Abstract
A substantial proportion of children underachieve in reading (Brooks & Tough, 2006; Doubek & Cooper, 2007; Tymms & Merrill, 2007). This paper builds on the existing evidence base for the link between reading and rhythm, presenting findings of an investigation into a rhythm-based music intervention. Following participation in the intervention, an entrainment strategy, significant gains occurred in the children’s reading comprehension, reading accuracy and reading rate. An analysis of reading fluency revealed significant gains in the prosodic features of reading behaviour, highlighting syllable division, grammatical structures and phrase contours, and explicating the children’s deeper engagement with connected text. Rhythm-based training in music emphasized group interaction and the children themselves also reported positive changes in their sense of well-being.
Introduction
A growing body of evidence supports the argument that some children experience a difficulty with processing rhythm and that this impacts deleteriously on reading behaviour (Anvari, Trainor, Woodside, & Levy, 2002; Corriveau & Goswami, 2009; Douglas & Willats, 1994; Overy, 2000, 2008; Tallal & Gaab, 2006; Wolff, 2002; Wolff, Michel, Ovrut, & Drake, 1990). For example, scholars have found individual differences in children’s ability to process rhythms at 2.5 years of age, particularly in their ability to keep time with a musical beat (Kirschner & Tomasello, 2009). In a study of young children, researchers found that their ability to cope with patterns and rhythmic complexity typically increases between 5 and 7 years of age (Davison & Colley, 1987), whereupon they reproduce one-bar rhythms to the level of an adult non-musician (Drake, 1993).
Children’s drawings of rhythmic figures have helped researchers to study the way in which their representations of metrical accents and time values corresponded with their rhythmic tapping abilities (Smith, Cuddy, & Upitis, 1994; Upitis, 1987). Studies of young children’s compositions and improvisations showed that undeveloped repetitive rhythmic patterns were typical and that the ability to keep a steady pulse increased with age (Kratus, 1985; Swanwick & Tillman, 1986; Brophy, 2002). According to Paananen (2006) there is wide variation in the development of rhythmic improvisation among children. From 8 years, children developed the ability to synchronize with a musical beat and to organize different note lengths within a metrical structure. When grouping and metre conflicted, younger children were unable to process rhythmic anomalies, but from 8 years, older children understood this level of complexity if they improvised using two or three time values, implying a hierarchical organization of time in the mind of the child. Indeed, ethnographic studies of musical play in children from many different cultures have shown that children as young as 7 years use polymetric relationships in clapping against the duple metre of the text and have suggested that a tendency exists for music educators to over-simplify materials for classroom use (Marsh & Young, 2006).
Furthermore, children place value and purpose on learning through movement. They participate in communal ownership of repertoire and experience social synchrony (Turino, 2008) by sharing a ‘strong sense of groove’ (Harwood & Marsh, 2012, p. 326) and by learning musically challenging games of skill. Observing that some children are more adept than others in terms of competence in executing movement patterns, Marsh and Young (2006) noted that in the playground, children who had better coordination tutored their friends and younger children, who struggled with the patterns. Historically too, music educators have observed varying levels of rhythmic organization in children. According to Jaques-Dalcroze (1966 [1905]), their individual differences in reproducing rhythms and in measuring time related to an irregular gait which could be corrected through rhythmical training. He developed eurhythmics, a systematic approach for regulating the natural rhythms of the body through smooth, uninterrupted and flowing sounds (see Bachmann, 1993). In explaining his methods, Jaques-Dalcroze stated that through continuous repetition of action, responses would pass beyond the conscious control of the brain and become automatic. He identified two fundamental problems in the perception and production of rhythm: arrhythmia (inattention to rhythm) and errhyhma (misapprehension of rhythm).
Jaques-Dalcroze categorized the various difficulties that his young pupils experienced. Some displayed very good hearing and rhythmic sense but were poor readers and struggled to analyse what they had heard. He identified a second group as those who perceived rhythms accurately but for whom reproducing the rhythms in the body or lower limbs or combinations of limbs was difficult. He found a third group who were slow in responding and reacting to changes in the speed, shading or shaping of a rhythm. Critics of his approach found too much emphasis on rhythmic elements and a neglect of an embodied response to the melodic features of music (e.g. Repp, 1993). However, the types of rhythmic weakness described by Jaques-Dalcroze have also been identified among children with reading difficulties (Atterbury, 1985).
Scholars have identified a general ‘pan-sensory’ timing difficulty in association with poor literacy, a weakness in coordination and efforts to learn new skills (Denckla & Rudel, 1976; Lovegrove, Garzia, & Nicholson, 1990; Nicolson, Fawcett, & Dean, 1995; Wolff et al., 1990). Visual impairments include slower than average rapid naming of letters, digits, colours and objects (Denckla & Rudel, 1976). Movement coordination appears to be impaired by slower than average responses to finger tapping tests (Tallal, Miller, & Fitch, 1993; Wolff et al., 1990). An acoustic processing delay in both identifying the sounds of language (phonemic awareness) and in tapping to a metronome has been found in undergraduates with dyslexia (Pasquini, Corriveau, & Goswami, 2007). These findings support Wolff’s assertion that the precise serial ordering of features is associated with timing processes that underlie communicative speech, music and skilled motor performance. Wolff has found that dyslexic children anticipate metronome clicks more slowly, less precisely, and are slower in adapting to abrupt changes in the rate of metronome clicks (Wolff, 2002).
Variation in attainment in children’s spelling and reading is found to be related to performance on a test of rhythmic discrimination (David, Wade-Woolley, Kirby, & Smithrim, 2007; Douglas & Willats, 1994). These scholars argue that an ability to discriminate the stress patterns of language helps children recognize multi-syllabic words. According to Lerdahl (2001), features of grouping, metre, duration, pitch contour and timbre (the character of a sound as distinct from its pitch and strength) are shared by both language and music and have been recognized in the musical elements of poetry, and are known as prosody. Scholars have measured the rise and falls of pitch (intonation), rhythm, stress (emphasis and graded volume) and pauses that emphasize meaning and expressive effect (Benjamin & Schwanenflugel, 2010; Herrera, Lorenzo, Defior, Fernandez-Smith, & Costa-Giomi, 2011; Hirschberg, 2002; Kuhn, Schwanenflugel, Meisinger, Levy, & Rasinski, 2010).
Sensitivity to complex features of stress in speech develops in infancy (Jusczyk, 1999). However, children vary in their ability to interpret meaningful and expressive features of speech (Wood & Terrell, 1998). Scholars have found significant improvements in 6-year-old children’s ability to interpret prosodic cues in natural speech when they compared groups of children trained in either music or in drama with an untrained group: both groups of trained children outperformed the untrained children. The musically trained children performed at the same level as the children trained in drama, indicating that training in music improved children’s interpretation of the musical (prosodic) aspects of speech. During musical training, the children had received no specific training to enhance their perception of prosody. Therefore, this study provided evidence of a transfer effect between music and language and suggested links between cognitive processing mechanisms for music and language (Thompson, Schellenberg, & Husain, 2004). In addition, positive effects of musical training on verbal memory have been described (Ho, Cheung, & Chan, 2003; Jakobson, Cuddy, & Kilgour, 2003).
However, the demographic profile of the sample tends to be omitted in quasi-experimental studies of musical training such as those described here. Scholars have revealed that in the United States, music programmes consistently attract academic achievers and children from privileged family backgrounds; in particular, scholars have described an abnormally high incidence of families in which both parents have degrees (Costa-Giomi, 2012; Duke et al., 1997; Elpus & Abril, 2011; Stewart, 1991). According to Costa-Giomi, it is possible that non-musical benefits of musical training, particularly in long-term studies, may be explained by demographic features and personality characteristics.
Nonetheless, the effects of training in music may have a positive effect on the early reading skills of young children as demonstrated by Anvari et al. (2002). The tests included measuring their sensitivity to the sounds of language (phonemic awareness), which are known to underpin children’s reading development (Bradley & Bryant, 1983). Tests of pitch and rhythm discrimination were used to determine musical perception. Children’s sensitivity to pitch at 5 years predicted their performance in reading letters and single words. Therefore, the scholars concluded that increasing sensitivity to the sounds of the language through music may improve phonological awareness, which is thought to enhance young children’s readiness for reading. These outcomes confirmed those of previous investigations (Barwick, Valentine, West, & Wilding, 1989; Lamb & Gregory, 1993), where higher scores in musical pitch sensitivity corresponded with higher scores in reading abilities and phonemic awareness.
More recently, Besson, Schon, Moreno, Santos, and Magne (2007) compared language and music processing skills of a group of children receiving musical training through the Kodály method with a group that received training in painting. The Kodály method draws on the work of Jaques-Dalcroze; it incorporates walking, running, skipping, clapping and marching in the early stages of the programme, alongside tuition in melodic musical concepts involving singing, listening and notating music (Choksy, 1999). The researchers recorded young children’s reaction times and errors on tasks related to expressive features in speech; they also recorded changes in the children’s brain electrical activity. Their findings showed that after six months of music training using the Kodály method, the children displayed reductions in error rates and enhanced sensitivity to pitch contours in speech compared with children who had been intensively trained in painting. This study supported their previous findings (Moreno & Besson, 2006; Schon, Magne, & Besson, 2004).
Hurwitz, Wolff, Bortnick, and Kokas (1975) predicted that the initial stage of Kodály training, designed for young children, would influence their performance on sequencing tasks, spatial functioning and reading behaviour. Their findings revealed significant gains in performance on the Stroop test (a test of ability to inhibit inappropriate response) and in reading comprehension for boys, but not for girls. In further investigations of the Kodály method, scholars also found significant gains in mathematical abilities (Neufeld, 1986; Staines, 1999), and gains in both mathematical and reading attainment for girls and boys following Kodály training (Gardiner, Fox, Knowles, & Jeffrey, 1996). The present study explores key features of the Dalcroze-inspired elements of Kodály training, designed to strengthen children’s alignment with metre. This study explored whether training children to perform increasingly complex tasks while ‘keeping time’ to a musical accompaniment would have a positive effect on the temporal organization of their reading skills.
Method
Recruitment and sample
The study was located in a school for children aged 9–13 years, in a large English town. The children, in their first term at middle school, had been educated in either very small rural lower schools 1 located in neighbouring villages or larger lower schools in the outskirts of the town. As a group, these children spanned a fairly broad range of prior educational experiences.
One notable demographic feature of this school was that a large number of children, 45.5 per cent, were eligible for free school meals. 2 In terms of year-on-year attainment, the pupils of this school performed less well than did pupils of neighbouring schools and less well when compared with pupils across the UK. The 15 children recruited for this study were 9 and 10 years of age. The school had identified them as weak readers and taught them together as a low-ability group. According to school staff, the school’s rationale for streaming was that it allowed pupils to work at their own pace so that they would feel more comfortable, but the school did not specifically support these pupils through a targeted intervention. School inspectors had recently judged systems for raising academic attainment in the school to be unsatisfactory and the rhythm-based intervention was delivered by school staff specifically to demonstrate that these pupils were appropriately supported. The pupils took part in rhythm-based musical training between November and December, allowing them time to settle into their new school before conducting baseline testing.
Design
The rationale for the design of this study was based on the fact that it followed a randomized-controlled investigation of the rhythm-based intervention (Long & Hallam, 2012). In the earlier study, randomization was carried out on a pair basis and pupils were matched for reading ability. When compared with controls, the findings recorded significant gains in reading comprehension for the intervention group (Long & Hallam, 2012). It was not feasible to repeat the two-group design as a randomized controlled trial. The extent of the individual differences between pupils in this type of investigation would necessitate a very large sample size, and that was beyond the scope of this study. The present study, however, aimed to trial the efficacy of the intervention when delivered to a whole class by their usual teacher, the rationale being to gauge whether the efficacy of the intervention was sustained when delivered by school staff. Strengthening the ecological validity of the approach was a priority in this study. However, as a single group study it was not possible to control for variables that are typical of everyday life in schools, such as individual differences and experiences of the pupils, and this limited the interpretation of the findings.
In preserving the natural setting of the study, the aim was to investigate the efficacy of the intervention in a typical classroom. Therefore, the investigation aimed to accommodate a whole class approach and single-group repeated measures design was judged most appropriate under the circumstances. The intervention consisting of 10 minutes of rhythm-based exercises was administered at the start of the children’s usual weekly curriculum music lessons. The children’s school music teacher 3 had been trained to deliver the intervention in two sessions of 30 minutes with an interim period of one week, which provided the teacher with an opportunity to assimilate the exercises and then to ask any questions in the second session. The children were tested on their reading before and after the six-week intervention period.
The rhythm-based music intervention
The rhythm-based exercises were modelled for the children in three stages (see Table 1(i)). The teacher modelled keeping time with the beat of a tuneful musical accompaniment. The intervention required mental anticipation and inhibitory control in order to lift one foot while striking the other against the floor in synchrony with the strong beat of the musical accompaniment and the actions of the other children in the room. This entrainment activity applies theoretical modelling of (i) normal anti-synchrony, as ‘one foot comes up the other goes down’ (Clayton, Sager, & Will, 2004, p. 9), and (ii) a stable hierarchical distribution of cognitive attention (see Clayton, Sager, & Will, 2004; Jones, 1976; Jones & Boltz, 1989; Jones & Yee, 1993).
A rhythm-based music intervention: A three stage method with nine steps to progression.
Cognitively, the children were required to plan ahead, synchronize, monitor and integrate multi-level physical coordination, which in turn required anticipatory and inhibitory control, and to keep time with the musical accompaniment and the other children in the room. Lastly, the teacher modelled reading simple staff notation, chanting on a monotone the alphabet letter names of music notation. Chanting was synchronized in time with stamping and clapping actions (Table 1(vii)). During the training, in addition to reading pitch, the durations were occasionally varied to include: one beat and two beat notes, and also one beat and two beat rests.
Measuring reading behaviour
A standardized and commercially available measure of reading was used to assess the impact of the rhythm-based exercise on reading scores. The Neale analysis of reading ability (NARA) (Neale, 1989) measures oral reading ability and comprehension of the text. This reading test is a more sensitive and ecologically valid tool than single word reading tests or forced choice reading comprehension tests. It measures integration of connected text with other context driven features of reading behaviour such as vocabulary, schematic and general knowledge.
Measuring reading accuracy involved recording the number of omitted, added, mispronounced, reversed or substituted words in the reading. Reading comprehension was measured as the number of correct responses to questions, posed immediately after the child had read the passage of text. Literal responses required the child to simply retrieve information directly from the text, whereas inferential responses required judgement and application of general knowledge in the context of the passage.
The reading comprehension scores were recorded as a percentage of correct comprehension responses given. Rate of reading scores were recorded as words per minute (wpm). Reading accuracy raw scores were calculated by deducting the number of incorrect scores from a maximum score as directed in the manual. The raw scores for reading accuracy were then converted to a percentage. This form of testing formed an adequate basis for before and after comparisons to be made. The manual allows for retesting after six weeks.
There are a number of definitions of reading fluency. Benjamin and Schwanenflugel (2010) recommended that studies of reading fluency should examine prosody and chunking of text into grammatical (syntactic) units, and that passages of connected text should be used rather than single words or sentences. In this investigation prosodic components of reading behaviour were recorded at sentence by sentence level: (i) the sustained presence of a rhythmic foot for the duration of a sentence as a result of strong and weak stress and alternating long and short vowel lengths at the syllable level; (ii) the sustained presence of expressive features allocated to key syllables, indicating that the child had detected the grammatical organization (the subject and predicate) of the sentence; (iii) the sustained presence of a pitch contour across a segment of the passage (e.g. a phrase, clause, question or sentence). Measuring these features allowed for detection of temporal organization as a hierarchical structure in children’s reading of passages of connected text.
In scoring the NARA the pupils’ responses were spoken rather than written. Although the NARA was scored by a researcher with expertise in this particular reading assessment, which facilitated accurate and reliable scoring, a data audit was required as a validity check. The auditor listened to selections from all of the reading samples and checked that the scoring was reliable by followed the scoring procedures in the manual. For example, for the purposes of this study, the comprehension responses were required to match those listed in the manual to score a point. In this respect any subjectivity was removed from NARA scoring. Similarly rigorous scrutiny was applied to the analysis of reading fluency. In order to achieve a point on an item in this measure of reading, the feature was sustained along the entire length of the sentence. Partial or majority success in a single sentence was scored as zero. This stringency determined that scoring was objective and produced reliable categorical data. Scoring was given as a percentage based on the number of sentences read fluently divided by the total number of sentences read in the passage. Age norms were not available for this measure.
Results
The data were entered by hand into a statistical software package (SPSS). A within groups one-way ANOVA revealed significant gains and a large effect size in reading comprehension (F (1, 14) = 48.124, p < .0001), significant gains and a medium effect size in reading accuracy (F (1, 14) = 7.02, p = .019), and significant gains and a large effect size in rate of reading (F (1,14)= 31.821, p < .0001) (see Table 2).
Mean scores and standard deviations for reading comprehension, reading accuracy and rate of reading.
NB: mean accuracy and comprehension scores are given as a percentage of the total possible score for the passages that were read; mean rate of reading is given as words per minute.
The mean scores for each child were converted to reading ages according to the instructions in the NARA (Neale, 1989) manual. The changes in reading age are set alongside the children’s perceptions of the rhythm-based music intervention and their perceptions of reading (see Table 3). The children were asked, ‘What did you think of the stamping exercise?’ and ‘Have you noticed any change in your reading?’ Ten of the children were aware that their reading had become ‘easier’ or ‘better’, five of the children referred to being able to ‘understand’ more of the text.
Standardized reading ages set alongside children’s perceptions of the rhythm-based training and their reading behaviour.
To investigate whether the improvements in reading scores were matched by improvements in features of reading fluency at sentence level, a one way within groups ANOVA was conducted and this revealed statistically significant gains and large effect sizes in the sustained presence of (i) a regular rhythmic foot (F (1, 14) = 49.91, p <.0001), (ii) parsing of subject and predicate (F (1, 14) = 34.325, p < .0001), and (iii) pitch contour (F (1, 14) =10 .99, p = .005) (see Table 4).
Mean scores and standard deviations for three features of prosody in reading:(i) a sustained regular rhythmic foot, (ii) consistent allocation of prosodic features to reflect functional grammar and (iii) sustained pitch (intonation) contours within a sentence.
NB: mean scores show the number of read sentences that sustained each prosodic feature as a percentage of total sentences read.
Discussion
These findings showed that statistically significant changes occurred in children’s reading of connected text after participation in a rhythm-based intervention. After six training sessions of 10 minutes, the children demonstrated substantial gains in reading behaviour with large effect sizes in reading comprehension and rate, and a medium effect size in reading accuracy.
The children’s reading ages improved dramatically. Four of the 15 children were reading at below chronological age level on one or more sub-tests of the NARA before the intervention, but after training in music, read at their chronological age level or higher across all three sub-tests. Eight of the children made gains ranging from four months to 33 months in one or more sub-tests. The pupils in this group had been identified as learning more slowly than other pupils in the cohort. Their progress after six weeks of training in music suggests that the lowest attaining pupils would continue to improve and catch up with peers (of the same age) if they had access to additional rhythm-based music intervention sessions. Two children in the group read at or above their chronological age level in all of the baseline sub-tests. They too made considerable progress during the intervention period, indicating that this method is suitable for use in mixed ability group teaching. One child had avoided participating in the intervention; he read very quickly during the post-tests and opted not to answer most of the comprehension questions. His scores were included in the analysis to safeguard the ecological validity of the study and to reflect issues that arise in a typical classroom setting.
We can be confident that these results are not due to practice effects. The manual states that retesting after six weeks is permissible. In earlier investigations which involved control groups, where practice effects would be expected to occur equally, gains in NARA subtests were very modest for the control group children. Practice effects had clearly not occurred (Long, 2007; Long & Hallam, 2012). In the present study, the children who took part had been assigned to the low ability learning group and it is highly unlikely that children identified as low ability would have the capacity to retain several passages of text and comprehension questions over a six week period. Further, the most able readers, achieving chronological age scores on the baseline tests, made relatively modest gains when compared with weaker readers. Arguably, if these results were a product of practice effects, the more confident readers would be expected to display greater gains than their less able classmates. Finally, the children were not told whether or not their baseline responses to the comprehension questions were accurate, and for all these reasons the use of the same reading passages for the baseline tests and post-intervention tests were unlikely to have threatened the experimental design.
The findings of the present study are consistent with the gains in reading behaviour described by Long and Hallam (2012). The present study indicates that the rhythm-based music intervention is effective when delivered in a typical classroom setting. The focus on high levels of ecological validity in the present study meant that pupils were not randomly assigned to different conditions. For this reason, there was no control group in this study. Therefore, we cannot completely attribute the change in reading scores to the pupils’ participation in the rhythm-based exercises. The changes in reading scores might be alternatively explained by (i) a ‘catch-up’ intervention in the school during this period, (ii) the new curriculum at the new school may have impacted positively on the students, or (iii) a Hawthorn Effect which is the possible effect of ‘being studied’ on behaviour.
Firstly, according to UK Government statistics, pupils at this school persistently underperformed in relation to pupils in neighbouring comparator schools. The study took place during the period in which the school had been placed under Special Measures 4 by school inspectors as it was a failing school. The inspection report, dated three months after the intervention took place, stated that tracking data was not used by teachers to inform lesson planning, which indicated that the pupils had not received a targeted intervention as part of their curriculum. Further, the school music teacher confirmed that this group had been separated from other pupils so that they were able to work at their own pace but they had not received additional support. Secondly, the transition points in the National Curriculum for England are between Key Stage 1 and Key Stage 2 (at 7 years of age) and between Key Stage 2 and Key Stage 3 (at age 11 years). Given that the children in this study transferred to their new school at age 9, it is highly unlikely that the gains in reading scores occurred because of a change in the curriculum. Thirdly, steps were taken to strengthen the ecological validity of the present study; these also minimized the occurrence of a Hawthorn Effect. Apart from the data collection, in which individual children read aloud for 20 minutes, they had no contact with research staff. The intervention took place during the children’s usual class music lesson, and the children stood behind their usual chairs to perform the rhythm-based exercises. There were no researchers present during the class music lessons.
The interpretation of the findings must be made cautiously because this exploratory small-scale study was limited both by its scope and single group design. A lack of comparator data permits little more than speculation on the changes in the children’s reading. Nonetheless, the aim of the study was to investigate whether the gains made in a previous investigation delivered by a researcher might be sustained when the same intervention was delivered by school staff. The findings appear to suggest that school staff delivered the intervention effectively.
Further, consideration of these findings in relation to previous research might indicate that training in rhythm-based aspects of music and reading of connected text are linked and one explanation for this may be that particular cognitive processes overlap in both musical and reading activities (Moreno et al., 2011; Wolff, 2002). Wolff reasoned that anticipation is a fundamental attribute of skilled motor action, language processing and cognition, and that unstable or abnormal anticipation times would deleteriously impact on coordinated behaviour patterns that depend on precise timing and serial ordering of their elements such as communicative speech and the reading of connected text. In deconstructing the impact of the intervention, it is appropriate to discuss anticipation times and inhibitory control. However, the measurement of these attributes lay beyond the scope of the ecological and naturalistic approach of this study.
In an everyday sense, the development of children’s anticipation and inhibitory control can be observed by teachers and parents in the initial stages of acquiring new procedural knowledge. For example, anticipation and inhibitory control are evident when children learn to cross a road safely, and also when children recreate or invent playground games that involve metrical stepping, chanting or skipping as described by Gaunt (2006). Marsh observed that the children preferred a constant state of self-imposed challenge, eventually achieved by combining a duple metre chant at syllable-level with choreographed movements, which were often in a faster triple pattern to create a ‘polymetric relationship’ (Marsh, 2008, p. 306). Some of the observed children who focused on the words first, added the movements later when they could participate fully thus: ‘synchronised, precise, sharp and complex rhythmical body movements combined with singing, chanting and verbal play’ (Malone, 1996, p. 188, cf. Marsh, 2008, p. 296). In particular, Marsh’s study captured the prolific and rapid mechanisms of transmission of new clapping games between large numbers of children through proximity, physical contact and membership of friendship groups. The children described learning new games by attentive observing and allowing their bodies to recognize and remember the movement formulae from other games that they had already mastered. In this sense they were using a form of kinaesthetic memory and the movements may have operated as a type of mnemonic device (Marsh, 2008, p. 143). They even used the word ‘glide’ to describe the process of picking up and participating in the rhythm of a clapping game (Marsh, 2008, p. 141). Children appeared to use clapping games to manage the tension and frustration caused by boredom; in other words, they were perhaps self-regulating their emotional experiences. They tended to spontaneously play these games in situations when their personal autonomy was constrained, thus avoiding boredom during in-between times, when freedom of movement and time were in limited supply: between lessons, waiting in line and on bus journeys (Marsh, 2008).
According to the social-constructivist perspective on learning, the notion of a cultural tool offers children an opportunity to progress from a shared experience (inter-mental) to an individual (intra-mental) internalized learning experience (Vygotsky, 1962). The effectiveness of the rhythm-based music intervention in a typical classroom setting illustrates this. Although two pupils strongly resisted the weekly opportunity to participate, many progressed from a shared (entrained) response to an individual (internalized) response by the end of the six-week period. These pupils used the words ‘fun’, ‘easier’, ‘happier’ and ‘more together with your body’ to describe their experience of the intervention. The children’s self-perceptions of their experience suggested that (i) conscious awareness of weak inhibitory control was superseded by increased consciousness of anticipation, and (ii) conscious awareness of weak coordination was superseded by increased consciousness of relaxation to form an ‘anticipate/relax cycle’ (see Table 1, rows (b) and (c)). In short, this means that pupils’ sense of temporal organization became increasingly poised, natural and relaxed.
After six weeks of music training, the prosodic aspects of the children’s reading behaviour had shifted. Instead of reading syllables with equal prominence, the children allocated stress appropriately, reflecting the stress-timed syllabic pattern of spoken English. The proportion of sentences featuring stress-timed syllables increased from 47 per cent to 88 per cent of total sentences read. Also, the children improved in their ability to allocate prosodic features to show that they had detected the grammatical functions of words. Awareness of the subject and predicate was demonstrated in 66 per cent of the total number of sentences read, compared with 38 per cent in baseline testing. Most children were better able to sustain an intonation contour across a sentence, as featured in 18 per cent of sentences before training, and in 44 per cent of sentences after training. Thus, the children had automatically integrated word recognition with syntactic knowledge and segmented the passage of text into meaningful units, demonstrating the coordination of the sub-skills of fluent reading.
It is possible that rhythm-based training may have influenced the children’s reading ability in three ways. First, it may have facilitated the natural tendency in English to apply a stress-timed pattern. Second, it may have generated self-sustaining momentum. An optimum level of momentum sustains the co-ordination of various sub-components of reading, enabling readers to construct meaning from the text (National Institute of Child Health and Human Development, 2000). This goes some way to explaining why 13 of the 15 children improved in their rate of reading scores and why two children did not. Although, pupil 12 did not show an improvement in rate of reading, his baseline reading comprehension age indicated that he was not under-performing in reading comprehension prior to the intervention and so it is not surprising that his score remained the same after six weeks. On the other hand, pupil 14 showed an increase in rate and accuracy of reading. The decrease in reading comprehension coupled with uneven baseline scores suggested that an underlying specific issue might account for the difference in her response compared with the majority of the group. Third, assimilating the rhythm-based training may have helped the children to track the direction, length and shape of the phrases embedded within the melodic and harmonic structure of the musical accompaniment.
The improvements in reading behaviour might be explained as having an activating effect of sharpened definition of phonemes, in terms of their prosodic attributes and their relative positions in adjacent syllables and non-adjacent syllables, on sub-cortical automatic language processing mechanisms. These findings develop further the notion that a link exists between children’s processing of music and language, involving a sharpening of awareness of sensitivity to types of sound (Anvari et al., 2002; Overy, 2003; Tallal & Gaab, 2006). Before training, many children applied stress to every syllable that they read, but after training, the edges of syllables acquired finely graded definition, and became more reminiscent of natural speech, facilitating assignment of syllables to grammatical function. These findings concur with previous studies, which found that children with good comprehension skills generally displayed exaggerated prosody (see Benjamin & Schwanenflugel, 2010).
The musical accompaniment was internalized as an integral feature of the training. The particularly regular organization and hierarchical structure of the musical accompaniment may have strengthened predictive language processing mechanisms and also supported phrase-level prosody in reading, which in turn would strengthen reading comprehension skills. The findings recorded gains of three months or more in 12 of the 15 children, six of whom had improved by 12 months or more, and this is consistent with the finding that phrase-level prosody predicted a significant amount of variance in reading comprehension after word reading accuracy, phonological awareness and general rhythmic sensitivity had been accounted for (Whalley & Hansen, 2006). It is suggested that predictive language mechanisms may also account for the reduced error rate in reading accuracy: 12 of the 15 children made gains in reading accuracy age of three months or more, five of whom improved by 12 months or more.
Concluding thoughts
This study aimed to investigate the impact of a rhythm-based music intervention on reading behaviour. It followed a randomized-controlled study in which findings showed significant gains in reading comprehension for the rhythm-based music intervention group when compared with controls (Long & Hallam, 2012). The present study considered the pupil’s self perceptions of their experiences of rhythm-based music intervention alongside quantitative analysis of their reading fluency, accuracy and comprehension. In addition to recording significant gains in pupils’ reading behaviour, the findings implied that the rhythm-based music intervention had strengthened pupils’ inhibitory control and anticipation in an everyday sense. This was hardly surprising because the rhythm-based intervention is an entrainment strategy, which involved pupils moving and reading in time with background music.
Inhibitory control, as one of three major subcomponents of executive function in the brain, allows behaviour to become progressively more strategic, accurate and organized. In terms of reading, the majority of the children in this study were able to rapidly shift from slow sequential word recognition to the retrieval of larger units, having resolved weak anticipatory and/or inhibitory control through rhythm-based training. The medium and large effect sizes of the statistically significant findings from such a small sample suggested that rhythm-based training may have enabled struggling readers to quickly improve. However, these findings should be viewed with caution because this exploratory small scale study was limited firstly by its scope and secondly by its lack of comparator data. For these reasons, it is not appropriate to attribute causality to the intervention, not is it appropriate to generalize widely from these findings. In the future, longitudinal research should be carried out to measure the extent to which the changes in reading behaviour are sustained over time.
Footnotes
Funding
This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.