Training four-syllable CV patterns in individuals with acquired apraxia of speech: Theoretical implications

Background: Acquired apraxia of speech (AOS) is known as a motor speech disorder because it affects both the motor planning and motor programming components of speech production (Darley, Aronson, & Brown, 1975; Kent & Rosenbek, 1983; Square & Martin, 1994). Although the clinical characteristics of acquired AOS are generally accepted, what specifically causes individuals to make the errors they do is less well understood. Four theories have been proposed that attempt to explain these errors. Few treatment studies have taken into account these theories when designing their studies. This could account for the limited generalisation effects that have been reported (Ballard, 2001).Aims: This study investigated the effects of a treatment for three individuals with moderate acquired AOS and concomitant aphasia. Treatment focused on training four-syllable CV patterns of varying levels of difficulty. Acquisition, as well as generalisation effects and error patterns were examined and interpreted based on the proposed theoretical models of AOS in the literature.Methods & Procedures: A single-subject multiple baseline design across behaviours and subjects was utilised to train each participant's ability to produce four-syllable CV patterns varying in levels of difficulty. A total of 277 stimuli were used in the study: 32 trained and untrained four-syllable CV patterns, 203 words and nonsense words, divided into one- to five-syllable words, and 42 common phrases. The CV patterns were further divided into specified levels of difficulty. Generalisation to untrained four-syllable CV patterns at the same level and across levels, as well as to words and phrases, was examined.Outcomes & Results: All participants were successful in learning to produce the four-syllable CV patterns on which they were trained. No generalisation to the untrained patterns at the same level occurred. This finding suggested that some aspects of the phonological representations, specifically voicing, are not intact prior to the production being sent to the phonetic-motoric level, as the phonetic-motoric theory suggests (McNeil, Odell, Miller, & Hunter, 1995; McNeil, Robin, & Schmidt, 1997; Shriberg, Aram, & Kwaitowski, 1997a, 1997b, 1997c; Van der Merwe, 1997). Generalisation to untrained patterns across levels that were considered less complex than the pattern being trained was seen. This finding supports the conclusions of Gierut (1998), Thompson, Ballard, and Shapiro, (1998), and Plaut (1996) that training complex behaviours generalises to less complex behaviours. Improvement of word/phrase production as treatment on the four-syllable CV patterns proceeded was not consistent with the motor theory of AOS (Ziegler, 2003). Instead it supports the more integrative theory of speech production proposed by Ballard, Robin, and Folkin (2003).Conclusions: This study provided the opportunity to examine the theories attempting to explain the characteristics of AOS. Future treatment studies should consider these theories when designing their studies to further test and to advance our understanding of AOS.