis & MacNeilage, 1995). Hearing children produce mostly oral
consonants by the onset of babbling (Davis & MacNeilage, 1995)
Consonant frequencies are determined by how easily the sound can be produced
as well as where in the oral cavity it occurs. Infants initially have a greater facility with
front consonants such as the alveolars (/d, t/), and labials (/m, b, p/) (Locke, 1983). The
alveolar /d/ is found to be the largest category present in a study of six infants by Davis
and MacNeilage in 1995. Nasals (/m, n, ŋ/) and glides (/w, j/), occur less frequently than
stop consonants (Locke,1983; Davis & MacNeilage, 1995), and velars (/g, k/), produced
by the tongue and the back of the throat, have very little incidence. Complex sounds such
as liquids (/l, r/) and consonant clusters are infrequent or are not present (Vihman, 1985).
Locke’s study (1983) of consonant frequency illustrates the prevalence of
consonants within an infant’s sound repertoire (See Table 4). He found all six stop
consonants to be frequent, both glides, as well as an appearance of the first fricative (/s/).
This study additionally provides information about the role that voicing plays in
babbling. The more common sounds, with the exception of the glottal /h/, were all
voiced. Within each minimal pair (two sounds that share the same articulatory
Copyright © 2007 Bilinguistics All Rights Reserved
10
movements) the voiced sound was normally present more often (Locke, 1983). What
Davis and MacNeilage (1995) and Locke (1983) suggest bases phonemic development on
maturation. Theoretically, consonant inventories and prelinguistic vocalizations of this
study’s participant should be commensurate with these previous findings independent of
ambient language.
Table 4: Consonant Frequency. This chart illustrates the percentage that time
that each consonant was present during sound sampling.
More Frequent Consonants
Less Frequent Consonants
Sound
A
Sound
A
h
31.77%
v
1.03%
d
20.58
l
0.96
b
9.79
θ
0.85
m
6.69
z
0.56
t
4.34
f
0.37
g
4.15
ʃ
0.37
s
3.45
ð
0.34
w
3.39
ŋ
0.33
n
2.65
Ʒ
0.1
k
2.12
r
0.1
j
1.77
ƫƒ
0
p
1.63
ʤ
0
Totals
92.33%
Totals
5%
Note: From Locke (1983).
Copyright © 2007 Bilinguistics All Rights Reserved
11
Consonant Vowel Co-occurrences
The combination of vowels and consonants create the syllable patterns that are
often perceived as speech. Early in babbling, three common patterns appear to be
universal (MacNeilage & Davis, 2000). Most often, labial consonants occur with central
vowels (/bΛ/), alveolars occur with front vowels (/dæ/), and velars occur with back
vowels (/gu/) (Davis & MacNeilage, 1995). In the case where labial consonants occur
with central vowels, it is known as a “pure frame” (MacNeilage et al., 2000, MacNeilage
& Davis, 2000). It is pure in the sense that the sound is solely the result of mandibular
movement. The consonant is formed as the lips come together and the vowel is formed
while the tongue is in a resting position (MacNeilage & Davis, 2000). Figure 2 illustrates
the oral positions of these CV co-occurrences.
Syllables appear in an infant’s repertoire as monosyllables (Vihman, 1985).
Normally these syllables begin with a consonant and end with a vowel (MacNeilage,
Davis, Kinney, & Matyear, 2000). MacNeilage et al. (2000) state that the CV sequence is
so important that it is “often given the status of the only universal syllable type”
(MacNeilage et al.; 154, 2000). Following the appearance of these syllables in isolation,
an infant begins to reduplicate (Oller, 1983).
Copyright © 2007 Bilinguistics All Rights Reserved
12
Table 5: Consonant Vowel Co-occurrence Constraints in Babbling.
Note: From MacNeilage & Davis, (2000).
Reduplicated syllable production (babbling) is the result of repeated oscillations
of the mandible (Davis & MacNeilage, 1995). Mandibular movements appear earlier
than lingual or labial movements in canonical babbling (Davis & MacNeilage, 1995,
Davis et al., 2002) because oscillation is easiest when the articulators stay in one place
(Davis & MacNeilage, 1995; MacNeilage and Davis, 2000; MacNeilage, 2000). Initially,
50% of the total syllable is reduplicated and 67% of the consonants are reduplicated
(Davis and MacNeilage, 1995). As the child starts to produce his or her first words, 30%
of the total syllables are reduplicated and 73% of the consonants are reduplicated
(MacNeilage, Davis & Matyear, 1997). As variegated babble leads into full words, there
is a very little reduplication in the child’s attempts at speech (MacNeilage, Davis,
Kinney, & Matyear, 2000).
Copyright © 2007 Bilinguistics All Rights Reserved
13
Table 6: Observed-to-Expected Frequencies of Consonant-Vowel Co-occurrence.
Consonants
Vowels
Coronal
Labial
Dorsal
Front
B
1.28
0.57
0.95
C
1.21
0.85
0.89
W
1.18
0.75
0.66
B
0.84
1.34
0.96
C
0.86
1.27
0.89
W
0.85
1.2
1.1
B
0.64
1.22
1.22
C
0.85
0.79
1.84
W
1.08
0.76
1.24
Central
Back
Note: B = prespeech babbling, C = babbling concurrent with first words, W = first words.
Taken from: MacNeilage et al.; p.155 (2000).\
Summary and Conclusion
In conclusion, prelinguistic babbling is the biological result of mandibular
oscillations that are accompanied by phonation (Davis & MacNeilage, 1995; MacNeilage
and Davis, 2000; MacNeilage, 2000). Many of these syllable patterns that are created
during babbling are universal, without influence of ambient (native) language
(MacNeilage, 2000). This being said, an explanation of speech development, even for a
Copyright © 2007 Bilinguistics All Rights Reserved
14
Spanish-speaking child, for example, would be generic up until babbling makes the
transition into first words (Oller, 1980; Locke, 1983).
References
Davis, B.L., MacNeilage, P.F., & Matyear, C.L. (2002). Acquisition of serial complexity
in speech production: a comparison of phonetic and phonological approaches to
first word productions. Phontica, 1-33.
Davis, B.L., & MacNeilage, P.F. (1995). The articulatory basis for babbling. Journal of
Speech and Hearing Research, 38(6), 1199-1211.
Elbers, L. (1982). Operating principles in repetitive babbling: a cognitive continuity
approach. Cognition, 12, 45-63.
Locke, J.L. (1983). Phonological acquisition and change. New York: Academic Press.
MacNeilage, P.F., Davis, B.L., Kinney, A., & Matyear, C.L. (2000). The motor core of
speech: A comparison of serial organization patterns in infants and languages.
Child Development, 71(1), 153-163.
MacNeilage, P.F., Davis, B.L., & Matyear, C.L. (1997). Babbling and first words:
Phonetic similarities and differences. Speech Communication, 22, 269-277.
Mitchell, P.R., & Kent, R.D. (1990). Phonetic variation in multisyllable babbling.
Journal of Child Language, 17, 247-265.
Moore, J.A., Prath, S., Arrieta, A. (2007). Early Spanish Speech Acquisition Following
Cochlear Implantation. The Volta Review, 106(3), 321-341.
Nathani, S., Oller, K. (2001). Beyond ba-ba and gu-gu: challenges and strategies in
coding infant vocalizations. Behavior Research Methods, Instruments and
Computers, 33(3), 321-330.
Copyright © 2007 Bilinguistics All Rights Reserved
15
Oller, D.K. (1980). The emergence of the sounds of speech in infancy; Chapter 6. In G.
Yeni-Komshian, J.F. Kavanagh, & C.A. Ferguson (eds.), Child Phonology: Vol.
1. Production (pp.93-112). New York, Academic Press.
Stark, R.E. (1980). Stages of speech development in the first year of life. In G. YeniKomshian, J.F. Kavanagh, & C.A. Ferguson (Eds.), Child Phonology: Vol. 1.
Production (pp.93-112). New York, Academic Press.
Vihman, M.M., Macken, M.A., Miller, R., Simmons, H., & Miller, J. (1985). From
babbling to speech: A re-assessment of the continuity issue. Language. 61(2),
397-445.
Vihman, M.M., Ferguson, C.A., & Elbert, M. (1986). Phonological development from
babbling to speech: common tendencies and individual differences. Applied
Psycholinguistics, 7, 3-40.
Copyright © 2007 Bilinguistics All Rights Reserved
16