Contemporary issues in apraxia of speech

ABSTRACT

Background: Apraxia of speech (AOS) has been central in aphasiology for over 200 years. It has generated theoretical and clinical controversy and contributed significantly to the development of knowledge of how speech is formulated and generated by the brain.

Main Contribution: This short review introduces theoretical, experimental and clinical studies concerning current issues in apraxia of speech.

Conclusions: Progress in our understanding of apraxia of speech is advanced by the contemporary issues presented in the studies making up this special issue in aphasiology.

KEYWORDS:

• Apraxia of speech
• motor speech disorders
• speech production models
• treatment of apraxia of speech
• the neural architecture of apraxia of speech

Introduction

Apraxia of speech (AOS) has been central in aphasiology since the mid-19th century when Baudelaire, Broca, Hughlings Jackson, Liepmann and others first began to differentiate the condition from aphasia in modern times (Tesak & Code). In more contemporary decades, interest and controversy developed with the work of Fred Darley and his colleagues at the Mayo Clinic in the 1950s and 1960s who did the necessary work to establish some of the foundational groundwork. The main result of this work was an acceptance by most that AOS was a distinct and separate condition to aphasia. Darley and colleagues (1975) described a three-stage model of speech production where AOS was due to a breakdown at the “motor speech programmer”.

Apart from more central issues, what the condition is actually called has been a topic of discussion, and even now it goes by a number of pseudonyms: aphemia, cortical anarthria, phonetic disintegration, motor aphasia, verbal and articulatory a/dyspraxia, to list the main ones used in different languages and by different disciplines. In continental Europe, especially in the medical profession, AOS still goes by different names like aphemia (Broca’s own preference) and anarthria, and many continue to feel that the term “motor aphasia” is a useful syndrome label, both theoretically and clinically. But the majority of the contemporary aphasiological clinical and research community are happy to settle on apraxia of speech.

There are a few who still cling to the view that there is no such thing as AOS, failing to recognise, or choosing to ignore, a significant body of experimental and clinical research from several disciplinary groups. Many believe that Broca’s (1861) famous patient Louis Victor Laborgne had AoS, a condition Broca called aphemia. Broca described Leborgne’s problem as a severe inability to articulate in the absence of muscular paralysis or incoordination and, following autopsy, found a massive old lesion in the 3rd frontal gyrus of the left inferior frontal lobe. He proclaimed that this area of the brain was “la siège de la faculté du langage articulé” (the seat of the faculty of articulated language), supporting Jean-Baptiste Bouillaud’s (1796–1881) earlier claims concerning the role of the frontal brain in speech production. Broca emphasized that AOS is a disorder of speech, and not of language. LaBorgne had little speech except famously for the repeated nonlexical speech automatism,/tan, tan/. Not generally known is that LeBorgne had a second automatism, this was the lexical speech automatism, Sacre Nom de Dieu (Sacred name of God), made up of real words Code (2013).

Subsequently, AoS is seen in contemporary times as a disorder resulting from damage between the abstract phonological component of language representation and the phonetic execution of articulatory speech causing speech initiation and coordination problems, articulatory groping and searching, disturbed prosody, omissions, substitutions and distortions of speech phones, and problems sequencing the correct order of phone production. The disorder is seen as one caused by disruption to the speech planning and programming system that translates the abstract phonological representation into speech.

The diagnosis of AoS depends on the presence of certain symptoms, but controversy about what those symptoms are continues. The established traditional symptoms are articulatory searching and groping for speech, failures in initiation of speech and a disassociation of voluntary and involuntary actions. The individual is typically able to carry out involuntary and automatic actions, which are relatively preserved, but impaired in their attempts to produce voluntary actions – the classical failure to lick the lips to command when able to carry out the same behaviour automatically when eating and drinking. Omissions, substitutions, distortions and additions of speech phones are classic features with issues in sequencing the correct order of phones. Rosenbek et al. (1984) list the main features as errors increase as the complexity of the motor task increases. Thus, vowels are easier than consonants to produce and single consonants are easier than clusters of consonants; errors occur more often at low frequencies than at higher frequency consonants; errors increase with word length and occur more in imitation than in spontaneous speech; errors occur less often on automatic than on more propositional speech (see Code, 1998, for review). McNeil et al. (2009) have identified what they contend are the two “core symptoms” of AoS: Lengthened Segment Durations and Lengthened Intersegment Durations that all people with AoS have. Other features may be secondary, and searching and groping they consider not to be a differential symptom.

Several papers on this issue are concerned with developing and describing models of the speech production process and its impairment. Most linguistic models assume an abstract phonology. It is here that speech sounds are encoded into strings that are transmitted to a component that plans and programmes the co-ordinated utterances for speech execution. It is impairment at this planning and programming mechanism that is considered to be the locus for AoS. On Levelt’s (1989) model buffering holds the phonetic plan ready for execution. The phonetic plan is a representation of instructions for articulation of the utterance. Levelt’s influential model was utilised by Van Der Merwe (2021) to develop a model of AoS where planning entails the retrieval of the core motor plan which specifies the voice, place and manner of articulation and programming lays down the pattern of muscle tone, resistance, and force of movement. Impairments of programming result in muscle tone or reflex deficits (Van Der Merwe, 2021). Invoking the model, others have restricted AoS to impairments of planning rather than speech programming, the latter component produces either dysarthric or paraphasic speech (Van Der Merwe, 2021).

The influential Directions into Velocities of Articulators (DIVA, responsible for speech motor control) and Gradient Order DIVA (GODIVA, responsible for speech sequencing) models (Guenther et al., 2006; Miller & Guenther, 2021) has been developed in recent decades utilising brain imaging and computational techniques. These models have been applied to AOS, where impairments in the selection of the correct motor program are the proposed core impairment (Miller & Guenther, 2021).

A nonlinear gestural model of AOS was developed and is described in this volume by Ziegler and colleagues (Ziegler et al., 2012, 2021) that posits that speech is laid down in procedural motor memory in development in the inferior frontal cortex. Connections are established here between articulatory gestures as the foundation for syllable and word forms. The model is informed by articulatory phonology (Goldstein et al., 2006) which brings together phonetic fact and abstract phonology such that the abstract representations of words produced by the phonology are bonded gestural units whose articulated form is constrained by the mechanical limitations of the phono-articulatory mechanism. AoS disrupts these learnt gestural connections established and stored in procedural motor memory.

Many approaches to the treatment of AoS have been proposed and researched over the years. Several systematic reviews have been undertaken to examine the range and effectiveness of treatments for AoS in recent years and provide guidelines for treatment (Wambaugh et al., 2006a, 2006b; Ballard et al., 2015) covering all AOS treatment research published in English until 2002. In this issue, Wambaugh (2021) has filled the gap since the 2015 review and provided an update on new developments in AOS treatment that have occurred since. She shows that research into treatments reveals that the most popular approaches around the world are articulatory-kinematic approaches, theoretically based on the principles of motor learning (PML). Wambaugh (2021) discusses new evidence supporting the use of Motor Learning Guided (MLG) treatment, which utilises PML. Musical/rhythmic approaches are also well represented and there have been advances in technologically enhanced behavioural treatments, such as transcranial magnetic stimulation variants.

Haley et al. (2021a) described a treatment programme for a woman with AOS and aphasia that involved her crucially building her own practice programme based on the personal vocabulary she wanted, together with coaching by the clinician. Options for practice are provided by the clinician, but chosen by the participant, and structured on a custom app. installed on a tablet computer.

A long-running issue in AOS research has been the degree to which apraxic speech differs from aphasic-phonemic paraphasia, and in what ways. In this issue two papers examine the issue. Haley et al. (2021a) set out to determine whether speakers with AOS make consistent (in type) and invariable (in location) segmental sound errors compared to speakers with aphasia with paraphasia. They compare error consistency in a large sample of speakers with aphasia and AOS and speakers with aphasia but no AOS, using sequential multisyllabic word repetition. They reported that in their participants with AOS sound errors were inconsistent rather than consistent, historically established as an important diagnostic criterion. The findings strongly support the view that the diagnosis of AOS compared to aphasia should not be based on finding more consistency in error production, as has been proposed (McNeil et al., 2009) and should be removed from diagnostic checklists until further supportive research has been produced.

Another question asked in contemporary times is: do different varieties or subtypes of AoS occur? Three articles on this issue address this question (Duffy et al., 2021; Mailend & Maas, 2001; Ramoo et al. 2021). Mailend and Maas review the experimental and clinical research and find that there have been few studies examining the question. They conclude that a great deal more experimental research is needed before the question can be addressed. Duffy et al. (2021) described a recent Mayo Clinic study (Utianski et al., 2018) that suggested that separate prosodic and phonetic subtypes of progressive AOS are found (see Duffy et al., 2021 for discussion). Ramoo et al. (2021) undertook extensive analysis of the speech of a single individual who was diagnosed with a conduction aphasia and AOS, with good semantic, syntactic, and phonological input processing, and good phonological short-term memory. His problems lie mainly in word production involving stages following the lexical access stage with particular difficulties computing articulatory plans rather than in selection and retention of phonological representations. The case shows that there is variability in how AoS impairments can present themselves and the difficulties in identifying different varieties of phonological and articulatory disorders that should be distinguished in clinical work, beyond a simple dichotomy between conduction aphasia and AoS. These separate potential limitations involve transfer from phonology to articulation limitations and difficulties in computing, selecting and/or initiating articulatory plans.

The neural architecture of apraxia of speech

A range of functional networks essential for speech and language converge in the left inferior frontal cortex. Imaging and electrophysiological studies with neuro-typical volunteers have identified an enlarged left frontal region engaged in various aspects of speech and language, including, but beyond, Broca’s area in inferior frontal cortex BA 44 (pars opercularis) and BA 45 (pars triangularis), including the supplementary motor area (SMA) in ventral premotor cortex (BA 6). In the enlarged Broca’s area, called Broca’s complex by Hagoort (2005), activation is seen while participants engage in semantic, syntactic and phonological processing and speech planning. A central cognitive selection and binding or unification role for Broca’s complex is advocated with a particular role in maintaining information online while the binding operations can take place. This maintenance of information is achieved via working memory. The inferior frontal cortex is considered to be the area where the articulatory loop component of working memory engages in the storage and manipulation (Fiez et al., 1996; Paulesu et al., 1993). The region has been identified as the location for the learnt procedural knowledge for the production of the speech sound patterns is assembled and stored in development. It evolved into a neural basis for speech motor planning (Ziegler et al., 2021).

The stroke that causes AOS, by definition, is swift. It results from impairment to the supply of oxygen-carrying blood by the middle cerebral artery. The area damaged can be extensive and diffuse, and in larger vascular strokes, subcortical areas are affected. Progressive conditions, in contrast, are insidious neural disease processes that destroy tissue and inevitably lead to cortical and subcortical atrophy. Primary progressive AOS (ppAOS) begins in premotor and supplementary motor area, inferior frontal cortex and anterior temporal areas. In many cases of ppAOS atrophy travels to subcortical areas and the syndrome devolves to progressive supranuclear palsy and cortico-basal syndrome (Duffy et al., 2021).

Conclusions

Overall, the collection of papers included here addresses contemporary issues and controversies currently in the apraxia of speech. They present experimental and discursive questions contributing to the foundation for future research programmes for theoretical, experimental and clinical advancement.

Disclosure statement

No potential conflict of interest was reported by the author(s).