Acoustic communication has become the focus of intensive research aiming to assess, delineate, and interpret the integrity of neural functions within different theoretical frameworks. For example, an increasing number of studies have aimed to document early difficulties in this domain and their potential implications with participants of various developmental disorders, such as autism spectrum disorders or Rett syndrome (RTT). In this research a series of peculiarities have been reported. Publications on delay in acquisition of milestones are increasingly complemented by documentation of qualitative deviances, even at the earliest stages of speech-language acquisition such as cooing and babbling vocalizations (e.g., Marschik et al. Reference Marschik, Pini, Bartl-Pokorny, Duckworth, Gugatschka, Vollmann, Zappella and Einspieler2012; Reference Marschik, Kaufmann, Sigafoos, Wolin, Zhang, Bartl-Pokorny, Pini, Zappella, Tager-Flusberg, Einspieler and Johnston2013; Paul et al. Reference Paul, Fuerst, Ramsay, Chawarska and Klin2011).

Our species-unique ability to frame our world with words and the required neurobiological underpinnings that enable it have fascinated researchers studying phylogenetic and ontogenetic perspectives of language and communication. Contemplations about the origin, evolution, and development of verbal communicative abilities have led to many assumptions, speculations, theories, and attempts to deliver the one and only plausible explanation. In their article, Ackermann et al. postulate an ontogenetic model that assumes age-dependent interactions between basal ganglia and their cortical targets. We discuss the plausibility of and support for this assumption by reviewing recent findings on early vocalizations in infants with neurodevelopmental disorders, more specifically RTT.

From an ontogenetic perspective, early vocalizations – both nonhuman (e.g., as shown for pygmy marmosets) and human – actively promote the proximity and attention of caregivers, and therefore represent an advantage for the babbling infant (Elowson et al. Reference Elowson, Snowdon and Lazaro-Perea1998). But what can we say at this point about the neurobiological substrates of these vocalizations? And, more pragmatically, what does it tell us with regard to progressive neurodevelopmental conditions, such as RTT? Prosodic features of spoken language were reported to be dependent on the integrity of the basal ganglia, especially the striatum (Darkins et al. Reference Darkins, Fromkin and Benson1988; Van Lancker Sidtis et al. Reference Van Lancker Sidtis, Pachana, Cummings and Sidtis2006). From an evolutionary perspective Ackermann et al. argue that a “structural reorganization of the basal ganglia during hominin evolution may have been a pivotal prerequisite for the emergence of spoken language” (sect. 1.2, para. 3). A great body of clinical evidence for the involvement of the basal ganglia in speech-language functions stems from patients with basal ganglia dysfunctions, such as Parkinson's disease or Tourette syndrome. The focus has been on the substantia nigra pars reticulata that exerts inhibitory control of the midbrain periaqueductal gray matter (PAG), a major relay of the descending motor system across vertebrates, and its role in converting emotional and cognitive commands into vocalization (Kittelberger & Bass Reference Kittelberger and Bass2013; Menuet et al. Reference Menuet, Cazals, Gestreau, Borghgraef, Gielis, Dutschmann, Van Leuven and Hilaire2011).

The PAG does not directly control the coordinated activity of respiratory movements, and laryngeal and orofacial muscle groups, but rather projects to the closely related brainstem central pattern generators (CPGs; Hikosaka Reference Hikosaka, Tepper, Abercrombie and Bolam2007). CPGs are neuronal circuits that can produce rhythmic motor patterns in the absence of oscillatory input. Some CPGs operate continuously (e.g., respiratory movements), whereas others are activated to perform specific behavioral tasks (e.g., locomotion). To provide motor output flexibility, supraspinal projections activate, inhibit, and, most of all, modulate the CPG-activity, as does sensory feedback (Einspieler & Marschik Reference Einspieler and Marschik2012; Grillner et al. Reference Grillner, Deliagina, Ekeberg, el Manira, Hill, Lansner, Orlovsky and Wallén1995). CPGs for vocalization have been studied to a great extent not only in amphibians and avians, but also in mammals such as cats (CPGs located in the nucleus retroambiguus; Zhang et al. Reference Zhang, Bandler and Davis1995) or squirrel monkeys (CPGs in the parvocellular reticular formation around the nucleus ambiguus; Hage & Jürgens Reference Hage and Jürgens2006). Barlow et al. (Reference Barlow, Lund, Estep, Kolta and Brudzynski2009) have suggested the same mechanism for early human vocalizations/babbling. A rudimentary understanding of the CPG-circuitry for respiration and mouth movements suggests multiple loci in the brainstem, with a significant role for integration among subsystems and the PAG (Barlow & Estep Reference Barlow and Estep2006).

The above-mentioned tight interconnection of CPGs (Barlow et al. Reference Barlow, Lund, Estep, Kolta and Brudzynski2009) becomes functionally evident when observing individuals with RTT, a neurodevelopmental disorder mainly arising from mutations in the X-linked MECP2 gene (Neul et al. Reference Neul, Kaufmann, Glaze, Christodolou, Clarke, Bahi-Buisson, Leonard, Bailey, Schanen, Zappella, Renieri, Huppke and Percy2010). We have speculated that the interconnectivity of CPGs is pictured in RTT by the apparent evolution of early atypical vocalizations, with inspiratory-modulated sound patterns, into oro-motor dyspraxia and breathing irregularities later in childhood (Marschik et al. Reference Marschik, Pini, Bartl-Pokorny, Duckworth, Gugatschka, Vollmann, Zappella and Einspieler2012). We propose the neuropathology of RTT, a condition with well-documented early atypical vocalizations in both humans and animal models (De Filippis et al. Reference De Filippis, Ricceri and Laviola2010; Marschik et al. Reference Marschik, Pini, Bartl-Pokorny, Duckworth, Gugatschka, Vollmann, Zappella and Einspieler2012, Reference Marschik, Kaufmann, Sigafoos, Wolin, Zhang, Bartl-Pokorny, Pini, Zappella, Tager-Flusberg, Einspieler and Johnston2013), as a model for elucidating abnormalities and their mechanisms involving the CPGs.

In terms of neurobiological substrates, studies of knock-out mouse models of RTT have revealed reduced striatal dopamine release after stimulation that coincided with motor abnormalities (Gantz et al. Reference Gantz, Ford, Neve and Williams2011). Whether such a nigro-striatal pathway involvement could also be associated with abnormal ultrasonic vocalizations, as demonstrated in the Mecp2-308 mouse model (De Filippis et al. Reference De Filippis, Ricceri and Laviola2010), remains open. Of relevance to the vocalization-generating circuitry is the demonstration of decreased PAG volume and length in yet another RTT mouse model (Mecp2B; Belichenko et al. Reference Belichenko, Belichenko, Li, Mobley and Francke2008). Ultimately, the ontogeny of MeCP2 expression in the human brain (Kaufmann et al. Reference Kaufmann, Johnston and Blue2005) supports an early involvement of brainstem monoaminergic nuclei and related brain regions in the pathogenesis of multiple neurologic deficits, including language.

In conclusion, developmental delays and atypicalities in verbal behaviors and other neurologic functions in RTT support CPG and, consequently, brainstem involvement. Future human and animal model studies are needed to further elucidate developing brain–behavior interfaces, disentangle specific traits, and help detect affected children at an earlier age. The “vocal pattern generator” together with other CPGs seems to have great potential in disentangling neurodevelopmental disorders and potentially predict neurological development.