Grammatical comprehension in developmental semantic-pragmatic disorder

This is a clinical case study of 105 children; 35 children demonstrated ′semantic-pragmatic′ type of specific language impairment, 35 children presented with delayed language development due to mental retardation, and 35 children with normal language development. In contrast to the disturbances of linguistic form that characterize the most frequently reported type of specific language impairment, the semantic-pragmatic disorder is characterized mostly by ′inappropriate′ language use. A cognitive explanation has been proposed, which accounts for all instances of inappropriate language thus far studied in this population. An expressive linguistic explanation has also been proposed, which accounts for some instances of inappropriate language. This study investigates the possibility that a receptive linguistic explanation can account for inappropriate responses to questions. A linguistic explanation with receptive and expressive components could account for some of the semantic-pragmatic behaviors previously accounted for by the cognitive explanation. To test this proposal, the linguistic profiles of all children were examined using the Arabic language test. The results indicate that grammatical comprehension was impaired in about 45.7% of the children examined. Therefore, instances of inappropriate language use that appear to express disordered concepts may reflect receptive linguistic deficits instead.


Introduction
Specifi c language impairment (SLI) presents when children present language maturation at least 12 months behind their chronological age in the absence of sensory or intellectual defi cits, pervasive developmental disorders, evident cerebral damage, and adequate social and emotional conditions [1]. SLI cases can present great variability in clinical manifestations concerning language. One approach to dealing with this heterogeneity involves the identifi cation of clinical subtypes of SLI [2].
For example, Rapin and Allen [3] described three subtypes of SLI and six profi les of language problems based on the evaluation of spontaneous and directed language, taking into account the level of linguistic analysis, in terms of phonological, morphosyntactic, semantic-lexical, and pragmatic analysis. Th e three subtypes of SLI were as follows: (1) Mixed receptive expressive disorders: Verbal auditory agnosia and phonologic-syntactic defi cit disorder. (2) Expressive disorders: Verbal dyspraxia and phonological programming defi cit disorder. (3) Higher-order processing disorders: Lexical-syntactic defi cit disorder and semantic-pragmatic disorder.
Five of these six profi les have found empirical confi rmation in a study by Conti-Ramsden et al. [4] that combined clinical and test information. A cluster analysis was performed on the children's performance, and the children were divided into six groups. One of the groups had no match with Rapin and Allen's categories because it comprised children who appeared to be performing within the normal range. In contrast, the 'verbal auditory agnosia' group described by Rapin and Allen was not found. Van der Lely [5] proposed SLI classifi cation of fi ve subtypes, which includes the following: semantic-pragmatic SLI, primarily phonological SLI, primarily syntactic SLI, mixed primarily phonological SLI who show impairment in both phonology and syntax with more phonological than syntactic impairment, and mixed primarily syntactic SLI whose syntactic impairment is more apparent (or more persistent) compared with phonological defi cits.
Children with pragmatic impairment have problems using language appropriately in a given context. Although they can present in the early years with language delay and poor social development, often these diffi culties seem at least superfi cially to be overcome by the school years and children may be fl uent with seemingly normal syntax and adequate articulation. However, problems such as understanding discourse, literal use of language, impaired understanding of social inference, and the social use of language remain substantial [6]. Th ey may have diffi culty with speaker

Grammatical comprehension in developmental semantic-pragmatic disorder
Sally T. Kheir El-Din a , Yossra Sallam b This is a clinical case study of 105 children; 35 children demonstrated 'semantic-pragmatic' type of speci c language impairment, 35 children presented with delayed language development due to mental retardation, and 35 children with normal language development. In contrast to the disturbances of linguistic form that characterize the most frequently reported type of speci c language impairment, the semantic-pragmatic disorder is characterized mostly by 'inappropriate' language use. A cognitive explanation has been proposed, which accounts for all instances of inappropriate language thus far studied in this population. An expressive linguistic explanation has also been proposed, which accounts for some instances of inappropriate language. This study investigates the possibility that a receptive linguistic explanation can account for inappropriate responses to questions. A linguistic explanation with receptive and expressive components could account for some of the semantic-pragmatic behaviors previously accounted for by the cognitive explanation. To test this proposal, the linguistic pro les of all children were examined using the Arabic language test. The results indicate that grammatical comprehension was impaired in about 45.7% of the children examined. Therefore, instances of inappropriate language use that appear to express disordered concepts may re ect receptive linguistic de cits instead.
listener roles, turn taking, and the relevance of language in conversation. Th ese children do not compensate for their language diffi culties by developing a rich repertoire of nonverbal communication [7]. In addition, they also share a number of conversational problems with high-functioning autistic children (e.g. they have a tendency to talk about personal preoccupations, inappropriate questioning style, with repetitive speech and some stereotyped speech), although they do not appear to have the same degree of cognitive rigidity or obsession tendency as seen in those with autism spectrum disorders [8]. As a result, their language use leads to listener's impression of inappropriateness.
Rapin and Allen [9] pointed out that it is not known to what extent inappropriate language refl ects a 'thought disorder', or a cognitive defi cit, in addition to a 'communication disorder, or a linguistic defi cit. Cognitive defi cits refer to problems with concepts, whereas linguistic defi cits refer to problems with linguistic form -that is, the pairing of form and concepts. Bishop and Adams [10] suggested that cognitive defi cits could account for this inappropriateness and did not propose an alternative linguistic explanation.
In contrast, Snow [11] supported the linguistic explanation, which implies that the child's concepts are appropriate, but he or she has diffi culty selecting the correct linguistic form to convey the intended meaning. He proposed that the child's inappropriate responses refl ected poor comprehension of some of the questions addressed to him, which extended the scope of the linguistic explanation to include the receptive defi cit.
Th e aim of this study was to determine whether the linguistic hypothesis especially in its receptive part warrants further investigation in study of this population to prove that the breakdown in receptive syntax is linguistic in origin.

Participants and methods
This case-control study included 105 Arabicspeaking children (60 and 55 ) in the age range of 3-6 years (mean age 4.27 years) who were selected from the Phoniatric clinics of Al Demerdash and Al Zahra University Hospitals. Of the 105 children, 35 children presented with semantic-pragmatic SLI (the semantic-pragmatic group), 35 with delayed language development due to mental retardation (the delayed language group), and 35 children with normal language development (the control group). The study was approved by ethical committee.

Methods
Information about each child included full name, date of birth, age, address, and order of birth. Language assessment was based on the Arabic Language Test [12]. Psychometric evaluation was carried out using the Stanford Binet Intelligence Scale [13].

Statistical analysis
Data were analyzed using SPSS, version 15 (SPSS Inc., Chicago, Illinois, USA). Quantitative data were presented using mean and SD. Qualitative data were presented using frequency and related percentage. Comparison between groups was made using the 2 -test. A P value of 0.05 was chosen as level of signifi cance. Post-hoc test was used to compare the linguistic quotients (language age/chronological age) across the diff erent groups, including the control group.

Results
Th e dataset consisted of 105 children (60 and 55 ), with a mean age of 4.3 years (Table 1).
Th e age of various linguistic parameters varied from 2 to 8 years, with the lowest mean age for receptive syntax (3.97 years) and the highest mean age for semantics (4.07 years) ( Table 2).
Tables 3-5 provide estimates of the quotients of various linguistic parameters in the three groups. Because receptive syntax was impaired in 16 (45.7%) children of the 35 in the semantic-pragmatic group, receptive syntax quotient did not reach the normal level as expressive syntax and phonological quotients.
Th ere was a signifi cant diff erence between the semantic-pragmatic group and the control group in quotients of various linguistic parameters: semantic, pragmatic, and receptive syntax quotients, whereas there was no signifi cant diff erence between the two groups in quotients of expressive syntax and phonology (P = 0.000; P < 0.005).
Th ere was a signifi cant diff erence between the delayed language group and the control group in quotients of all linguistic parameters: semantic, pragmatic, receptive syntax, expressive syntax, and phonology (P = 0.000; P < 0.005).
Th ere was a signifi cant diff erence between the semantic-pragmatic group and the delayed language group in quotients of all linguistic parameters: semantic, pragmatic, receptive syntax, expressive syntax, and phonology (P = 0.000; P < 0.005).

Discussion
Although SLI has been viewed as a heterogeneous condition in which children may present with a wide range of different disturbances in language processing [14], sufficient information on grammatical comprehension in patients with semantic-pragmatic SLI was absent. In the present study, comparison between the control group and the semantic-pragmatic group showed a significant difference between the two groups in certain parameters only (semantic, pragmatic, and receptive syntactic), whereas the comparison between the control group and the delayed language group showed a significant difference in all parameters. This proves that the delayed language group due to mental retardation has a linguistic affection in all parameters, and this is due to delayed development and maturation in all aptitudes; even motoric development is delayed.
Th e receptive syntax may be aff ected in a large number of patients with semantic-pragmatic SLI, particularly with complex sentence, despite the fact that they can express the same sentence in a better way. Th is supports that the child's inappropriate responses refl ects poor comprehension of some of the questions addressed to him, which extends the scope of the linguistic explanation to include the receptive defi cit.
Rinaldi [15] indicated that impairment of receptive syntax in these children may be explained by their impaired semantic abilities. Semantic abilities concern the understanding of meaning of words or phrases and their expression [8]. Th is fi nding is in agreement with the criteria we fi nd in the language of children with semantic-pragmatic disorder and it is also proved by statistical analysis.
Th is result is in agreement with that reported by Leonard [16] as well. He suggested that children with SLI may be 1 year below age level in one set of features, 1.5 year below in another, 6 months below in a third, and so on.
In light of growing evidence that developmental disorders may also have neurogenic explanation [17], syndromes diff erentially aff ecting morphosyntactic and lexical form classes may also implicate diff erent sites of involvement [18].