Perceptual and Motor
Skills, 1975,40, 119-125. @ Perceptual and Motor Skills 1975
IMMEDIATE RECALL OF SEMANTICALLY VARIED "SENTENCES" BY LEARNING-DISABLED ADOLESCENTS ELlSABETH H. WIIG AND MARY ANNE ROACH Borton University Summary.-The immediate recall of 20 semantically and syntactically varied sentences was assessed and compared for 30 learning-disabled and 30 academically achieving adolescents. Learning-disabled adolescents repeated significantly fewer of the sentences verbatim than their achieving age peers. They exhibited significant reductions in the recall of sentences which violated semantic (selectional) rules, contained correctly and incorrectly sequenced modifier-strings, contained a random word-string, or were syntactically complex. Perseveration errors occurred more frequently among the learning-disabled adolescents than among the achievers and inter-sentence perseverative errors were exhibited only by those who were learning disabled. The rank order of difficulty for the sentences agreed for the two groups, suggesting primarily quantitative reductions in the immediate recall by the learning-disabled adolescents. The findings suggest that learning-disabled adolescents depend heavily upon semantic aspects for language processing, experience immediate memory and sequencing problems for modifier-strings, and exhibit a prevalence of interfering perseverative responses.
Learning-disabled children characteristically exhibit deficits in processing sentences which suggest that the language system has not been internalized. Morphological and syntactic deficits observed in such children may reflect limited auditory memory, limited ability to apply morphological and syntactic rules in sentence processing, and difficulties with logical operations and the conceptual synthesis underlying adult syntax (Farnham-Diggory, 1967; Lerner, 1971; Menyuk, 1964; Menyuk & Looney, 1972a, 1972b; Rosenthal, 1970; Wiig, et al., 1973; Wiig & Semel, 1973). There continues to be little research on languageprocessing abilities of adolescents with specific learning disabilities. Wiig and Semel (1974) observed quantitative reductions in the comprehension of linguistic concepts requiring logical operations by learning-disabled adolescents. The question remained whether such adolescents would exhibit significant deficits in the imrnediare recall of semantically and syntactically varied "sentences" when compared with academically achieving peers. Sentence-repetition tasks have been widely used to assess the linguistic competence of children and adults (Brown & Bellugi, 1964; Fraser, et al., 1963; Luterman & Bar, 1971; Menyuk, 1964, 1969; Menyuk & Looney, 1972a, 1972b; Newcombe & Marshall, 1967). Experimental evidence consistently indicates interactions between grammatical structure and semantic aspects and memory for sentences. These are interpreted by Slobin ( 1971) ro indicate chat meaning and form can be stored independently and that meaning (deep structure) is more persistent in memory than form (surface structure). Miller and Chomsky (1963) suggest that the surface structure of a sentence is computed within short-term memory and that the limitations of this memory relate directly to the
120
E. H. WlIG
6:
XI A. ROACH
transformational structure of language. Deep structure and semantic interpretation are considered to be derived within a larger and longer-term memory score. Sentence-repetition studies of language-disordered children have indicated that sentence length and structure constitute significant variables in recall (Menyuk, 1964, 1969; Menyuk & Looney, 1972a, 1972b). Newcombe and Marshall (1967) extended the analytic possibilities of a sentence-repetition task by varying semantic and syntactic constraints. Their research indicated that deficits in recall of "sentences" distinguished left-hemisphere-damaged from right-hemisphere-damaged and non-brain-damaged adults. The present study employed Newcombe and Marshall's "sentences" to assess and compare the effects of differences in semantic and syntactic constraints on immediate recall by learningdisabled and academically achieving adolescents.
Ss were 30 adolescents, 7 females and 23 males, with specific learning disabilities diagnosed by a psycho-educational team. They ranged in age from 12 yr., 5 mo. to 16 yr., 4 mo. ( M = 14 yr., 2 mo.) and attended middle-class suburban public schools in Grades 7, 8, and 9. WISC Full-Scale IQs ranged from 87 to 129 ( M = 9 6 ) . All scored within k 3 mo. of CA on the Peabody Picture Vocabulary Test (Dunn, 1959). Achievement scores in grade equivalents on the Metropolitan Achievement Tests (Durost, 1961 ) ranged from 2.7 to 6.7 (M = 4.5) for Reading, from 3.1 co 4.7 ( M = 4.1) for Language, and from 2.1 to 5.3 ( M = 4.2) for Mathematics. All exhibited academic retardation of at least 1.5 grade equivalent scores in two or more academic areas, normal auditory acuity and articulatory ability, and none had received speech or language therapy. The controls were 30 academically achieving adolescents from a middleclass suburban public school. They were matched for sex, grade level, age ( M = 14 yr., 1 mo.), and IQ (Range 88 to 130; M = 103). All scored within -+3 mo. of CA on the Peabody test and grade equivalents on the Metropolitan ranged from 7.5 to 9.5 ( M = 8.3). None received speech or language therapy and all had normal auditory acuity. Materialr and Procedrcre Intelligence and achievement were evaluated during the testing. The experimental "sentences" (Newcombe & Marshall, 1967) were read to all Ss individually, in random sequence to assess immediate recall. Ss were first given four practice sentences which included a random word-string and sentences with varied semantic and syntactic constraints. They were requested to repeat each sentence immediately after presentation and the instructions stressed that verbatim recall was required. Individual responses were recorded to permit analysis of
RECALL OF
SENTENCES BY
LEARNING DISABLED
12 1
errors. The number of correct, verbatim repetitions provided the raw data for statistical tests of the significance of mean differences and of differences in the responses to the various sentences.
RESULTS The 30 learning-disabled adolescents repeated 410 sentences verbatim (M = 13.67; SD = 0.38; range 8 to 20) while the academically achieving controls repeated 514 sentences verbatim (M = 17.12; SD = 1.94; range 1 3 to 20). The mean difference proved significant ( t = 2.12, p < .05), indicating that the learning-disabled adolescents made significantly more errors in repeating the sentences than their controls. The number of verbatim repetitions of each of the sentences by the learningdisabled and achieving adolescents were subsequently compared (Table 1). The former group made significantly more errors on ( a ) Sentences 2, 16, 19, and 20 which were syntactically well formed, but violated semantic (selectional) rules, ( b ) Sentences 11 and 13 which contained modifier-strings in correct and incorrect order, ( c ) Sentence 15 which contained a random word-string, and ( d ) Sentence 16 which was structurally complex with 'embedding.' TABLE 1
COMPARISON OF VERBATIM REPETITIONS BY 30 LEARNING-DISABLED AND 30 ACADEMICALLY ACHIEVING ADOLESCENTS "Sentence"
1. 2. 3. 4. 5. 6. 7. 8.
9. 10. 11. 12. 13. 14. 15. 16. 17.
The team of workers built the bridge. Colorless green ideas sleep furiously. The man posted the letter. The boy hit the girl. The politician nearly lost the election. The boy easily passed the examination. Didn'r the mechanic repair the van? Didn't the lion chase rhe tiger? Wasn't the stone wall built by the kind husband? Wasn't the rich uncle advised by the nice manager? She has bought five large brown leather cases. H e has sold the long heavy grey shiny car. She has washed plastic red small eight cups. Not in a tree to the lake ran with. Walk some by hard of clearly table very. The sky that the dream thought jumped cheaply. The burglar that the police found escaped easily. The chair roughly painted the fire. Wasn't the fat ceiling robbed by the tired pen? The man that the book read was interesting. .O5; t p .01; Sp .001.
18. 19. 20. *p
Skills, 1975,40, 119-125. @ Perceptual and Motor Skills 1975
IMMEDIATE RECALL OF SEMANTICALLY VARIED "SENTENCES" BY LEARNING-DISABLED ADOLESCENTS ELlSABETH H. WIIG AND MARY ANNE ROACH Borton University Summary.-The immediate recall of 20 semantically and syntactically varied sentences was assessed and compared for 30 learning-disabled and 30 academically achieving adolescents. Learning-disabled adolescents repeated significantly fewer of the sentences verbatim than their achieving age peers. They exhibited significant reductions in the recall of sentences which violated semantic (selectional) rules, contained correctly and incorrectly sequenced modifier-strings, contained a random word-string, or were syntactically complex. Perseveration errors occurred more frequently among the learning-disabled adolescents than among the achievers and inter-sentence perseverative errors were exhibited only by those who were learning disabled. The rank order of difficulty for the sentences agreed for the two groups, suggesting primarily quantitative reductions in the immediate recall by the learning-disabled adolescents. The findings suggest that learning-disabled adolescents depend heavily upon semantic aspects for language processing, experience immediate memory and sequencing problems for modifier-strings, and exhibit a prevalence of interfering perseverative responses.
Learning-disabled children characteristically exhibit deficits in processing sentences which suggest that the language system has not been internalized. Morphological and syntactic deficits observed in such children may reflect limited auditory memory, limited ability to apply morphological and syntactic rules in sentence processing, and difficulties with logical operations and the conceptual synthesis underlying adult syntax (Farnham-Diggory, 1967; Lerner, 1971; Menyuk, 1964; Menyuk & Looney, 1972a, 1972b; Rosenthal, 1970; Wiig, et al., 1973; Wiig & Semel, 1973). There continues to be little research on languageprocessing abilities of adolescents with specific learning disabilities. Wiig and Semel (1974) observed quantitative reductions in the comprehension of linguistic concepts requiring logical operations by learning-disabled adolescents. The question remained whether such adolescents would exhibit significant deficits in the imrnediare recall of semantically and syntactically varied "sentences" when compared with academically achieving peers. Sentence-repetition tasks have been widely used to assess the linguistic competence of children and adults (Brown & Bellugi, 1964; Fraser, et al., 1963; Luterman & Bar, 1971; Menyuk, 1964, 1969; Menyuk & Looney, 1972a, 1972b; Newcombe & Marshall, 1967). Experimental evidence consistently indicates interactions between grammatical structure and semantic aspects and memory for sentences. These are interpreted by Slobin ( 1971) ro indicate chat meaning and form can be stored independently and that meaning (deep structure) is more persistent in memory than form (surface structure). Miller and Chomsky (1963) suggest that the surface structure of a sentence is computed within short-term memory and that the limitations of this memory relate directly to the
120
E. H. WlIG
6:
XI A. ROACH
transformational structure of language. Deep structure and semantic interpretation are considered to be derived within a larger and longer-term memory score. Sentence-repetition studies of language-disordered children have indicated that sentence length and structure constitute significant variables in recall (Menyuk, 1964, 1969; Menyuk & Looney, 1972a, 1972b). Newcombe and Marshall (1967) extended the analytic possibilities of a sentence-repetition task by varying semantic and syntactic constraints. Their research indicated that deficits in recall of "sentences" distinguished left-hemisphere-damaged from right-hemisphere-damaged and non-brain-damaged adults. The present study employed Newcombe and Marshall's "sentences" to assess and compare the effects of differences in semantic and syntactic constraints on immediate recall by learningdisabled and academically achieving adolescents.
Ss were 30 adolescents, 7 females and 23 males, with specific learning disabilities diagnosed by a psycho-educational team. They ranged in age from 12 yr., 5 mo. to 16 yr., 4 mo. ( M = 14 yr., 2 mo.) and attended middle-class suburban public schools in Grades 7, 8, and 9. WISC Full-Scale IQs ranged from 87 to 129 ( M = 9 6 ) . All scored within k 3 mo. of CA on the Peabody Picture Vocabulary Test (Dunn, 1959). Achievement scores in grade equivalents on the Metropolitan Achievement Tests (Durost, 1961 ) ranged from 2.7 to 6.7 (M = 4.5) for Reading, from 3.1 co 4.7 ( M = 4.1) for Language, and from 2.1 to 5.3 ( M = 4.2) for Mathematics. All exhibited academic retardation of at least 1.5 grade equivalent scores in two or more academic areas, normal auditory acuity and articulatory ability, and none had received speech or language therapy. The controls were 30 academically achieving adolescents from a middleclass suburban public school. They were matched for sex, grade level, age ( M = 14 yr., 1 mo.), and IQ (Range 88 to 130; M = 103). All scored within -+3 mo. of CA on the Peabody test and grade equivalents on the Metropolitan ranged from 7.5 to 9.5 ( M = 8.3). None received speech or language therapy and all had normal auditory acuity. Materialr and Procedrcre Intelligence and achievement were evaluated during the testing. The experimental "sentences" (Newcombe & Marshall, 1967) were read to all Ss individually, in random sequence to assess immediate recall. Ss were first given four practice sentences which included a random word-string and sentences with varied semantic and syntactic constraints. They were requested to repeat each sentence immediately after presentation and the instructions stressed that verbatim recall was required. Individual responses were recorded to permit analysis of
RECALL OF
SENTENCES BY
LEARNING DISABLED
12 1
errors. The number of correct, verbatim repetitions provided the raw data for statistical tests of the significance of mean differences and of differences in the responses to the various sentences.
RESULTS The 30 learning-disabled adolescents repeated 410 sentences verbatim (M = 13.67; SD = 0.38; range 8 to 20) while the academically achieving controls repeated 514 sentences verbatim (M = 17.12; SD = 1.94; range 1 3 to 20). The mean difference proved significant ( t = 2.12, p < .05), indicating that the learning-disabled adolescents made significantly more errors in repeating the sentences than their controls. The number of verbatim repetitions of each of the sentences by the learningdisabled and achieving adolescents were subsequently compared (Table 1). The former group made significantly more errors on ( a ) Sentences 2, 16, 19, and 20 which were syntactically well formed, but violated semantic (selectional) rules, ( b ) Sentences 11 and 13 which contained modifier-strings in correct and incorrect order, ( c ) Sentence 15 which contained a random word-string, and ( d ) Sentence 16 which was structurally complex with 'embedding.' TABLE 1
COMPARISON OF VERBATIM REPETITIONS BY 30 LEARNING-DISABLED AND 30 ACADEMICALLY ACHIEVING ADOLESCENTS "Sentence"
1. 2. 3. 4. 5. 6. 7. 8.
9. 10. 11. 12. 13. 14. 15. 16. 17.
The team of workers built the bridge. Colorless green ideas sleep furiously. The man posted the letter. The boy hit the girl. The politician nearly lost the election. The boy easily passed the examination. Didn'r the mechanic repair the van? Didn't the lion chase rhe tiger? Wasn't the stone wall built by the kind husband? Wasn't the rich uncle advised by the nice manager? She has bought five large brown leather cases. H e has sold the long heavy grey shiny car. She has washed plastic red small eight cups. Not in a tree to the lake ran with. Walk some by hard of clearly table very. The sky that the dream thought jumped cheaply. The burglar that the police found escaped easily. The chair roughly painted the fire. Wasn't the fat ceiling robbed by the tired pen? The man that the book read was interesting. .O5; t p .01; Sp .001.
18. 19. 20. *p
