Relationship Between Expressive Phonological Impairment (EPI) and Phonological Awareness

A growing body of evidence indicates that young children with severe EPI go on to experience problems in literacy. As well, results from another line of research indicate that individuals with reading disabilities evidence more phonological production difficulties (e.g., with multisyllabic words) than their peers with typical reading abilities (Catts, 1986). Bird, Bishop, and Freeman (1995) found that the children who had severe EPI experienced greater difficulty with phonological awareness tasks than their ability-matched peers, even when the tasks did not require a verbal response. Clarke-Klein and Hodson (1995) obtained similar results for spelling. Larivee and Catts (1999), who tested children first in kindergarten and again 1 year later, found that expressive phonology (measured by a multisyllabic word and nonword production task) and phonological awareness scores in kindergarten accounted for significant amounts of variance in first-grade reading.

Several investigators (e.g., Bishop and Adams, 1990; Catts, 1993), however, have reported that phonological impairments alone do not have as great an impact on literacy as language impairments do. A possible explanation for this discrepancy may be the level of EPI severity in the participants in their studies.

Severity Considerations

A common practice in the articulation/phonology literature is to report the number of errors on an articulation test. Not all speech sound errors are equal, however. For example, if two children have 16 errors on the same test, some examiners might view them as equal. If, however, one child evidences a lisp for all sibilants and the other has 16 omissions, the impact on intelligibility will be vastly different. Moreover, the child with extensive omissions might be identified as having a language impairment because of the omission of final consonants (which would affect the production of word-final morphemes on an expressive language measure). Some highly unintelligible children who are considered to have a language impairment may, in fact, have a severe phonological impairment with intact receptive language abilities. Typically such children produce final morphemes as they learn the phonological pattern of word-final consonants.

Phonological Awareness Treatment Studies for Children with EPI

Although there have been numerous studies reporting the results of phonological awareness treatment, only a few investigators have focused on children with phonological or language impairments. van Kleeck, Gillam, and McFadden (1998) provided classroom-based phonological awareness treatment (15 minutes twice a week) to 16 children with speech and/or language disorders (8 in a preschool class and 8 in a prekindergarten class). The small-group sessions focused on rhyming during the first semester and on phoneme awareness during the second semester. The treatment groups and a nontreatment comparison group all made substantial gains in rhyming. Children in the treatment groups, however, made markedly greater gains on phonemic awareness tasks than children in the nontreatment group. Information on changes in expressive phonology or language was not provided by the investigators.

Howell and Dean (1994) used their Metaphon program to provide both phonological awareness and production treatment for 13 preschool children with EPI in Scotland. In phase 1 of this program, children progress from the concept/sound (not speech) level to the phoneme level to the word level. Minimal pairs are used extensively during phase 1. In phase 2, the progression is from word level to sentence level. The children attended between 11 and 34 30-minute sessions weekly. Single subject case study results indicated that the children improved on both phonological production and phonological awareness tasks (sentence and phoneme segmentation).

Harbers, Paden, and Halle (1999) provided individual treatment to four preschool children with EPI for 6–9 months that focused on both feature awareness and production for three phonological patterns that the children lacked. All four children targeted /s/ clusters. Three targeted strident singletons, two targeted velars, two targeted liquids, and one targeted final consonants. The investigators used a combination of the Metaphon (Howell and Dean, 1994) and Cycles (Hodson and Paden, 1991) treatment approaches. Improvement in the production of /s/ clusters coincided with gains in recognizing /s/ cluster features for two of the four children targeting /s/ clusters. Both of the children targeting velars also evidenced concomitant gains in production and awareness. For the remaining targets, there was a slight tendency for the two variables (phonological awareness and production) to move in similar directions, but inconsistencies occurred.

Gillon (2000) conducted a phonological awareness treatment study in New Zealand that involved 91 children with “spoken language impairment” between the ages of 5 and 7 years. Twenty-three children participated in an experimental “integrated” treatment program. A second group of 23 children received traditional speech-sound treatment. Two additional groups served as controls. One treatment group of 15 children who received “minimal” intervention, and the other consisted of 30 phonologically normal children. Children in the first treatment group received two 60-minute sessions per week until a total of 20 hours of intervention had been completed. The second group participated in phoneme-oriented sessions for the same amount of time. All of the children continued participating in their regular classroom literacy instruction, which was based on a “Whole Language” model.

The children in the first group did not receive direct production treatment for EPI during the course of the study. Additional stimulus items for children's individual speech sound errors were integrated into some of the activities, however. The phonological awareness treatment focused on the development of skills at the phonemic level and integrated phonological awareness activities with grapheme-phoneme correspondence training. Activities included (a) picture Bingo and oddity games for rhyme awareness, (b) identification of initial and final sounds, and sometimes medial sounds, (c) phoneme segmentation, (d) phoneme blending, and (e) linking speech to print. The children in this group made significantly greater gains in phonological awareness and reading scores than the children in the other groups. Moreover, the children also made greater gains in phonological production than children in the other groups with EPI. The results of this investigation lend support to the contention that it is important to incorporate phonological awareness tasks into treatment sessions for children with EPI.

Enhancing Phonological Awareness Skills

Available tasks range in difficulty from simple “yes-no” judgments regarding whether two words rhyme to complex phonological manipulation activities (e.g., pig Latin, spoonerisms). Moreover, many activities that are commonly used in treatment sessions have phonological awareness components. When children are taught how a sound is produced and how it feels, they develop awareness about place, manner, and voicing aspects of the sounds in their phonological system. One phonological awareness treatment program (Lindamood and Lindamood, 1998) has a component that specifically addresses teaching the articulatory characteristics of phonemes to all children with reading disabilities, even when there are no phonological production problems. Moreover, when children learn about where a sound is located in a word (initial, medial, or final position), they develop awareness about word positions.

One phonological awareness activity that has proved to be particularly effective is the “Say-It-And-Move-It” task, using Elkonin cards (Ball and Blachman, 1991, adapted from Elkonin, 1963). Children are taught to represent the sounds in one- (e.g., a), two- (e.g., up), or three-phoneme (e.g., cat) words by using manipulatives. Initially blank tiles or blocks are used. Tiles with graphemes are incorporated after the child demonstrates recognition of the sounds for the letters. The top half of the paper has a picture of a word. The bottom half has the appropriate number of boxes for the phonemes needed for the word. Children are taught to say each word slowly and to move one manipulative for each sound into the boxes from left to right.

Another phonological awareness activity that is widely used both for assessment and for segmentation practice is categorization. This task requires matching and oddity awareness skills. Typically the child is given four pictures and is to identify the one that does not match the others in some aspect (e.g., rhyme) and thus is the “odd one out.” Categorization also is used for individual sounds (e.g., initial consonants).

Learning to blend phonological segments to make words is another important task and one that is extremely difficult for some children. Blending tasks commonly start at the word level with compound words (e.g., ice plus cream) followed by blending syllables (e.g., can plus dee; candy). Blending intrasyllabic units (e.g., onset and rime, as sh plus eep; and body and coda, as shee plus p) should precede blending individual phonemes (e.g., sh plus ee plus p).

Another task that has been found to be highly correlated with success in reading is deletion (e.g., elision task, Rosner and Simon, 1971). As with blending, it is important to begin with the larger segments (e.g., compound words). The child says the word (e.g., cowboy), and then, after part of the word is removed (e.g., boy), says the new word (cow). After a child demonstrates success at the larger unit levels, individual phonemes are deleted (e.g., take away /t/ from note/, leaving no).

The task that consistently has accounted for the greatest amount of variance in predicting decoding success is manipulation. Children who are most successful performing phoneme manipulation tasks such as spoonerisms typically are the best decoders (Strattman, 2001). Phonological manipulation in “pattern” songs (e.g., “Apples and Bananas”) seems to be an extremely enjoyable task for very young children and can help them be more aware of sounds and word structures.

Implications for Best Practices

Because children with EPI appear to be at risk for the development of normal reading and writing skills even after they no longer have intelligibility issues, it seems prudent to incorporate activities to enhance phonological awareness skills while they are receiving treatment for phonological production. Moreover, results from Gillon's (2000) study indicate that enhancing phonological awareness skills leads to improvement in phonological production. Thus, enhancing phonological awareness skills appears to serve a dual purpose for children with expressive phonological impairments.