Special thanks to Merriam School of Music who helped us in music reading projects!
A first study investigated how music reading is being taught through an inventory of musical signs and reading concepts found in most common piano method books. Each musical piece was scanned and with the help of software developed at the Piano Laboratory, we were able to get a complete inventory of all music notation symbols. A detailed analysis of this inventory provided a framework to classify the types of symbols and to examine the frequency of each symbol's occurrence. The results showed a wide variety from one method book to another in the number of symbols introduced, which ranged from 62 to 262 (Nassrallah & Comeau, 2010). We are now conducting a systematic content analysis of the same popular piano method books to find out how music symbols are paced, sequenced and reinforced. With this codification, it will be possible to compare how the learners' visual environment is different from one method to another.
Young children beginning piano lessons are presented with music-reading materials that contain a wide variety of colorful illustrations. However, no one seems to be concerned about the instructional value of these illustrations or question whether they affect reading. Since it has been demonstrated that there is a close relationship between where the eyes are fixated and where the cognitive attention is engaged while processing visual information, we decided to investigate where young piano students are looking when they are reading from a page in a piano method book to find out to what extent illustrations are attracting a learner’s attention. In a first pilot project, we analysed young piano students’ eye movements when previewing a piece for performance and when playing that piece at sight and we measured the length of time they inspect the illustration zones and the music zones. The results showed clear indication that students are dividing their cognitive attention between processing the musical signs on the page and processing the various colourful illustrations. Based on the results of this first project, we are presently engaged in a larger study to understand the extent to which decorative illustrations are directing the participants’ attention to evaluate the effect those illustrations might have on the development of music-reading skills. We need empirical evidence to clarify the real impact of decorative illustrations on music-reading outcomes.
Another study addressed early music-reading skill acquisition and the possible impact of two different music-reading teaching approaches found in method books, Middle-C and Intervallic. Using cognitive modeling, we observed through computer simulation the problem- solving and decision-making tasks involved in decoding a simple musical score. Inspection of the simulation results revealed differences in terms of cognitive processing demands. In particular, the Intervallic method required a larger amount of declarative knowledge related to notes and more execution planning than the Middle-C method (Emond & Comeau, 2013).
In order to compare three common methods of sight-reading assessment, the well-known Watkins-Farnum Performance Scale for assessing wind instrumentalists was adapted for pianists and then compared with two other assessment methods, a scoring algorithm and expert examiners. It was found that these methods, regularly used in music-reading research, differed greatly in their assessment results. Therefore these assessment methods used to rank research participants are not comparable and should be used with great caution when conducting research on music reading (Lemay, 2008).
Considering the variety of tests used to evaluate music-reading performances, we have undertaken a classification of the existing measurement methods. By providing a similar and consistent body of information on each one, this compilation and classification will provide direction and stability for future research (Jensen, 2016).
Having demonstrated the inconsistencies between the different assessment methods and the limitations and drawbacks of the various approaches to music-reading assessment, we knew that no reliable tool existed to measure music-reading performances by pianists. So, the Piano Lab undertook the development of a test to measure music-reading skills. Using original musical stimuli of increasing difficulty, a system for codifying errors and a scoring grid for evaluating music-reading performances of beginning to advanced-level pianists, this assessment tool is presently being tested. It will be the first measuring instrument specifically designed to evaluate a pianist’s reading skills and will become a valuable tool as reading assessment is essential for measuring teaching strategies or developing research protocols.
To understand the eye-movement patterns of piano students reading a musical score, a series of experiments were undertaken using eye-tracking. The first project was designed to study the perceptual span of university piano students during a sight-playing activity. A moving window ensures that the reader sees only the section of the musical score where the eyes are fixating. The player must move the eyes to see the following notes. Eye movements were recorded and analysed, and we found that contrary to popular belief, sight readers benefit from a similar field of vision, no matter their level of competence. In order words, expert sight readers do not use a wider perceptual span to be successful in the reading task (Liu, Comeau, & Balasubramaniam, 2013).
Based on the assumption of a probable link between the way the brain processes language syntax and music syntax, we studied the processing of musical and linguistic syntactic incongruities. Participants would read aloud phrases whose syntax is either correct or deliberately incongruous and would play on the piano musical sequences that were also correct or incongruous. The eye tracker made it possible to study the effects of incongruous language and music syntax on eye movements and on reading time. It was found that syntactic incongruities in both domains were associated with an increase in the duration of fixations in the region of interest. The results are consistent with the growing evidence of a shared network of neural structures for syntactic processing in music and in language, while not ruling out the possibility of independent networks for each domain (Ahken, Comeau, Hébert, & Balasubramaniam, 2012).
Future project: We aim to develop a theoretical model that would explain the development of the processes involved in the acquisition of music reading. Such a model would deepen our understanding of the precursors of musical literacy, the factors and skills necessary to succeed, the growth trajectory, and the consequences of different types of teaching approaches. This study on music reading will offer a unique perspective on the acquisition of a writing system; clarifying the learning process required to master a skill learned sequentially and progressively will illuminate the important components for a unified theory of the acquisition of complex skills in different learning contexts.