Children with speech sound disorders form a heterogeneous group whose problems differ in severity, scope, etiology, course of recovery, and social consequences. Beyond manifest problems with speech production and use, their problems can include reduced intelligibility, risk for broader communication disorders, and academic difficulties, as well as social stigma.

Because of the heterogeneity of children's speech sound disorders, the description and classification of these disorders have been attempted from a variety of perspectives, with persisting controversy as a predictable result. Nonetheless, one distinction that has garnered relatively universal support is the division of children's speech disorders into those that are developmental (with onset in early or middle childhood, e.g., before age 9) and those that are nondevelopmental (occurring after that time period and resulting from known causes). Developmental disorders have received substantially more research attention to date.

A second widely accepted distinction separates developmental disorders with known causes from those without. For developmental speech disorders of known causes in children, the terminology has been relatively stable and has typically referenced etiological factors (e.g., speech disorders due to mental retardation, cleft palate). In contrast, the terminology for children's speech disorders of unknown origin is less stable, reflecting uncertainty about their nature and origin. During the past 30 years, commonly used terms have included functional articulation disorders, phonological disorders (Locke, 1983), articulation and phonological disorders (Bernthal and Bankson, 1998), and persistent sound system disorders (Shelton, 1993).

Proposed classifications of child speech disorders have been advanced along descriptive, predictive, and clinical grounds (Shriberg, 1994). Three classifications currently warrant particular attention either because of empirical support (those associated with Shriberg and Dodd) or practical significance (the Diagnostic and Statistical Manual of Mental Disorders-IV-TR; American Psychiatric Association, 2000).

Currently, the most comprehensive and rigorously studied classification is the Speech Disorders Classification System, developed through a 20-year program of research by Shriberg and his colleagues (Shriberg, 1994, 1999; Shriberg and Kwiatkowski, 1982, 1994a, 1994b, 1994c; Shriberg et al., 1997). This evolving classification is designed to provide a framework for identifying and describing subtypes and testing etiological hypotheses. At this time, it is primarily a research tool.

Within this classification, children's speech sound disorders of unknown origin are divided into speech delay and residual error categories. Speech delay, with an estimated prevalence of 3.8% among 6-year-olds (Shriberg, Tomblin, and McSweeney, 1999), is characterized by reduced intelligibility and increased risk for broader communication and academic difficulties. It encompasses more severe forms of speech disorder. Residual speech errors, with a tentatively estimated prevalence of 5% among individuals older than age 9 (Shriberg, 1994), is characterized by the presence of at least one speech sound error (often involving distortion of a sibilant fricative or liquid) that persists past the developmental period. Although the category of residual speech errors encompasses less severe forms of speech disorder that are associated with neither reduced intelligibility nor broader communication difficulties, disorders in this category remain of interest for theoretical reasons (i.e., genetic versus environmental origin) and for their potential social and vocational costs, which for some individuals can continue throughout life.

Within the Speech Disorders Classification System, the major categories of speech delay and residual speech errors are further divided according to suspected etiological factors or developmental pattern. Five subtypes of speech delay are postulated in relation to the following possible causes: genetic transmission, early history of recurrent otitis media with effusion (Shriberg et al., 2000), motor speech involvement associated with developmental apraxia of speech (Shriberg, Aram, and Kwiatkowski, 1997), motor speech involvement associated with mild dysarthria, and developmental psychosocial involvement. In each case the etiological factor is considered dominant in a mechanism that is suspected to be multifactorial in nature. Two subtypes of residual speech errors are proposed, those found in association with a documented history of speech delay (residual error-A) and those for which no previous history of speech disorder was reported (residual error-B) (e.g., Shriberg et al., 2001). Ongoing research is aimed at increasing understanding of the causal, developmental, and cognitive processing mechanisms underlying each of these five subtypes.

The classification description described by Dodd (1995) is well motivated from theoretical perspectives and based on processing accounts of child speech disorders, but has been less thoroughly validated than the Speech Disorders Classification System. Nontheless, Dodd's classification system has been supported by studies that examine characteristics of clinical populations (Dodd, 1995), error patterns across languages (e.g., Fox and Dodd, 2001), bilingual children's generalization patterns (Holm and Dodd, 2001), and treatment efficacy (Dodd and Bradford, 2000). Thus, its empirical support is growing rapidly. It is intended primarily to be used to aid in differential diagnosis and clinical management, and was proposed as a system that uniquely combined four historical approaches to classifying speech disorders. These four approaches were based on age at onset, severity, causal and maintenance factors, and description of symptoms, respectively.

Dodd's classification system recognizes five subtypes: articulation disorder, delayed phonological acquisition, consistent deviant disorder, inconsistent disorder, and other. Within this system, an articulation disorder is defined as inability to produce an undistorted version of a speech sound or sounds that are expected, given the child's age. English sounds that are often affected in such disorders include /s/, /r/, and the interdental fricatives. This label is applied regardless of whether the cause is an anatomical anomaly or is unknown. Delayed phonological acquisition is defined in cases where a child's speech errors are consistent with those seen in younger, normally developing children. Consistent deviant disorder is the label applied when a child demonstrates a reduced variety of syllable structure use as well as errors that are atypical of those seen in normal development.

Inconsistent disorder is identified when a child's productions are inconsistent in ways that cannot be explained by complex phonological rules or the effects of linguistic load on production. Ozanne (1995) described a study suggesting that inconsistent disorder represents one subgroup associated with developmental verbal dyspraxia (the condition referred to by Shriberg and others as developmental [or childhood] apraxia of speech). Inconsistent disorder is operationally defined using a 25-item word list. The child is asked to produce each word three times, with inconsistency noted when the child produces at least ten of the words differently on two of the three elicited productions. Dodd's “other” category encompasses suspected motor speech disorders.

The DSM-IV-TR (American Psychiatric Association, 2000) classification represents the most streamlined classification of children's speech disorders and one that is perhaps more familiar than others to a broad range of speech pathologists, who use it for billing purposes, and non-speech-language pathologists who come in contact with children with childhood speech disorders. Within this classification, Phonological Disorders 315.39 (formerly Developmental Articulation Disorders) is nested within Communication Disorders. Communication Disorders, in turns, falls under the relatively large category Disorders Usually First Diagnosed in Infancy, Childhood, or Adolescence. This category includes, among others, mental retardation, learning disorders (learning disabilities), and pervasive developmental disorders.

In the most recent DSM-IV classification, phonological disorders is defined by the failure to use speech sounds that are expected given the child's age and dialect. Although subtypes are not described, the American Psychiatric Association's description of the category phonological disorders acknowledges that errors may reflect difficulties in peripheral production as well as more abstract difficulties in the child's representation and use of the sound system of the target language. Under comments on differential diagnosis, phonological disorders is described as a possible secondary diagnosis when speech errors in excess of expectations are noted in association with disorders that might be considered known causes for speech difficulties (viz., mental retardation, hearing impairment or other sensory deficit, speech motor deficit, and severe environmental deprivation). Speech difficulties that may be associated with the term “Developmental (or Childhood) Apraxia of Speech” are addressed neither in the DSM-IV criteria nor as a subclassification, although that term is described as a possible label for some forms of phonological disorder in DSM-IV-TR—a revision designed to increase the currency of the DSM without changing the actual classificatory categories.

Classifications of children's speech disorders are encumbered by demands that they address numerous audiences and unresolved controversies. Among the current audiences to be served are clinicians, researchers, and administrators. Each of the three classifications described here addresses those audiences to a different degree. One of the unresolved controversies that has been addressed to some degree by each is the status of clinically postulated entities such as developmental or childhood apraxia of speech. A second controversy, and one that is being addressed by Shriberg and colleagues, relates to how child speech sound disorders should be conceptualized in relation to other communication disorders that frequently co-occur, but are associated with causal mechanisms that are more ill-defined than those for conditions that fall under child speech disorders with known causes (e.g., hearing loss, cleft palate). Two that are of particular interest are specific language impairment (Shriberg, Tomblin, and McSweeney, 1999) and stuttering (Guitar, 1998). In addition, classifications are ideally consistent with developmental as well as psychological processing accounts of the manifest behaviors associated with child speech disorders. The accounts of Shriberg, Dodd, and their colleagues appear to be pursuing these challenging issues.