Saturday, October 12, 2013

The Underlying Nature of Specific Language Impairment


Bishop, D. V. (1992). The underlying nature of specific language impairment. The Journal of Child Psychology and Psychiatry and Allied Disciplines, 33, 3–66.

In this classic paper entitled "The underlying nature of specific language impairment", Bishop reviews and evaluates six theories that describe SLI in terms of a core deficit in cognitive processing. The aim of this posting is to provide a brief summary of each hypothesis discussed in in this 1992 article.

Hypothesis 1 – Language Impairment as an Output Disorder
According to a view of SLI as an output disorder, the deficit lies in the ability to transform a concept in the mind to speech for output. The idea is that articulatory errors occur with such frequency in these children that this output eventually impairs the lexical route on which the concept based. However, many children with SLI do not have a significant articulatory disorder. As a result, this hypothesis cannot account for SLI.

Hypothesis 2 – Language Impairment as an Auditory Disorder
In a series of studies, Tallal and her colleagues provided evidence for auditory deficits in children with SLI. These children tended to have problems in the perception and production of rapid sequences of stimuli (Tallal & Piercy, 1973a,b).  Several examples of how the auditory perceptual impairment theory can account for several deficits seen in children with SLI are provided including phonological problems (e,g., Tallal, & Piercy, 1974), comprehension difficulties (e.g., Frumkin & Rapin, 1980), and expressive grammatical problems (Leonard et al., 1987). According to Bishop (1992), further work is necessary to fully understanding how the auditory limitations theory can account for different linguistic and non-verbal deficits seen in children with SLI.

Hypothesis 3 – Linguistic Interpretations of SLI
The third hypothesis reviewed by Bishop (1992), addresses the possibility that children with SLI are born with a deficit in an “innate language acquisition device” that has been proposed to be responsible for mastering the grammatical relations of a language. There are many ways in which this language acquisition device may cause the characteristic impairments that are observed in SLI.  One possibility is that children with SLI are born with the inability to extract the hierarchical structure of their nature language (Cromer, 1978). Although hierarchical processing deficits have been reported in children with SLI, the same pattern of performance has also been observed in other populations such as children who are deaf (Bishop, 1982), providing contradictory evidence against this theory. An alternative theory is that children with SLI are impaired in their ability to link grammatical structure to meaning (Pinker, 1989). This theory has been applied to children with SLI and their language matched peers, with the finding that children with SLI have a deficit in what appears to be an innate ability to link grammar to meaning, while typically developing children do not show the same pattern of results (Van der Lely, 1990). Although plausible, this finding does not exclude the possibility that auditory perception issues could be the core deficit in SLI.  Bishops concludes his review by stating that the current findings are not well understood and that more sophisticated analyses will need to be conducted before fully rejecting or accepting this theory.

Hypothesis 4 - Explanation of SLI in Terms of Piagetian Theory
Bishops fourth hypothesis explains SLI as being caused by a conceptual deficit. The conceptual deficit is linked to Piaget’s theory of cognitive development: He saw the ability to represent something symbolically (e.g.: using a symbol as a stand-in for an object) as being important for language. Language itself is inherently symbolic, as we are using words to represent our thoughts, actions, beliefs, etc. Some researchers have hypothesized that children with SLI are impaired in their representation of concepts, but the research findings on this are mixed. Some of the tasks used to assess conceptual development in children require verbal skills to complete, so children with SLI naturally do poorly on these. Also, children with SLI seem to do as well as their typically developing peers, it just takes them longer to do the tasks possibly implicating a speed of processing deficit and not a conceptual deficit. Overall, it appears that conceptual development is not impaired in children with SLI, and that their impairment is due to other factors.

Hypothesis 5- Learning Strategies in SLI
In the fifth theory discussed in this paper, Bishop posits that the underlying problem in SLI is a deficit in the rule-learning process. Specifically, children with SLI may have difficulty testing hypotheses about grammar, a process that is thought to be necessary in learning language. Research has been carried out to test how well children with SLI implicitly learn rules in different kinds of tasks, but results vary depending on the difficulty of the tasks. In some cases, children with SLI learn the rules as well as typically developing peers, but have less success in others. Bishop concludes that children with SLI seem to struggle with rule learning mainly when they are required to retain the rules for a time.

Hypothesis 6- Limited Information Processing Capacity as an Explanation for SLI
A limited information processing capacity model has been proposed as an explanation for SLI, where children with SLI have been thought to have a reduced capacity for verbal processing of information. This model has since been altered to account for a limited information processing system that is not restricted to verbal processing but also accounts for SLI deficits reported for some visuospatial and non-verbal information processing tasks as well. One problem with a limited information processing capacity account of SLI is that the theory is so general that it can explain almost any pattern of results in terms of capacity of limitations. Nevertheless, processing capacity in children with SLI remains an area of continued research interest.


Bloggers for hypotheses 1 through 6, respectively: Alberto Filgueirass, Areej Balilah, Alexandra Smith,  Nicolette Noonan, Laura Pauls, & Monica DaSilva