Supporting Piano Education
Article written by Andrew Eales
The topic of singing in aural tests has long been a contentious one. Not only have growing numbers of teachers noted how unpopular the singing tests are, but concerns have increasingly been raised about access and inclusion.
In recent years, scientific research has also cast significant doubt on the long-held assumption that singing provides a reliable measure of a candidate’s underlying pitch perception, raising further questions of validity.
As we consider these points, we must be careful not to conflate the importance of promoting singing in music education with an insistence on assessing a pianist’s singing in a piano exam. These are two quite different issues. Piano lessons should certainly not be limited to including only exam syllabus requirements. Nor should singing be marginalised to aural test preparation.
The central question is not therefore whether singing has educational value. It plainly does. Rather, it is whether a singing task provides a reliable, valid, and inclusive measure of the listening skills that aural tests are intended to assess. And if not, what are the alternatives?
Recognising there is a problem
Recent thinking in education has increasingly been influenced by the principles of Universal Design for Learning (UDL). This encourages educators to design curriculum, syllabus, and assessment with learner variability in mind from the outset, rather than making individual adjustments only after barriers to inclusion have been identified.
UDL is neither an unassailable theory, nor without its problems. However, it commendably encourages a more proactive rather than a reactive approach to inclusion. From this perspective, an aural assessment should let all candidates demonstrate their listening skills without adding extra barriers, such as singing.
Some candidates may find accurate singing particularly challenging for reasons that are unrelated to their aural perception, for example:
- older adults with increasingly frail voices
- teenage boys with changing voices
- those with a vocal-motor control impairment
- some with disabilities or muscular problems
- those suffering with chronic or acute respiratory conditions
- those with certain allergies, asthma, hay fever, etc
- those prone to situational anxiety
- some neurodiverse learners
Few would argue that any of these groups are inherently unable to perceive pitch simply because they lack confidence or accuracy when singing under pressure in an exam room.
It’s hardly surprising that many dislike singing tests, and examples seemingly abound of pianists who switch to using different exam boards to avoid them. Probing this, they seem not to be seeking an easier path, but simply one which is more inclusive and relevant to their piano journey.
Anecdotally, I have heard distressing stories of teenagers crying and even vomiting, overwhelmed by situational anxiety. I have personally witnessed a few suffer adverse reactions when asked to sing in aural tests.
Examiners perhaps rarely see this for themselves, because such learners will often avoid making the trip to the exam room, or else choose a board more attentive to their access requirements.
A singing test?
The point is often made that examiners are not assessing the quality of singing or vocal technique during aural tests, but merely listening for basic accuracy of pitch and rhythm. This reassurance might help put some at ease, but to what extent is it credible? Let’s consider an example.
In ABRSM Grade 4 Test B, the candidate must sight-sing a sequence of five pitches, written as semibreves. Introducing the test, the examiner instructs the candidate as follows:
“Please sing the notes at number… on this page. Sing them slowly, and I’ll help by giving you the right note if you sing a wrong one. Here is the key-chord [name and play] and this is your starting note [name and play]… thank you.”
How exactly can vocal production be separated from this assessment, when the candidate’s physical ability to reproduce pitch through the voice is part of the task? A basic understanding informs us that pitching notes with confidence and secure intonation requires control of the voice as an instrument.
Surely a candidate who sings the notes accurately, in tune and in time, should score more highly in this test than one who doesn’t. We would all rightly be concerned were that not the case. But marks lost here could reduce an overall Distinction to a Merit, or a Pass to a Fail. To suggest that the candidate’s singing ability does not impact their final result seems disingenuous.
At this point, a colleague will usually chip in with helpful suggestions about how I could teach my piano students to sing more confidently.
But this overlooks an important fact: they come to me primarily to develop their piano playing, not to learn voice as an assessed instrument. Singing to a stranger in an exam room simply isn’t something most anticipate, or a musical goal many pianists have.
Sight-singing and ‘audiation’
So why are singing tests included in piano exams? The main reason for aural tests is that they help evaluate listening skills and musical perception, like pitch, pulse, and rhythm. But do sung responses truly show musical listening skills and perception, or as Edwin Gordon calls it, ‘audiation’?
Returning to our previous ABRSM Grade 4 aural test example, the point here would seem to be to determine whether a candidate can, from the notation, form an aural perception of how it should sound.
Singing may seem like an obvious way to evaluate this, were it not for the fact that introducing vocal techniques materially alters the test by adding a whole different skill set, requiring physical abilities that many people are not proficient and comfortable with.
In any case, the relationship between singing and audiation has in recent years become a topic of scientific research, and we now have concrete answers to the complex and intriguing question of whether we can trust singing as an accurate way of assessing pitch perception.
What does the research tell us?
According to a paper published in 2012 by the International Laboratory for Brain, Music and Sound Research in Montréal, an inability to sing the correct pitch is common, but this is rarely due to a perceptual (aural) deficit.
Sean Hutchins and Isabelle Peretz established that in tests, 95% were able to discriminate and match pitch accurately using a pitch slider. But crucially, many of them proved unable to match the pitch by singing or humming it back. Their scientific findings offer empirical evidence that raises serious questions about the validity of using singing as evidence of audiation.
Digging deeper into their research, Hutchins and Peretz found that:
- 20% of participants displayed evidence of a vocal-motor control impairment that prevented them from singing the correct pitch
- 35% had a sensori-motor (timbre) deficit, meaning that they had difficulty matching vocal pitch to notes played on a piano
In a further piece published in Frontiers in Psychology in 2013, Sean Hutchins and Sylvain Moreno considered dozens of similar studies from the last two decades, and concluded,
“Using a variety of different singing and pitch perception tasks, some studies have found evidence of a correlation between the two abilities. However, many others, using similar designs, have failed to find a significant correlation.”
These findings indicate that singing accuracy is not a reliable measure of a person’s listening skills.
In assessment terminology, this again raises the issue of construct validity: whether a test measures the ability it claims to measure, rather than introducing additional factors which may influence the outcome.
And this is clearly problematic, because there will inevitably have been many examination results over the years in which marks awarded for singing materially affected the overall outcome.
Existing alternatives
If there are better, more effective approaches to assessing aural it seems to me incumbent on exam boards to adopt them as a priority. Happily for ABRSM they have already developed suitable alternatives.
In their Practical Musicianship exams they include a test in which the candidate compares a written score with an altered version played by the examiner. Trinity College London include a similar test within the aural component of their practical grade exams, which commendably avoid singing altogether.
These tests clearly evaluate the candidate’s ability to connect sound and symbol, including other musical elements such as dynamics, articulation, harmony and tempo. They not only offer a more accurate and inclusive approach to the assessment of audiation, but a more comprehensive one, too.
ABRSM also have at their fingertips a fully written, benchmarked set of aural tests for all grades that don’t include singing at all. Created for candidates with selective mutism, you can read about these adapted aural tests on the their website here.
Given that ABRSM already has up-and-running alternatives, and assuming they have done a good job in benchmarking them, why limit their access? Better, surely, to offer them as an alternative option for any candidate who would prefer not to sing in the exam room.
The introduction of ‘Performance Grades‘ by ABRSM and others has further demonstrated that it is possible to assess musical achievement successfully without a separate aural component. While these qualifications currently serve a different purpose from Practical Grades, they nevertheless illustrate that alternative assessment models are entirely feasible.
Where next?
In this article we have considered evidence that compulsory singing in aural tests is neither reliable as a means to assessing musical listening skills and perception, nor appropriate in the context of the statutory requirements for access and inclusion.
Meanwhile, the research of Hutchins and others adds scientific weight to our concerns. If an accurate sung response cannot reliably be treated as evidence of underlying aural perception, then singing cannot provide a sufficiently valid basis for assessing audiation.
Having largely focused on ABRSM, I should also note that MTB, established after Hutchins’ research had already been published, nevertheless adopted listening tests with a strong emphasis on singing. Given the available evidence, that decision is difficult to understand.
Designing valid, reliable and accessible assessments is no simple task, and examination boards inevitably balance many competing educational priorities. As UDL gains traction as a framework for assessment design, it is more urgent than ever for exam boards to offer a syllabus designed around proactive rather than reactive inclusion.
Whether in response to the science, educational rationale, or to promote wider inclusion in their exam rooms, I hope that all the boards will equally commit to removing singing tests from piano exams completely, refreshing their aural requirements so that they assess listening skills in ways that are valid, inclusive and appropriate for the 21st century..
Pianodao offers over 700 articles and reviews, all FREE to access.
If you appreciate this content, please support and follow the site:
