Pragmatics may be defined as “the study of the rules governing the use of language in social contexts” (McTear and Conti-Ramsden, 1992, p. 19). Although there is some debate as to what should be included under the heading of pragmatics, traditionally it has been thought to incorporate behaviors such as communicative intent (speech acts), conversational management (turn taking, topic manipulation, etc.), presuppositional knowledge, and culturally determined rules for linguistic politeness. Some authors, working from a framework in which pragmatics is seen as the motivating force behind other components of language such as syntax and semantics, include an expanded list of behaviors within this domain. For example, from this latter perspective, behaviors such as those occurring in the interactive context of early language acquisition would be considered pragmatic. Despite difficulty in determining where the boundaries of pragmatics should be drawn, as one considers how language is used in real interactions with other people, it is impossible not to cross over into areas more typically seen as social or cultural rather than linguistic (Ninio and Snow, 1996).

Although the social implications of impaired communication skills have been considered for some time, the pragmatic aspects of language impairment did not become a serious topic of study until the mid-1970s, following the lead of researchers studying typical language acquisition (see Leonard, 1998). The innovations for language assessment and intervention that stemmed from this work motivated Duchan (1984) to characterize these efforts as “the pragmatics revolution.” Gallagher (1991) summarized the contributions of pragmatics to assessment by noting that greater awareness of the pragmatic aspects of language resulted in a larger set of behaviors on which a diagnosis of language impairment could be made. It also highlighted the importance of contextual variables in spontaneous language production. Clinicians gained an understanding that controlling or standardizing these variables would fundamentally alter the nature of the interaction.

With respect to language intervention, goals were expanded to include a wide range of pragmatic behaviors. Further, providing intervention in more naturalistic contexts, thereby allowing communication to be motivated and reinforced by natural consequences, was highlighted. At the same time the value of using routines, scripts, and similar procedures to provide greater contextual support for language usage also gained favor among clinicians (Gallagher, 1991).

The study of pragmatics also brought insights into how communication might be linked to other aspects of behavior. A prime example, stemming from work with persons with pervasive developmental disabilities, was the insight that challenging behavior may have communicative intent. Further, providing the individual with a more appropriate means of communicating the same intent would often result in a notable decrease in the undesirable behavior (Carr and Durrand, 1985).

Additionally, much of the recent work focusing on the social skills and peer relationships of children with language impairment (e.g., Hadley and Rice, 1991; Craig and Washington, 1993; Brinton et al., 1997) is a natural extension of earlier work studying the interactional skills of these children. As Gallagher (1999) noted, “it was inevitable that the pragmatic language focus on communication would eventually lead to questions about the interpersonal and intrapersonal roles of language” (p. vi).

Despite the positive contributions to assessment and intervention procedures that have resulted from the study of pragmatics, a clear sense of the role of pragmatic behaviors in language impairment has been difficult to achieve. Research with some groups of children with language impairment has documented the presence of serious pragmatic problems. In other groups of children the nature of pragmatic difficulties has been more challenging to characterize. This variability can be seen by contrasting two groups for which language problems play a major role: children with autism spectrum disorders (ASD) and children with specific language impairment (SLI).

Children with ASD have communication deficits that may be grouped within two categories: (1) the capacity for joint attention to objects and events with other persons, and (2) the ability to understand the symbolic function of language (Wetherby, Prizant, and Schuler, 2000). Pragmatic deficits figure prominently within both categories. For example, with respect to joint attention, children with ASD produce a limited range of communicative intentions. They may communicate to direct the behavior of others but not for purposes requiring joint attention with another person, such as to share feelings or experiences. Children with ASD also have difficulty interpreting and responding to the emotional states of others. This may be reflected in a lack of responsiveness to positive affect as well as in behaviors such as the failure to appropriately coordinate eye gaze during interaction.

Problematic behaviors stemming from difficulty with symbol use include the often cited reliance on reenactment strategies (e.g., a preschool child who used the phrase “do ah” to mean that he was not feeling well, stemming from appropriate use in a prior context), as well as developing maladaptive ways of communicating to compensate for a lack of more conventional means (e.g., using head banging to communicate the desire to avoid an unpleasant task) (Wetherby, Prizant, and Schuler, 2000). Both of these examples of language use have important pragmatic implications.

Whereas it is accepted that pragmatic difficulties constitute a basic problem for children with ASD, the situation is less clear for children with SLI. It is well established that children in the latter group have difficulty with aspects of syntactic, morphological, and semantic development. Studies examining pragmatic behaviors have yielded more equivocal findings. For example, children with SLI have been found to be less capable of responding to stacked requests for clarification (Brinton, Fujiki, and Sonnenberg, 1988), less able to initiate utterances in conversation (Conti-Ramsden and Friel-Patti, 1983), and less adept at entering ongoing conversations (Craig and Washington, 1993) than typically developing peers at the same language level. Other researchers, however, have found that children with SLI performed similarly to typically developing peers on pragmatic variables when language level was controlled (e.g., Fey and Leonard, 1984; Leonard, 1986). In summarizing this work, it appears that many (but not all) children with SLI have difficulty with many (but not all) aspects of pragmatic language behavior. For some of these children, pragmatic deficits stem from problems with language form and content. For other children, pragmatic impairment is a central component of their language difficulty.

One way in which researchers have addressed the variability noted above has been to place children with SLI into subgroups more specifically characterizing the nature of their impairment. Several groups of researchers (e.g., Bishop and Rosenbloom, 1987; Conti-Ramsden, Crutchley, and Botting, 1997) have developed taxonomies identifying a group of children whose language impairments are pragmatic in nature. Labeled as “semantic pragmatic deficit syndrome,” these children are described as having a variety of pragmatic problems in the face of relatively typical structural language skills. Areas of deficit include inappropriate topic manipulation, difficulty assessing shared information, an overly high level of verbosity, and a lack of responsiveness to questions. Word-finding problems and difficulty comprehending language are also associated with this subgroup. In work by Conti-Ramsden, Crutchley, and Botting (1997), semantic pragmatic deficit syndrome characterized approximately 10% of 242 participants with SLI. In follow-up work, Conti-Ramsden and Botting (1999) found that although the subcategories of impairment were stable over the course of a year, many individual children moved between subcategories.

It might be argued that children with pragmatic problems in the face of relatively good structural language (and who do not meet criteria for ASD) might form a separate category of impairment. Bishop (2000) argued that this is not an ideal solution because pragmatic impairment is not limited to children with good structural skills, nor is it always found in association with semantic difficulties. Rather, it may be more productive to view structurally based SLI and ASD as two ends of a continuum on which many combinations of pragmatic and structural language deficits may occur.

In summary, pragmatic impairment may be found in a wide range of children with language problems. In some cases, it may constitute a central component of the impairment. For other children, the problem may be secondary to other types of language problems. From a clinical standpoint, it is important to recognize that children with a variety of disabilities may have pragmatic problems. Given the close link between language and social behavior, it is perhaps as important to recognize that even language impairments that do not involve pragmatic behaviors are likely to have implications for social interaction. To be most productive, interventions should be structured not only to improve specific language skills but also to facilitate the use of language in interactions to improve relationships with peers and adults in the child's social world.