313
Paradigmatic Language Training for Learning Disabled Children
Lora M. Cartelli, PhD
In last month's Journal, Suiter and Potter demonstrated that like normals, LD children use categories to organize material for recall. Similarly, this article shows that LD children can benefit from preorganized material and that the positive effects are not lost soon after teaching, as is often reported in clinical observations of LD children. These studies may suggest avenues of strength —avenues differing in their conceptual level as opposed to their modality — which can be used in the teaching process. — G.M.S. The effect that training paradigmatic language structures has upon the reading process is investigated with LD children. The effects of intelligence, special training, and training over time were analyzed in a pretest-posttest control group design. Findings reveal that paradigmatic thought products are both testable and teachable structures and are inherent in both oral and written communication. Children who received training made rapid and positive gains on all criterion measures, and gains were residual over time. It appears that the ability to manipulate and apply paradigmatic operations has a favorable
effect upon the understanding of written language and helps LD children to process and organize information.
R
esearch investigating the language factors in learning problems in academic areas generally conclude that academic deficits appear to be a function of attention and memory disorders which apparently interfere with the manipulation and organization of information (Freston & Drew 1974, Senf & Freundl 1971, Underwood 1964). Consequently when children with learning difficulties are required to Journal of Learning Disabilities
54 Downloaded from ldx.sagepub.com at OAKLAND UNIV on June 11, 2015
314 communicate in oral or written language, they fall short of normal expectations. It is impossible to consider learning independent of language and its meaning. Dinnan (1971) has postulated that paradigmatic responding is the key to learning and communication in any symbolic code. Paradigmatic responses are those of the same form class as the stimulus word on free wordassociation tests, and are the "expected" mode of operation in schools and standardized tests. Thus, the ability to organize information paradigmatically allows the child to communicate and think in the same mode as the instructor — ensuring an educational interchange. Syntagmatic responses, on the other hand, are "unexpected," or outside a paradigmatic framework. They are based on an internalized set of experiences (e.g., answering "twirling" to the stimulus word "top") or are phonetically close to the stimulus (answering "pop" to "top"). A student who uses such a personalized referent base may respond in ways quite meaningful to himself but meaningless to others in the instructional setting. Several researchers have investigated the language difficulties manifested in freeassociation tasks. Bartel, Grill, and Bartel (1973) investigated whether learning disabled (LD) children experience specific problems in shifting from syntagmatic to paradigmatic language responding (P/S shift) and whether these problems were related to the variety of language deficits often associated with LD children. Results of their study and of one conducted by Freston and Drew (1974) reveal that there are no significant differences between normal and LD children in making the P/S shift. The children made most paradigmatic responses as age increased, and neither age nor IQ alone was adequate to explain the transition. Their findings support previous research (Deese 1962, Entwisle, Forsyth, & Muuss 1964); however, the fact that paradigmatic responding occurs as expected by age norms does not ensure academic success for the LD child — the children
in the studies still performed at a lower level on achievement tests when compared to their normal counterparts. Apparently, the ability to make paradigmatic responses in oral language tasks does not ensure that a child will automatically transfer and apply this ability to written language tasks such as reading. It maybe precisely in this transition that LD children differ most significantly from normals and fall short of normal academic expectations. Research has shown that the base referents of the paradigmatic language code comprise the key components of the language used in schools and in most standardized achievement tests (Bickley, Dinnan, & Jones 1971, Dinnan 1973, Dinnan & Lodge 1976, Lammarre 1976, McNinch, Hornsby, & Richmond 1970). These base referents appear in 50$ of oral and written communications and are the common denominator used by schools in teaching. It could be that LD children use different base referents to organize percepts, a deviation from what is normally expected in most academic situations, or else they just fail to perceive these base referents in written language and do not make the transfer. Thus, this study tries to ascertain whether the subjects are predominantly paradigmatic or syntagmatic in thought structure and whether they are actually making the transition to written language tasks (in this study, reading). Secondly, this study seeks to implement a training program specifically designed to teach the meaning and use of paradigmatic thought structures to assist the child with reading tasks.
SUBJECTS The sample consisted of 46 children identified as learning disabled according to the following criteria set forth by the Florida state regulations (Bureau of Education 1975, pp. 25-26): (1) a score of not less than -2 SD on the Stanford-Binet, Wechsler Intelligence Scales for Children, or the Wechsler Preschool or Primary Scale of Intelligence;
Volume 11, Number 5, May 1978 Downloaded from ldx.sagepub.com at OAKLAND UNIV on June 11, 2015
55
315 (2) normal visual and auditory acuity confirmed by examination, and no evidence of a primary physical handicap; (3) based on expectancy age, a score on a standardized test in the lower tenth percentile in one or more of the following academic areas: reading, writing, arithmetic, spelling, and prerequisite skills; (4) no evidence of severe emotional disturbance as indicated by test data or the clinical judgment of a qualified psychologist; (5) based on expectancy age, a score of less than -2 SD on standardized tests which measure the following basic psychological processes: visual processing, auditory processing, haptic processing, language (any meaningful communication) including receptive and expressive, and sensory integration including visual-motor, visual-auditory (vocal), and auditory-visual. A deficit in one area is sufficient for eligibility; however, where a score less than -2 to -2.5 SD occurs, another test must confirm the deficit. Of the seven elementary schools in Collier County, Florida, four were randomly chosen for this study. Two schools served as the experimental group and two as the control group. The mean chronological age for the experimental group was 8.64 (SD = 1.16) and for the control group 8.81 (SD = 1.03). The mean IQ for the experimental group was 106.00 (SD = 7.72) and for the control group 104.30 (SD = 8.91). Analysis of this dimension showed no main effect or interactions with any of the other variables.
PROCEDURE AND METHOD All children in the study were given pretests on the Metropolitan Reading Tests, Primary II Battery, Form F, and the P/S Oral Language Inventory, Form A (Dinnan 1971b). The P/S Inventory was selected as the verbal language test to measure the thought process. It measures the degree of paradigmatic responses to specific stimuli in a free-association task. Approximately 80 words were identified from Coleman's (1968) rank order of the Thorndike and Lorge
Frequency List that had contrast associations, i.e., up-down, front-back. This listing was limited to the first 700 words because of Coleman's conclusion that these words constitute 67% of adult communication. Thirty words were randomly selected from the 80 to comprise the initial P/S Inventory. Stanford-Binet scores were used to determine if initial differences in intellectual ability existed. No statistically significant differences (F < 1 on all criterion measures) were found between the groups. Training in paradigmatic language was begun immediately following the collection of pretest data. Each experimental school received paradigmatic language training for 19 consecutive school days, approximately 2-and-a-half hours a day, for a total of about 45 to 50 hours per school. The control group received no special training other than that normally conducted in their LD programs. The learning disabilities teachers in each experimental school were present throughout the entire experimental program; however, they were not required to implement any training procedures. The Key to Learning Series, Kits I and II (Dinnan & Co wart 1976), was used to train students in the acquisition and application of paradigmatic language structures. The students in each experimental school were divided into two groups for training purposes; however, they were treated as one experimental group in the analysis of the data. Group I consisted of students who had not yet acquired the base referents of the paradigmatic language code as shown by their performance on the P/S Inventory (mean paradigmatic response was less than 15). Training for this group began with Kit I, which is designed to develop a cognitive awareness of the paradigmatic language framework by teaching the base referents of this code through contrast relations. Since the meaning and understanding of these base referents are on a cognitive and not a linguistic level and are not restricted to semantic and syntactic constraints, students are taught to manipulate and organize mentally these referents into meaningful relations within the Journal of Learning Disabilities
56 Downloaded from ldx.sagepub.com at OAKLAND UNIV on June 11, 2015
316 oral language code. Oral language readiness, then, involved the mental manipulation of these key verbal elements to progressively higher levels of organization (classes and systems) and progressed toward the application of paradigmatic language structures to the reading processes. Kit II was used when the students were ready for the transition to written language. Group 2 consisted of students who had acquired the base referents as shown by a score of 15 or more on the pretest of the P/S Inventory. Training with these students began with a review of the nucleus of the English language, followed by the sequential lessons outlined in Kit II, which is designed to teach reading readiness for the written symbolic code. When these students could classify various objects, pictures, events, etc. according to the relational referents, the experimenter felt they were ready for written language or reading. Then they were taught sight word recognition of the prime frequency words in contrast sets. The first posttest was conducted immediately after the completion of the experimental training program. The P/S Inventory, Form B and the Metropolitan Reading Tests, Primary II Battery, Form G were administered to all children at this time. Follow-up testing was conducted about 10 weeks after the collection of the initial posttest data, no special training being conducted during this 10-week interval. P/S Inventory, Form C and the Metropolitan Reading Tests, Form H were then administered.
RESULTS AND FINDINGS Tables I to IV report the means and standard deviations for all groups on the four criterion measures. From the results of the analysis of the data the following findings are presented: (1) LD children who received paradigmatic language training performed significantly better on all criterion measures than LD children who did not. (2) LD children, regardless of group membership, performed significantly better on
the criterion follow-up tests than the criterion posttests (see Tables I through IV). (3) The interaction effects of time of testing and special training were also statistically significant across all criterion measures. The interaction effects were found to be of an ordinal nature; that is, LD children who received special training performed better on all criterion measures than LD children who did not receive special training, regardless of time of testing. (4) A comparison of means from one testing time to another showed that children receiving special training made significantly greater gains from criterion posttests to the criterion follow-up tests than children not receiving special training; that is, the effects of paradigmatic language training were residual over time.
CONCLUSIONS AND DISCUSSION The findings of this study reveal that paradigmatic language structures are teachable and testable thought products. Inspection of the means and standard deviations for the groups reveals that the gains for children receiving special training were rapid and positive. Of a possible 30 points on the P/S Inventory, the experimental groups had already reached a mean score near the ceiling on the first posttest. Once the acquisition of the base referents of the paradigmatic language code had been achieved, they became a part of the cognitive process and did not dissipate over time. Beauvois (1973), Dinnan (1971a), and Francis (1972) similarly find the P/S classification system to be a valid criterion for the evaluation of high-level mental processing. The acquisition and application of paradigmatic language structures was not restricted to semantic and syntactic constraints, but involved the ability to perform the mental manipulations of these referents, first in the oral language code and then in the written symbolic code. Olson (1970) suggested that a person's knowledge of referents and the ability to organize meaningful relationships among them
Volume Uy Number 5, May 1978 Downloaded from ldx.sagepub.com at OAKLAND UNIV on June 11, 2015
57
317 TABLE I. Group means and standard deviations on P/S language tests.
TABLE II. Group means and standard deviations on word knowledge tests. Experimental
Experimental
Pretest Posttest Follow-up Total*
7.70 25.17 27.65 26.41
6.23 2.04 1.40 2.04
Control
7.39 8.05 9.09 8.57
5.14 5.94 7.28 6.38
Pretest* Posttest Follow-up Totalf
M 2.07 2.60 3.17 2.89
SD .70 .84 .60 .75
Control M 2.04 2.00 2.10 2.05
SD .61 .53 .53 .51
*Time of testing; F = 17.7, df = 1, p < .01 tExperimental vs. control; F = 26.9; df = 1; p
Paradigmatic Language Training for Learning Disabled Children
Lora M. Cartelli, PhD
In last month's Journal, Suiter and Potter demonstrated that like normals, LD children use categories to organize material for recall. Similarly, this article shows that LD children can benefit from preorganized material and that the positive effects are not lost soon after teaching, as is often reported in clinical observations of LD children. These studies may suggest avenues of strength —avenues differing in their conceptual level as opposed to their modality — which can be used in the teaching process. — G.M.S. The effect that training paradigmatic language structures has upon the reading process is investigated with LD children. The effects of intelligence, special training, and training over time were analyzed in a pretest-posttest control group design. Findings reveal that paradigmatic thought products are both testable and teachable structures and are inherent in both oral and written communication. Children who received training made rapid and positive gains on all criterion measures, and gains were residual over time. It appears that the ability to manipulate and apply paradigmatic operations has a favorable
effect upon the understanding of written language and helps LD children to process and organize information.
R
esearch investigating the language factors in learning problems in academic areas generally conclude that academic deficits appear to be a function of attention and memory disorders which apparently interfere with the manipulation and organization of information (Freston & Drew 1974, Senf & Freundl 1971, Underwood 1964). Consequently when children with learning difficulties are required to Journal of Learning Disabilities
54 Downloaded from ldx.sagepub.com at OAKLAND UNIV on June 11, 2015
314 communicate in oral or written language, they fall short of normal expectations. It is impossible to consider learning independent of language and its meaning. Dinnan (1971) has postulated that paradigmatic responding is the key to learning and communication in any symbolic code. Paradigmatic responses are those of the same form class as the stimulus word on free wordassociation tests, and are the "expected" mode of operation in schools and standardized tests. Thus, the ability to organize information paradigmatically allows the child to communicate and think in the same mode as the instructor — ensuring an educational interchange. Syntagmatic responses, on the other hand, are "unexpected," or outside a paradigmatic framework. They are based on an internalized set of experiences (e.g., answering "twirling" to the stimulus word "top") or are phonetically close to the stimulus (answering "pop" to "top"). A student who uses such a personalized referent base may respond in ways quite meaningful to himself but meaningless to others in the instructional setting. Several researchers have investigated the language difficulties manifested in freeassociation tasks. Bartel, Grill, and Bartel (1973) investigated whether learning disabled (LD) children experience specific problems in shifting from syntagmatic to paradigmatic language responding (P/S shift) and whether these problems were related to the variety of language deficits often associated with LD children. Results of their study and of one conducted by Freston and Drew (1974) reveal that there are no significant differences between normal and LD children in making the P/S shift. The children made most paradigmatic responses as age increased, and neither age nor IQ alone was adequate to explain the transition. Their findings support previous research (Deese 1962, Entwisle, Forsyth, & Muuss 1964); however, the fact that paradigmatic responding occurs as expected by age norms does not ensure academic success for the LD child — the children
in the studies still performed at a lower level on achievement tests when compared to their normal counterparts. Apparently, the ability to make paradigmatic responses in oral language tasks does not ensure that a child will automatically transfer and apply this ability to written language tasks such as reading. It maybe precisely in this transition that LD children differ most significantly from normals and fall short of normal academic expectations. Research has shown that the base referents of the paradigmatic language code comprise the key components of the language used in schools and in most standardized achievement tests (Bickley, Dinnan, & Jones 1971, Dinnan 1973, Dinnan & Lodge 1976, Lammarre 1976, McNinch, Hornsby, & Richmond 1970). These base referents appear in 50$ of oral and written communications and are the common denominator used by schools in teaching. It could be that LD children use different base referents to organize percepts, a deviation from what is normally expected in most academic situations, or else they just fail to perceive these base referents in written language and do not make the transfer. Thus, this study tries to ascertain whether the subjects are predominantly paradigmatic or syntagmatic in thought structure and whether they are actually making the transition to written language tasks (in this study, reading). Secondly, this study seeks to implement a training program specifically designed to teach the meaning and use of paradigmatic thought structures to assist the child with reading tasks.
SUBJECTS The sample consisted of 46 children identified as learning disabled according to the following criteria set forth by the Florida state regulations (Bureau of Education 1975, pp. 25-26): (1) a score of not less than -2 SD on the Stanford-Binet, Wechsler Intelligence Scales for Children, or the Wechsler Preschool or Primary Scale of Intelligence;
Volume 11, Number 5, May 1978 Downloaded from ldx.sagepub.com at OAKLAND UNIV on June 11, 2015
55
315 (2) normal visual and auditory acuity confirmed by examination, and no evidence of a primary physical handicap; (3) based on expectancy age, a score on a standardized test in the lower tenth percentile in one or more of the following academic areas: reading, writing, arithmetic, spelling, and prerequisite skills; (4) no evidence of severe emotional disturbance as indicated by test data or the clinical judgment of a qualified psychologist; (5) based on expectancy age, a score of less than -2 SD on standardized tests which measure the following basic psychological processes: visual processing, auditory processing, haptic processing, language (any meaningful communication) including receptive and expressive, and sensory integration including visual-motor, visual-auditory (vocal), and auditory-visual. A deficit in one area is sufficient for eligibility; however, where a score less than -2 to -2.5 SD occurs, another test must confirm the deficit. Of the seven elementary schools in Collier County, Florida, four were randomly chosen for this study. Two schools served as the experimental group and two as the control group. The mean chronological age for the experimental group was 8.64 (SD = 1.16) and for the control group 8.81 (SD = 1.03). The mean IQ for the experimental group was 106.00 (SD = 7.72) and for the control group 104.30 (SD = 8.91). Analysis of this dimension showed no main effect or interactions with any of the other variables.
PROCEDURE AND METHOD All children in the study were given pretests on the Metropolitan Reading Tests, Primary II Battery, Form F, and the P/S Oral Language Inventory, Form A (Dinnan 1971b). The P/S Inventory was selected as the verbal language test to measure the thought process. It measures the degree of paradigmatic responses to specific stimuli in a free-association task. Approximately 80 words were identified from Coleman's (1968) rank order of the Thorndike and Lorge
Frequency List that had contrast associations, i.e., up-down, front-back. This listing was limited to the first 700 words because of Coleman's conclusion that these words constitute 67% of adult communication. Thirty words were randomly selected from the 80 to comprise the initial P/S Inventory. Stanford-Binet scores were used to determine if initial differences in intellectual ability existed. No statistically significant differences (F < 1 on all criterion measures) were found between the groups. Training in paradigmatic language was begun immediately following the collection of pretest data. Each experimental school received paradigmatic language training for 19 consecutive school days, approximately 2-and-a-half hours a day, for a total of about 45 to 50 hours per school. The control group received no special training other than that normally conducted in their LD programs. The learning disabilities teachers in each experimental school were present throughout the entire experimental program; however, they were not required to implement any training procedures. The Key to Learning Series, Kits I and II (Dinnan & Co wart 1976), was used to train students in the acquisition and application of paradigmatic language structures. The students in each experimental school were divided into two groups for training purposes; however, they were treated as one experimental group in the analysis of the data. Group I consisted of students who had not yet acquired the base referents of the paradigmatic language code as shown by their performance on the P/S Inventory (mean paradigmatic response was less than 15). Training for this group began with Kit I, which is designed to develop a cognitive awareness of the paradigmatic language framework by teaching the base referents of this code through contrast relations. Since the meaning and understanding of these base referents are on a cognitive and not a linguistic level and are not restricted to semantic and syntactic constraints, students are taught to manipulate and organize mentally these referents into meaningful relations within the Journal of Learning Disabilities
56 Downloaded from ldx.sagepub.com at OAKLAND UNIV on June 11, 2015
316 oral language code. Oral language readiness, then, involved the mental manipulation of these key verbal elements to progressively higher levels of organization (classes and systems) and progressed toward the application of paradigmatic language structures to the reading processes. Kit II was used when the students were ready for the transition to written language. Group 2 consisted of students who had acquired the base referents as shown by a score of 15 or more on the pretest of the P/S Inventory. Training with these students began with a review of the nucleus of the English language, followed by the sequential lessons outlined in Kit II, which is designed to teach reading readiness for the written symbolic code. When these students could classify various objects, pictures, events, etc. according to the relational referents, the experimenter felt they were ready for written language or reading. Then they were taught sight word recognition of the prime frequency words in contrast sets. The first posttest was conducted immediately after the completion of the experimental training program. The P/S Inventory, Form B and the Metropolitan Reading Tests, Primary II Battery, Form G were administered to all children at this time. Follow-up testing was conducted about 10 weeks after the collection of the initial posttest data, no special training being conducted during this 10-week interval. P/S Inventory, Form C and the Metropolitan Reading Tests, Form H were then administered.
RESULTS AND FINDINGS Tables I to IV report the means and standard deviations for all groups on the four criterion measures. From the results of the analysis of the data the following findings are presented: (1) LD children who received paradigmatic language training performed significantly better on all criterion measures than LD children who did not. (2) LD children, regardless of group membership, performed significantly better on
the criterion follow-up tests than the criterion posttests (see Tables I through IV). (3) The interaction effects of time of testing and special training were also statistically significant across all criterion measures. The interaction effects were found to be of an ordinal nature; that is, LD children who received special training performed better on all criterion measures than LD children who did not receive special training, regardless of time of testing. (4) A comparison of means from one testing time to another showed that children receiving special training made significantly greater gains from criterion posttests to the criterion follow-up tests than children not receiving special training; that is, the effects of paradigmatic language training were residual over time.
CONCLUSIONS AND DISCUSSION The findings of this study reveal that paradigmatic language structures are teachable and testable thought products. Inspection of the means and standard deviations for the groups reveals that the gains for children receiving special training were rapid and positive. Of a possible 30 points on the P/S Inventory, the experimental groups had already reached a mean score near the ceiling on the first posttest. Once the acquisition of the base referents of the paradigmatic language code had been achieved, they became a part of the cognitive process and did not dissipate over time. Beauvois (1973), Dinnan (1971a), and Francis (1972) similarly find the P/S classification system to be a valid criterion for the evaluation of high-level mental processing. The acquisition and application of paradigmatic language structures was not restricted to semantic and syntactic constraints, but involved the ability to perform the mental manipulations of these referents, first in the oral language code and then in the written symbolic code. Olson (1970) suggested that a person's knowledge of referents and the ability to organize meaningful relationships among them
Volume Uy Number 5, May 1978 Downloaded from ldx.sagepub.com at OAKLAND UNIV on June 11, 2015
57
317 TABLE I. Group means and standard deviations on P/S language tests.
TABLE II. Group means and standard deviations on word knowledge tests. Experimental
Experimental
Pretest Posttest Follow-up Total*
7.70 25.17 27.65 26.41
6.23 2.04 1.40 2.04
Control
7.39 8.05 9.09 8.57
5.14 5.94 7.28 6.38
Pretest* Posttest Follow-up Totalf
M 2.07 2.60 3.17 2.89
SD .70 .84 .60 .75
Control M 2.04 2.00 2.10 2.05
SD .61 .53 .53 .51
*Time of testing; F = 17.7, df = 1, p < .01 tExperimental vs. control; F = 26.9; df = 1; p
