By Saul Richardson, Princiapl Jazz Workshop Australia.
In this article I show why students should study with a teacher who plays the same instrument as the student.
It is common practice in some music programs for one teacher to tutor many different instruments. This is done for economic and scheduling reasons, and is not an educationally sound practice. In fact, it has been shown to be at best a hindrance to progress, and at worst harmful to the student.
Students should study with a teacher who actually plays the instrument they are teaching, and who plays the instrument during lessons to provide a model for the student to emulate.
Modeling is a term coined by Albert Bandura (1971, 1977), a Canadian psychologist and one of the most influential learning theorists in education today. A broad summary of his theory is that people learn by observing other people’s behaviours, actions, attitudes, and their outcomes.
When we watch a cooking or gardening show on television, we are learning through modeling. We observe the steps involved, how to carry them out, in what order, what the various techniques look like, and what the final result should look like. One of the most powerful ways people learn about various sports and athletic skills is by watching experts doing it and then trying to emulate them in training. The expert might be a coach, more capable peers, or professionals on television.
We learn to speak through modeling. We imitate, approximate and gradually master the sound and grammar of the language we hear. On this, Bandura wrote that “If children had no opportunity to hear speech it would be virtually impossible to teach them the linguistic skills that constitute a language…It is doubtful whether one could ever shape individual words by selective reinforcement of random vocalisations, let alone grammatical utterances” (Bandura, 1971).
Children who have been isolated from spoken language, through abuse for instance, do not develop effective language. An infamous example was the case of “Genie”, a thirteen year old Los Angeles girl who had been kept locked in a near-silent room for her whole childhood. She was discovered by police in 1970 and had developed almost no language whatsoever (Rymer, 1994).
Children learn to speak English when they grow up in an environment in which English is the spoken language and the principal model to which they are exposed. If they grow up in France, they will most likely become fluent in French, and so on.
In music there are several crucial aspects of performance for which effective learning relies on modeling. The first of these is tone production. Others include articulation, technique, posture, style, rhythm and improvisation in a given style.
It is possible to play most instruments with a vast array of different tones. In each genre of music there is a rather more limited spectrum along which some tones will be considered acceptable as “in style”, or characteristic or correct. All other possible tones, those that fall outside the range of what is widely accepted will be viewed as wrong. It seems that in “classical” music there tends to be a comparatively rigid definition of ideal tone for each instrument. Jazz, on the other hand, permits a broader spectrum. But even jazz has recognisable limits.
For example, there are many different tone qualities on the electric guitar that are recognisable as jazz, from Charlie Christian through Pat Metheny to John Scofield. However, even on electric guitar there are limits. The sound of a Heavy Metal lead guitar, for instance, falls outside the spectrum of generally accepted jazz guitar sounds.
How do students learn to produce an appropriate tone on their instrument? Through modeling. Models can and probably should include professionals in live and recorded performance. But for students lucky enough to have a teacher, that teacher is their most important model for tone. As Warren Haston, instrumental pedagogy researcher and professional saxophonist writes in the Music Educators Journal that “Students will imitate a model whether it is good or not, so modeling with characteristic tone quality is essential.” (Haston, 2007). In other words, if an instrumental teacher plays with an excellent tone, their students will imitate that. However, if the teacher plays with a poor tone, then that is what their students will develop.
A teacher who is an expert on one instrument but “doubles” on another will tend not to play their second instrument at the same level as their principal instrument. Therefore, in such cases it seems unlikely that if they were to teach their less favoured instrument, they will not be able to model good tone as effectively as they would on their main instrument. It follows that their students will develop a less than perfect tone.
Even worse, if a teacher plays instrument other than what their student is learning, that student will be profoundly disadvantaged. A trombone student, for instance, who only ever hears French horn being played during lessons, is unlikely ever to develop a good trombone sound.
Besides tone, there are other aspects of musical performance that are most effectively learnt through exposure to good models. In fact some, such as nuances of style, articulation, rhythm, and idiomatic improvisation can probably only be learnt this way. That is, students must actually hear the music as it should sound before they can possibly replicate it themselves.
In an experiment conducted in 1987 Richard Sang showed that teacher modeling was a powerful way of helping students become better performers. One of his key findings was that “Teachers who have stronger modeling skills and apply those skills in teaching are more likely to produce students who perform better than teachers who do not” (Sang, 1997).
Sang’s was not the first study into this area of music education, however. In 1992 Marc Dickey, an academic then at California State University, Fullerton, conducted an extensive review of research done into modeling in music teaching and learning. Here is how he summarised some of the key findings:
“Students learn to make increasingly complex musical discriminations through modeling, via both musically appropriate and inappropriate demonstrations and imitations. The models used to teach music affect the way students think music should be performed. Students’ performance preference, sense of correctness, group performance, and individual performance are all positively influenced by musical models.” (Dickey, 1992)
“In order to be effective, musical models (both appropriate and inappropriate, or correct and incorrect) must be accurate models. Teachers who model must have sufficient skills to demonstrate one musical performance variable (e.g., pitch, rhythm, tempo, timbre, style) correctly and incorrectly, while modeling all other musical performance elements consistently.” (ibid)
Another researcher, Michael Hewitt reached similar conclusions in his (2004) study of middle school brass and woodwind players: “Teachers who use modeling techniques in their instrumental classrooms generally elicit better performance results from their students than teachers who do not model” (Hewitt, 2004). Hewitt’s study found that model range and timbre did not significantly influence the accuracy of student performance safter a twenty minute period, but he acknowledged the limitiations of such a short experiment (p.15). He also points out that singers have been shown to try to imitate the tone quality of a teaching model (Hewitt, 2004; Price et al 1994) and that his study cannot say anything about the effect of long-term exposure to different timbres or registers on student performance.
A large body of research and literature on learning theory, as well as anecdotal experience show that modeling plays a critical role in the development of music students. We can draw several conclusions that are significant to instrumental music teachers, to employers of such teachers and to students or their parents.
1. Teachers should play their instrument during lessons to provide a good model for their students.
2. Teachers should play the same instrument that the student is learning. It is bad practice and ineffective to play a different instrument because it provides a poor model. It should be unacceptable to employers or parents.
3. Research has shown that teachers should speak less and model more.
4. Teachers should be able to demonstrate deliberately, for contrast, bad playing as well excellent playing.
5. Modeling is extremely effective in ensemble rehearsals as well as individual lessons and should be used extensively.
Bibliography and References
Bandura, A. (1971). Psychological modeling: conflicting theories. Chicago: Aldine•Atherton.
Bandura, Albert. (1977). Social Learning Theory. Englewood Cliffs, NJ. Prentice Hall.
Dickey, M. R. (1991). A comparison of verbal instruction and nonverbal teacher-student modeling in instrumental ensembles. Journal of Research in Music Education, 39, 132-142.
Dickey, M. R. (1992). A review of research on modeling in music teaching and learning. Bulletin of the Council for Research in Music Education, 113, 27-40.
Price, H. E., Yarbrough, C, Jones, M., & Moore, R. S. (1994). Effects of male timbre, falsetto, and sine-wave models on interval matching by inaccurate singers. Journal of Research in Music Education, 42, 269-2M.
Rosenthal, T. L. & Zimmerman, B. J. (1978). Social learning and cognition. New York: Academic Press.
Ryner, Russ. (1994). Genie: A Scientific Tragedy. New York: Harper-Collins.
Sang, R. C. (1987). A study of the relationship between instrumental music teachers’ modeling skills and pupil performance behaviors. Council for Research in Music Education Bulletin, 91, 155- 159.