Developmental apraxia of speech (DAS) is a developmental speech disorder frequently defined as difficulty in programming of sequential speech movements based on presumed underlying neurological differences. Theoretical constructs motivating understanding of DAS have been quite diverse. Motor-based or pre-motor planning speech output deficits (e.g., Hall, Jordon, and Robin, 1993), phonologically based deficits in representation (e.g., Velleman and Strand, 1993), or deficits in neural tissue with organizational consequences (e.g., Crary, 1984; Sussman, 1988) have been posited. Reflecting these varied views of causality, a variety of terms have been employed: developmental apraxia of speech, developmental verbal dyspraxia, and developmental articulatory dyspraxia. Clinically, DAS has most often been defined by exclusion from functional speech disorder or delay using a complex of behavioral symptoms (e.g., Stackhouse, 1992; Shriberg, Aram, and Kwiatkowski, 1997).

The characterization of DAS was originally derived from apraxia of speech in adults, a disorder category based on acquired brain damage resulting in difficulty in programming speech movements (Broca, 1861). Morley, Court, and Miller (1954) first applied the term dyspraxia to children based on a proposed similarity in behavioral correlates with adult apraxic symptoms. A neurological etiology was implied by the analogy but has not been conclusively delineated, even with increasingly sophisticated instrumental techniques for understanding brain-behavior relations (see Bennett and Netsell, 1999; LeNormand et al., 2000). Little coherence and consensus is available in this literature at present. In addition, despite nearly 40 years of research, differential diagnostic correlates and range of severity levels characterizing DAS remain imprecisely defined. Guyette and Deidrich (1981) have suggested that DAS may not be a theoretically or clinically definable entity, as current empirical evidence does not produce any behavioral symptom not overlapping with other categories of developmental speech disorder or delay. In addition, no currently available theoretical constructs specifically disprove other possible theories for the origins of DAS (see Davis, Jakielski, and Marquardt, 1998). In contrast to developmental disorder categories such as hearing impairment or cleft palate, lack of a link of underlying cause or theoretical base with behavioral correlates results in an “etiological” disorder label with no clearly established basis. Evidence for a neurological etiology for DAS is based on behavioral correlates that are ascribed to a neurological basis, thus achieving a circular argument structure for neural origins (Marquardt, Sussman, and Davis, 2000).

Despite the lack of consensus on theoretical motivation, etiology, or empirical evidence precisely defining behavioral correlates, there is some consensus among practicing clinicians as well as researchers (e.g., Shriberg, Aram, and Kwiatkowski, 1997) that DAS exists. It thus represents an incompletely understood disorder that poses important challenges both to practicing clinicians and to the establishment of a consistent research base for overall understanding. An ethical differential diagnosis for clinical intervention and research investigations should, accordingly, be based on awareness of the current state of empirically established data regarding theories and behavioral correlates defining this disorder. Cautious application of the diagnostic label should be the norm, founded on a clear understanding of positive benefits to the client in discerning long-term prognosis, appropriate decisions regarding clinical intervention, and valid theory building to understand the underlying nature of the disorder. Use of DAS as an “umbrella term for children with persisting and serious speech difficulties in the absence of obvious causation, regardless of the precise nature of their unintelligibility” (Stackhouse, 1992, p. 30) is to be avoided. Such practice continues to cloud the issue of precise definition of the presence and prevalence of the disorder in child clinical populations.

Accordingly, a review of the range of behavioral correlates presently in use is of crucial importance to careful definition and understanding of DAS. The relationship of behavioral correlates to differential diagnosis from “functional” speech disorder or delay is of primary importance to discriminating DAS as a subcategory of functional speech disorder. If no single defining characteristic or complex of characteristics emerges to define DAS, the utility of the label is seriously questionable for either clinical or research purposes. In every instance, observed behaviors need to be evaluated against developmental behaviors appropriate to the client's chronological age. In the case of very young clients, the differential diagnosis of DAS is complicated (Davis and Velleman, 2000). Some listed characteristics may be normal aspects of earliest periods of speech and language development (e.g., predominant use of simple syllable shapes or variability in production patterns at the onset of meaningful speech; see Vihman, 1997, for a review of normal phonetic and phonological development).

Before the clinical symptoms presently employed to define DAS are outlined, specific issues with available research will be reviewed briefly. It should be emphasized that behavioral inclusion criteria are not consistently reported and differing criteria are included across studies. Criteria for inclusion in studies then become recognized symptoms of involvement, achieving a circularity that is not helpful for producing valid characterization of the disorder (Stackhouse, 1992). Subject ages vary widely, from preschoolers (Bradford and Dodd, 1996) to adults (Ferry, Hall, and Hicks, 1975). Some studies include control populations of functional speech disorders for differential diagnosis (Stackhouse, 1992; Dodd, 1995); others do not (Horowitz, 1984). Associated language and praxis behaviors are included as differential diagnostic correlates in some studies (Crary and Towne, 1984), while others explicitly exclude these deficits (e.g., Hall, Jordon, and Robin, 1993). Severity is not reported consistently. When it is reported, the basis for assigning severity judgments is inconsistent across studies. A consequence of this inconsistency is lack of consensus on severity level appropriate to the DAS label. In some reports, the defining characteristic is severe and persistent disorder (e.g., Shriberg, Aram, and Kwiatkowski, 1997). In other reports (e.g., Thoonen et al., 1997), a continuum of severity is explored. In the latter conceptualization, DAS can manifest as mild, moderate, or severe speech disorder.

Despite the foregoing critique, the large available literature on DAS suggests some consensus on behavioral correlates that should be evaluated in establishing a differential diagnosis. The range of expression of these characteristics, although frequently cited, has not been specified quantitatively. Accordingly, these behaviors should not be considered definitive but suggestive of directions for future research as well as guidelines for the practicing clinician based on emerging research.

Exclusionary criteria for a differential diagnosis have been suggested in the areas of peripheral motor and sensory function, cognition, and receptive language. Exclusionary criteria frequently noted include (1) no peripheral organic disorder (e.g., cleft palate), (2) no sensory deficit (i.e., in vision or hearing), (3) no peripheral muscle weakness or dysfunction (e.g., dysarthria, cerebral palsy), (4) normal IQ, and (5) normal receptive language.

Phonological and phonetic correlates have also been listed. Descriptive terminology varies from phonetic (e.g., Murdoch et al., 1995) to phonological (Forrest and Morrisette, 1999; Velleman and Shriberg, 1999) according to the theoretical perspective of the researcher, complicating understanding of the nature of the disorder and comparison across studies. In addition, behavioral correlates have been established across studies with highly varied subject pools and differing exclusionary criteria. The range of expression of symptoms is not established (i.e., what types and severity of suprasegmental errors are necessary or sufficient for the diagnosis?). Some characteristics are in common with functional disorders and thus do not constitute a differential diagnostic characteristic (i.e., how limited does the consonant or vowel repertoire have to be to express DAS?). In addition, not all symptoms are consistently reported as being necessary to a diagnosis of DAS (e.g., not all clients show “groping postures of the articulators”). Long-term persistence of clinical symptoms in spite of intensive therapy has also frequently been associated with DAS. Phonological/phonetic correlates reported include (1) limited consonant and vowel phonetic inventory, (2) predominant use of simple syllable shapes, (3) frequent omission of errors, (4) a high incidence of vowel errors, (5) altered suprasegmental characteristics (including rate, pitch, loudness, and nasality), (6) variability and lack of consistent patterning in speech output, (7) increased errors on longer sequences, (8) groping postures, and (9) lack of willingness or ability to imitate a model.

Co-occurring characteristics of DAS in several related areas have also been mentioned frequently. However, co-occurrence may be optional for a differential diagnosis, because these characteristics have not been consistently tracked across available studies. Co-occurring characteristics frequently cited include (1) delays in gross and fine motor skills, (2) poor volitional oral nonverbal skills, (3) inconsistent diadokokinetic rates, (4) delay in syntactic development, and (5) reading and spelling delays.

Clearly, DAS is a problematic diagnostic category for both research and clinical practice. Although it has long been a focus of research and a subject of intense interest to clinicians, little consensus exists on definition, etiology, and characterization of behavioral or neural correlates. Circularity in the way in which etiology and behavioral correlates have been described and studied does not lend to precision in understanding DAS. Research utilizing consistent subject selection criteria is needed to begin to link understanding of DAS to ethical clinical practices in assessment and intervention and to elucidate the underlying causes of this disorder.

See also motor speech involvement in children.