28.2.1 Acoustic Similarities between Music and Speech Rhythm

Music and speech are acoustic signals that are structured and delivered in time, containing frequency (pitch), intensity (loudness), timbre (sound quality), and duration (timing) information. For each domain, the ordering and duration of the events in time are referred to as rhythm (London, Reference London2012; McAuley, Reference McAuley and Jones2010). Importantly for the current discussion, a major difference between music and speech rhythm is that music is typically structured in time so that events occur regularly, and their onsets can be easily predicted by listeners (Jones, Reference Jones2018; Patel and Morgan, Reference Patel and Morgan2017). In most music, this regularity in rhythm gives rise to the perception of an underlying regular (isochronous) beat, which listeners are able to extract and to move along with (e.g., Repp and Su, Reference Repp and Su2013), and which drives auditory-motor activation and connection (Morillon and Baillet, Reference Morillon and Baillet2017; Zatorre et al., Reference Zatorre, Chen and Penhune2007). As has been noted throughout this book (see Chapter 26), speech rhythm does not contain this same kind of regularity, and intervals between speech elements (syllables, words) are not perfectly regular. There is a large discussion in the literature around whether speech can be considered rhythmic (Nolan and Jeon, Reference Nolan and Jeon2014; Turk and Shattuck-Hufnagel, Reference Turk and Shattuck-Hufnagel2013), and a long history of searching for periodicity in speech has proven difficult (i.e., searching for measures to categorize “stress-timed” versus “syllable-timed” languages; see Chapters 7 and 9). However, research has shown that speech rhythm emerges through the interaction of accentual, prosodic, and lexical factors, which create patterns of stress (Beier and Ferreira, Reference Beier and Ferreira2018; Goswami and Leong, Reference Goswami and Leong2013; Kohler, Reference Kohler2009) and prominence (see Kohler, Reference Kohler2009). These rhythmic patterns, which have also been referred to more globally as prosodic structure and surface speech timing (Turk and Shattuck-Hufnagel, Reference Turk and Shattuck-Hufnagel2013), have been shown to aid the segmentation of the acoustic signal into syllables, words, and phrases (Cutler, Reference Cutler1994; Spinelli et al., Reference Spinelli, Grimault, Meunier and Welby2010), allow for the prediction of upcoming information (Beier and Ferreira, Reference Beier and Ferreira2018), and facilitate turn-taking between speakers (see Pickering and Garrod, Reference Pickering and Garrod2013).

In both music and speech, the structuring of acoustic events in time can be described as creating a hierarchical structure, where smaller event units are embedded within larger event groups or chunks. In music, interacting acoustic and temporal patterns create – based on the underlying regular pulse – metrical structures with a metric hierarchy, leading to the perception of strong and weak beats (Lerdahl and Jackendoff, Reference Lerdahl and Jackendoff1983; London, Reference London2012; Povel and Essens, Reference Povel and Essens1985). In speech, the interacting acoustic and temporal patterns relate to the combination of phonemes, syllables, and words into phrases and sentences that can also be integrated within a rhythmic hierarchy of more and less prominently accented elements (Beier and Ferreira, Reference Beier and Ferreira2018). The similarities and differences in acoustic cues, regularity, and hierarchically structured events suggest that the combined or at least comparative investigation of music and speech rhythm processing can provide novel insights into perceptual and cognitive processing and sequencing, with potentially shared mechanisms.

28.2.2 Cognitive Similarities between Music and Speech Rhythm Processing

Empirical research has suggested that the acoustic and hierarchical features of music and speech may be processed by the brain in a similar way (Heard and Lee, Reference Heard and Lee2020; Patel, Reference Patel2008), and several theoretical frameworks have been proposed aiming to better understand cognitive connections between music and speech rhythm processing as well as their neural correlates (e.g., Fujii and Wan, Reference Fujii and Wan2014; Patel, Reference Patel2011; Tierney and Kraus, Reference Tierney and Kraus2014). Here, we will outline the Processing Rhythm in Speech and Music (PRISM) framework, which focuses on three mechanisms underlying speech and music rhythm processing: precise auditory processing, synchronization/entrainment of neural oscillations to an external rhythm, and sensorimotor coupling (Fiveash et al., Reference Fiveash, Bedoin, Gordon and Tillmann2021; see Figure 28.1). PRISM was developed on the basis of previous theoretical frameworks of music, speech, or music and speech processing that each discuss some of the individual mechanisms potentially involved in the processing of one or both domains. The goal of PRISM is to provide a larger perspective that traverses music and speech. PRISM extracts and combines the most central mechanisms discussed across different theories into a single framework, which can inform new research. In addition, PRISM also aims to provide future directions to identify specific mechanisms that could be underlying the effects of music rhythm training in the long term, and rhythmic stimulation in the short term, on speech and language processing.

Figure 28.1

The PRISM framework.

The three mechanisms proposed in the PRISM framework (Fiveash et al., Reference Fiveash, Bedoin, Gordon and Tillmann2021).

Figure 28.1 Long description

It shows three interconnected circles, reflecting three mechanism for rhythmic processing in music and speech: Precise Auditory Processing, Synchronization or Entrainment of Neural Oscillations, and Sensorimotor Coupling. The arrows show the connections between mechanisms.

Copyright © 2021 by American Psychological Association. Reproduced with permission (Fiveash et al., 2021).

Among the incorporated frameworks, the following three were the most influential when creating PRISM: the Sound Envelope Processing and Synchronization and Entrainment to Pulse (SEP) hypothesis (Fujii and Wan, Reference Fujii and Wan2014), the Precise Auditory Timing Hypothesis (PATH) (Tierney and Kraus, Reference Tierney and Kraus2014), and the Temporal Sampling Framework of developmental dyslexia (TSF) (Goswami, Reference Goswami2011). Larger theoretical frameworks including the Overlap, Precision, Emotion, Repetition, and Attention (OPERA) hypothesis (e.g., Patel, Reference Patel2014), the Dynamic Attending Theory (Jones, Reference Jones1976, Reference Jones2018; Jones and Boltz, Reference Jones and Boltz1989), and active sensing approaches (Morillon et al., Reference Morillon, Hackett, Kajikawa and Schroeder2015; Schroeder et al., Reference Schroeder, Wilson, Radman, Scharfman and Lakatos2010) also informed the current framework. Each of these frameworks point towards different, though partly overlapping, aspects of the music and speech/language connection, which PRISM aimed to combine with the proposition of three mechanisms, as presented below. Outlining these three mechanisms and their potential interactions – both for deficits and expertise – lead to new research perspectives for both fundamental and clinical research.

28.3.1 Encoding, Segmentation, and Phonological Awareness

One link between musical rhythm skills and low-level speech processing is based on the processing of acoustic features. At the encoding stage of sound events (i.e., when acoustic features of a sound are first processed), musicians (compared to nonmusicians) show earlier and larger brainstem responses to the onset of both musical sounds and speech syllables (Musacchia et al., Reference Musacchia, Sams, Skoe and Kraus2007). Compared to nonmusicians, musicians also show enhanced syllable discrimination abilities based on amplitude rise time and voice onset time (Zuk et al., Reference Zuk, Ozernov-Palchik and Kim2013), perform better in speech-timing discrimination tasks (Sares et al., Reference Sares, Foster, Allen and Hyde2018), are quicker and more accurate at segmenting words in an artificial language (François et al., Reference François, Jaillet, Takerkart and Schön2014), and are more sensitive to metrically incongruous words (Marie et al., Reference Marie, Magne and Besson2011). Correlational studies have shown that (1) rhythm perception proficiency (as measured with a same-different rhythm task) is correlated with the amplitude of neural responses to violations in speech rhythm for nonmusicians (Magne et al., Reference Magne, Jordan and Gordon2016); (2) beat entrainment abilities in young children are correlated with more precise neural encoding of the speech envelope (Woodruff Carr et al., Reference Woodruff Carr, White-Schwoch, Tierney, Strait and Kraus2014); and (3) higher music rhythm perception abilities are correlated with more consistent grouping of speech rhythms in adults with varying musical ability (Boll-Avetisyan et al., Reference Boll-Avetisyan, Bhatara and Höhle2017). These rhythm-related skills as well as speech encoding and segmentation skills may contribute to phonological awareness abilities.

Phonological awareness is a central element of language learning and relates to the ability to manipulate and break up units of speech sounds, such as words, syllables, and phonemes (Ball, Reference Ball1993; Gordon et al., Reference Gordon, Fehd and McCandliss2015a). Several studies have shown links between phonological awareness and musical rhythm skills in school-aged children. For example, in typically developing children, phonological awareness measures correlate with performance in rhythmic discrimination tasks (Ozernov-Palchik et al., Reference Ozernov-Palchik, Wolf and Patel2018), as well as rhythm perception and production tasks (Degé et al., Reference Degé, Kubicek and Schwarzer2015), rhythm and melody repetition tasks (Cohrdes et al., Reference Cohrdes, Grolig and Schroeder2016), and rhythm production skills over development (David et al., Reference David, Wade-Woolley, Kirby and Smithrim2007). Links between rhythm perception and production skills and phonological awareness skills have also been shown in pathology, notably for children with dyslexia (Flaugnacco et al., Reference Flaugnacco, Lopez and Terribili2014; Huss et al., Reference Huss, Verney, Fosker, Mead and Goswami2011; Thomson and Goswami, Reference Thomson and Goswami2008) and developmental language disorder (Corriveau and Goswami, Reference Corriveau and Goswami2009). For example, Flaugnacco et al. (Reference Flaugnacco, Lopez and Terribili2014) showed that both rhythm reproduction and rhythm tapping correlated with phonological awareness in children with dyslexia (see also correlations with reading skills). Note that the potential relations between “phonological awareness” and “rhythm” abilities vary across studies, and that these patterns of correlation (versus noncorrelation) seem to be influenced by the types of tasks used to measure these broader constructs.Footnote ¹

28.3.2 Reading, Syntax, and Spelling

Music rhythm skills have also been shown to correlate with reading, syntax, and spelling abilities. For example, music aptitude (especially music perception skills) and reading ability were shown to correlate with subcortical responses to predictable speech sounds, and these subcortical responses were reduced in poor readers (Strait et al., Reference Strait, Hornickel and Kraus2011). Groups with musical training also showed enhanced novel word learning (through picture–word associations) for children (Dittinger et al., Reference Dittinger, Chobert, Ziegler and Besson2017) and adults (Dittinger et al., Reference Dittinger, Barbaroux and D’Imperio2016). Further, rhythm perception abilities correlate with rapid word naming (Bekius et al., Reference Bekius, Cope and Grube2016), enhanced grammatical skills (Gordon et al., Reference Gordon, Shivers and Wieland2015b), and both reading and spelling skills while controlling for the effect of verbal ability (Douglas and Willatts, Reference Douglas and Willatts1994), and reading attainment (Holliman et al., Reference Holliman, Wood and Sheehy2010). Converging results have been reported for rhythm production abilities. For example, paced tapping performance correlates with word and nonword reading (González‐Trujillo et al., Reference González‐Trujillo, Defior and Gutiérrez‐Palma2014; Tierney and Kraus, Reference Tierney and Kraus2013), and rhythm production skills at school entry (~six years) predict: (1) reading and spelling at the end of the first year of school (Kertész and Honbolygó, Reference Kertész and Honbolygó2021; Lundetræ and Thomson, Reference Lundetræ and Thomson2018), (2) reading skills at the end of the second year of school (Dellatolas et al., Reference Dellatolas, Watier, Le Normand, Lubart and Chevrie-Muller2009), and (3) reading and spelling skills in the third year of school (in the same children as listed in (1): Kertész and Honbolygó, Reference Kertész and Honbolygó2023).

Links between rhythm perception and production and reading skills have also been shown in children with dyslexia (Flaugnacco et al., Reference Flaugnacco, Lopez and Terribili2014; Goswami et al., Reference Goswami, Thomson and Richardson2002; Huss et al., Reference Huss, Verney, Fosker, Mead and Goswami2011; Thomson and Goswami, Reference Thomson and Goswami2008) and developmental language disorder (Corriveau and Goswami, Reference Corriveau and Goswami2009). For example, Corriveau and Goswami (Reference Corriveau and Goswami2009) showed that children with developmental language disorder were impaired in paced tapping compared to age-matched control children, and that their paced tapping performance was related to reading and spelling measures.

28.4.1 Short-Term Rhythmic Stimulation

Experimental manipulations of music and speech materials have shown that rhythmic stimulation and entrainment to music can influence the subsequent processing of a speech signal – an experimental paradigm referred to as rhythmic cueing or rhythmic priming.

Rhythmic cueing studies involve a short rhythmic cue that matches one to one with the accents (stressed syllables) of a subsequent sentence.Footnote ² These studies have shown that a matching rhythmic cue preceding a sentence enhances phoneme detection (Cason et al., Reference Cason, Astésano and Schön2015a; Cason and Schön, Reference Cason and Schön2012) and can enhance the neural response to a subsequent sentence (Falk et al., Reference Falk, Lanzilotti and Schön2017) compared to an irregular, mismatching cue. This effect can be boosted by simultaneous movement (e.g., tapping along; Falk and Dalla Bella, Reference Falk and Dalla Bella2016) and can also influence the sentence repetition performance of hearing-impaired children (Cason et al., Reference Cason, Hidalgo, Isoard, Roman and Schön2015b).

Rhythmic priming studies involve a longer rhythmic prime, typically 30 seconds, followed by a set of naturally spoken sentences. Such rhythmic priming experiments have shown that children and adults perform better at grammaticality judgments for auditorily presented sentences when presented after regular rhythmic primes compared to after irregular rhythmic primes or other control conditions (Canette et al., Reference Canette, Lalitte and Bedoin2020b; Chern et al., Reference Chern, Tillmann, Vaughan and Gordon2018; Fiveash et al., Reference Fiveash, Bedoin, Lalitte and Tillmann2020). This effect has also been observed for children with dyslexia and developmental language disorder (Bedoin et al., Reference Bedoin, Brisseau, Molinier, Roch and Tillmann2016; Ladányi et al., Reference Ladányi, Lukács and Gervain2021; Przybylski et al., Reference Przybylski, Bedoin and Krifi-Papoz2013), and can be seen in the brain response to syntax violations. The P600 neural response to grammatical errors increased after regular compared to irregular primes for adults with and without dyslexia (Canette et al., Reference Canette, Fiveash and Krzonowski2020a). Rhythmic priming also reinstated the P600 response for patients with basal ganglia lesions (Kotz et al., Reference Kotz, Gunter and Wonneberger2005), who typically do not show this response (Kotz et al., Reference Kotz, Frish, von Cramon and Friederici2003). In addition to grammatical error detection, recent research has shown a rhythmic priming effect for sentence repetition in children with and without developmental language disorder (Fiveash et al., Reference Fiveash, Ladányi and Camici2023b). These results show that in the short term, listening to a regular music rhythm with a strong metrical structure can improve the processing of the less regular speech signal in typically developing children and adults, as well as individuals with neurodevelopmental language disorders.

28.4.2 Long-Term Music (Rhythm) Training

Experimentally controlled music (rhythm) training studies have shown that speech and language abilities (especially speech rhythm processing and phonological awareness) can be enhanced following music training.Footnote ³ For example, Zhao and Kuhl (Reference Zhao and Kuhl2016) showed that musical activities focusing on temporal structures (12 sessions over a four-week period) enhanced nine-month-old infants’ sensitivity to the temporal structure of speech (and music), as measured by the neural response to rhythmic structural violations. A longitudinal study over two years showed that music training resulted in increased speech segmentation skills of eight-year-old children (François et al., Reference François, Chobert, Besson and Schön2013; see also Chobert et al., Reference Chobert, François, Velay and Besson2014, for a follow-up).

Other studies have reported that groups following experimentally implemented music training programs showed enhanced phonological awareness. For example, music training for 30 minutes a week over four months enhanced phoneme-segmentation fluency in kindergarten children (Gromko, Reference Gromko2005). In four- to six-year-old children, both music and phonological training, but not sports training, improved phonological awareness (Degé and Schwarzer, Reference Degé and Schwarzer2011; Patscheke et al., Reference Patscheke, Degé and Schwarzer2016). Several other studies have shown increased phonological awareness after music training (Bolduc and Lefebvre, Reference Bolduc and Lefebvre2012; Cogo-Moreira et al., Reference Cogo-Moreira, de Ávila, Ploubidis and Mari2013; Herrera et al., Reference Herrera, Lorenzo, Defior, Fernandez-Smith and Costa-Giomi2011; Moritz et al., Reference Moritz, Yampolsky, Papadelis, Thomson and Wolf2013). Phonological awareness has also been improved by music training (focusing specifically on rhythm) in dyslexic children (Flaugnacco et al., Reference Flaugnacco, Lopez and Terribili2015; see also Bonacina et al., Reference Bonacina, Cancer, Lanzi, Lorusso and Antonietti2015). These studies have also shown that the rhythmic training improved the dyslexic children’s reading performance, an effect that has been similarly shown in typically developing children.

Music (rhythm) training has also been shown to improve literacy skills. For example, Taub and Lazarus (Reference Taub and Lazarus2012) tested adolescents in Grades 9–10 (approximately 14–15 years old) and found that an experimental group who had 12 sessions of rhythm production training improved in their reading fluency and reading performance. Rautenberg (Reference Rautenberg2015) tested children in Grade 1 (approximately seven- to eight-years-old and showed that a group with eight months of music training (emphasizing both perception and production rhythmic skills) had enhanced reading accuracy. Further, eight-year-old children who underwent six months of music training (including rhythm, melody, harmony, and timbre training) showed improved reading performance (Moreno et al., Reference Moreno, Marques and Santos2009). In summary, music (rhythm) training programs have shown promising effects on different speech and language abilities, including improving speech rhythm processing, phonological awareness, and reading skills. Interestingly, the effect of the music training programs on phonological tasks, for example, can be similarly strong as that of specifically tailored language training programs (e.g., phonological training; see Bigand and Tillmann, Reference Bigand and Tillmann2022).