The research background

國

立 政 治 大 學

‧

N a tio na

l C h engchi U ni ve rs it y

3

attended the speech assessments do not have any apparent organic or environmental etiology (Shriberg, Kwiatkowski, Best, Hengst, & Terselic-Weber, 1986).

An alternative approach, which will be the one adopted in this study, for classifying children with speech disorder is the linguistic approach. With this approach, the linguistic characteristics of children’s speech are described primarily in terms of distinctive features, phonological rules (Crocker, 1969; Compton, 1976), or phonological processes (Edwards & Bernhardt, 1973; Grunwell, 1981; Dunn &

Davis, 1983). According to Stampe (1973), a phonological process is a mental operation that applies in speech in the way of substituting a class of sounds that go beyond the speech ability of an individual. Since Stampe has proposed the theory of natural phonology, analyses for phonological processes were widely adopted to examine children’s phonological developments.

Since developmental speech disorder is determined with reference to the phonological developments of normal children, the following section will start with some research background on the phonological developments in normally developing children.

1.1 The research background

Early work on language acquisition in the 1970s has described the language developments of a small number of children (Ingram, 1976). At the time, the

‧ 國

立 政 治 大 學

‧

N a tio na

l C h engchi U ni ve rs it y

4

researchers usually used their own children as subjects and kept diaries that recorded some representatives of the language behavior of the children. However, the small number of subjects makes it difficult to generalize the findings to the whole population; therefore, some large-scale studies, both longitudinal and cross-sectional, were carried out to examine the developmental age and sequence of phoneme acquisition (Prather, Hedrick, & Kern, 1975; Ferguson & Farwell, 1975; Ferguson, 1977; Ingram, 1981; Locke, 1983; Vihman, Macken, Miller, Simmons & Miller, 1985; Bortolini & Leonard, 1991; Dodd, Holm, Zhu, & Crosbie, 2003). Although the age and sequence of phoneme acquisition reported in the literature slightly differ, a similar pattern was found: stops, nasals, and glides are acquired first, followed by liquids, and finally fricatives and affricates (Stoel-Gammon & Dunn, 1985).

In addition to the age of the mastery of sounds, many researchers further investigated the substitution patterns, which involve the underlying phonological processes of consonant production (Edwards, 1974; Eckman, 1977; Irwin & Wong, 1983; Prather, Hedrick, & Kern, 1991; Dodd et al., 2003, among others).

Phonological processes that are not appropriate for children of the age will be regarded as delayed processes (Zhu & Dodd, 2000b).

Based on the findings of the phonological developments in normally developing children, many researchers further conducted comparative studies

‧ 國

立 政 治 大 學

‧

N a tio na

l C h engchi U ni ve rs it y

5

concerning the phonological developments of both normal and disordered children (Stoel-Gammon & Dunn, 1985; Zhu & Dodd, 2000b). Stoel-Gammon and Dunn (1985) found that compared to normal children at the same age, phonologically disordered children have static phonological systems that stop at an early level of development. In addition, an extreme variability on phoneme production, persistent uses of delayed processes and idiosyncratic processes, and restricted use of contrast are all crucial characteristics of phonologically disordered children. Dodd (1993) further proposed a four-subgroup categorization system to classify phonological disorder: articulation disorder, delayed phonological development, consistent disorder, and inconsistent disorder. He found that each subgroup of disordered children shared some similarities on consonant production. These findings all indicate that disordered children display different developmental patterns from that of normal children.

Many studies on normally developing children have reported that Mandarin [tɕ, tɕʰ, ɕ] are early-emerging yet late-acquired phonemes (Zhu & Dodd, 2000a; Hsu, 2003; Cho, 2008). In addition, it is believed that the acquisition data on both normal and disordered children may help understand the nature of affricates and fricatives.

Therefore, the development of these sounds in phonologically disordered children is worth investigating.

‧ 國

立 政 治 大 學

‧

N a tio na

l C h engchi U ni ve rs it y

6

In addition to the consonant acquisition in both normal and disordered children, consonant production in children also attracts many researchers’ attention. Some researchers have raised a question about what factors might influence children’s consonant production. Ingram, Christensen, Veach, & Webster (1980) found that the accuracy rates of word-initial fricatives in normally-developing English-speaking children are reduced in words with longer syllable length. Badar (2002) found that phoneme production consistency tends to decline with increasing syllable complexity. Jeng (2009) further investigated the effect of syllable complexity on consonant production in normal developing Taiwan Mandarin-speaking children, and she found that accuracy rates on the same segment increase in easy syllable structure while decrease in complex syllable structure.

Syllable complexity can be defined in many ways, such as the length of a syllable (Ingram et al., 1980), or a binary distinction between an easy CV structure and other more complex syllable structures (Jeng, 2009). In addition to the above ways of defining the complexity of a syllable, the notion of Sonority Dispersion

Principle (SDP) proposed by Clements (1990), was also used to define the complexity of a syllable in Romani & Calabrese’s (1998) study on the consonant

production in aphasic adult patients. The SDP has provided a better way of defining syllable complexity and will be the one adopted in this study.

‧ 國

立 政 治 大 學

‧

N a tio na

l C h engchi U ni ve rs it y

7

In sum, since previous studies have reported that disordered children usually

have difficulties in acquiring Mandarin palatals, it is worthwhile to have an overall investigation on disordered children’s palatal acquisition. In addition, the effect of

syllable complexity on palatal production in disordered children is also worth investigating because few studies have focused on whether syllable complexity might affect the consonant production of phonologically disordered children.

Investigations on both palatal acquisition and palatal production in disordered children might help parents have a preliminary judgment of identifying developmental speech disorder in their own children, and might also help speech therapists predict what kinds of errors disordered children might frequently make for Mandarin palatals [tɕ, tɕʰ, ɕ] during the process of disorder intervention.