Sight-singing and dictation are staples of university-level musical training. While it is a worthy ambition to "see what you hear and hear what you see," as many state on their aural skills syllabus, I believe that there is a better goal to pursue: musical fluency. And while the ability to perform from sight and to notate musical passages after a limited exposure is a natural (and readily assessable) by-product of musical fluency, it is not the same thing as musical fluency. Further, based on some experiments with class-time this year, I hypothesize that repeated practice of sight-singing and dictation in class is neither the best way to develop musical fluency nor even the best way to improve students' ability to sight-sing or dictate.
What is musical fluency?
I define musical fluency as the successful depositing of musical information (a "vocabulary," perhaps) into readily-accessible long-term memory, and the successful cognitive assimilation of musical concepts and structures. I take this largely from linguistic fluency. One is fluent in a language when one has deposited a large enough vocabulary in one's memory, can access that vocabulary in real time, and has assimilated a sufficient subset of that language's grammar that one can function comfortably in a culture dominated by that language. Notice that while fluency will lead to and may be indicated by one's performance with the language in a culture bound to it, fluency is not the same thing as the ability to perform a list of tasks associated with that performance. Musical fluency is similar. While it will lead to an ability to perform certain tasks, and may be indicated by those abilities, it is not the same thing.
This distinction is important for a musicianship class. It is easy (and common) in American education to gear class activities toward the final assessment. If the assessment involves sight-singing, then class will involve a heavy amount of sight-singing. That makes sense if the goal is an ability to sing from sight and if repeated sight-singing is the best way to improve one's ability to sing from sight. However, if musical fluency is the goal, we instructors need to ask 1) are our musicianship assessments in sight-singing and dictation the best ways to assess musical fluency? and 2) are sight-singing and dictation practice the best way to help students achieve musical fluency?
Regarding the first question, I think that sight-singing and dictation are helpful indicators of aspects of musical fluency, but not sufficient to stand on their own as the sole indicators we use. Instead, we should spend some time as a community of aural skills instructors reconsidering other methods of assessment (perhaps enhanced by developments in technology) that may better indicate fluency, or that may be placed alongside sight-singing and dictation to provide a more complete picture of a student's fluency.
(It should be noted that, just as linguistic fluency involves more than reading aloud and dictating sentences spoken in a language, musical fluency involves more than performance and listening. It necessarily involves "theory"—i.e., analysis, composition, improvisation, verbal description, and the construction of reasoned arguments. With that in mind, I'm a strong proponent of "Musicianship" courses that combine "theory" with "ear training." However, I believe that the points I raise in this post can apply even if these topics are treated in separate courses.)
Regarding the second question, I think that devoting the bulk of aural instruction to sight-singing and dictation is not the best use of class time in order to develop our students' musical fluency. Other activities (which I will discuss below and/or in future posts) have helped my students grow in their musical fluency, and I think that further exploration of these and other activities will reveal better ways to spend class time in order to promote fluency—and even to increase student ability to sight-sing and dictate.
Assimilation and what it looks like
Let me take a step back and describe in better detail what I think musical fluency looks like by way of an example: compound meter. What does it take for a musician to assimilate the concept of compound meter?
Those of us who teach music theory and/or aural skills know the many strategies employed by music students to skirt the concept of compound meter while performing or dictating it accurately. (Most involve shifting attention to beat divisions instead of beats.) Performing a piece in compound duple meter while attending to the divisions (six per bar) does not constitute assimilation of compound meter, but rather clever employment of a workaround. Compound meter has been assimilated when a student understands that a beat can be divided by two (simple meter) or three (compound meter), and can apply that difference—specifically the three-division variety—in a diversity of musical activities. We only know that a student has successfully assimilated this concept when the student can explain the difference between compound and simple meter, recognize the difference by ear or in a score, can perform a piece in compound meter while attending to the triply divided beat (perhaps indicated by conducting that pulse while singing), etc.
Students employing a workaround will not be assisted in assimilating this concept by repeating the tasks in which they are already employing the workaround (sight-reading and dictation). Other tasks must be devised. In my experience, playing a recording while giving students the counting pulse and asking them to sing both duple and triple divisions of that pulse and then determining the meter of the piece is helpful. Forcing students to conduct or tap beats and not divisions during practice performing is another helpful task. Ultimately, though, sight-singing and dictation neither help students assimilate this concept, nor are they indicators of successful assimilation. (And in my experience, quizzes asking "How many beats are in a bar of 6/8?" are not sufficient either, given how many students answer such questions correctly, yet still can only perform or dictate when attending to the division.)
Improving sight-singing without sight-singing
This semester, I tested the hypothesis that sight-singing could be improved without in-class practice in sight-singing. Instead, we focused on developing students' familiarity with and conceptual understanding of some of the elements they would find in melodies I would ask them to sing. Specifically, we focused on rhythmic performance (with a combination of repetition of stock rhythmic schemata, and some sight-reading of rhythmic lines without pitch) and species counterpoint. The species counterpoint unit emphasized pitch over rhythm (gradually introducing simple rhythmic complications between species), and included both composition and performance. (Students worked in pairs to compose counterpoints in each species in major and minor, above and below the cantus firmus, and then were assessed on both their composition and their performance of those exercises. Only correct exercises were accepted, and their grade was the number of species they were able to compose and perform with a partner without mistakes.)
This unit involved a lot of performing, but no pitch sight-reading. All pitches sung were composed by the students or, in early stages, supplemented by the keyboard as they analyzed model exercises (we have class in the piano lab). Prior to this unit, students had assessed (mostly successfully) singing stepwise melodies in major and minor with beats and divisions, and tonic-triad leaps at least in major (some successfully assessed tonic-triad leaps in both modes). This unit was designed to expose them to diatonic leaps not bound to the tonic triad.
After this unit of rhythmic study and species counterpoint, we returned to melodic sight-singing. Students were given melodies with the same rhythmic figures studied in isolation, and with a wide variety of types of leaps (Karpinski's text and anthology through Chapter 31). Students were quite successful in class, and when I remarked about this, one student said, "This is easy. It's just what we did in species counterpoint." Which was, of course, the idea. Despite spending far less time on sight-singing in class than in last spring's iteration of the same course—probably less than ten meetings all semester—their final sight-singing assessments of the semester were right in line with the previous year's students. (I used the same melodies and a similar grading scale.) The difference, of course, was more class time to focus on voice-leading and other topics, which has paid dividends in those areas.
The reason I believe that this process worked is that the task of sight-singing involves three primary components: fluent knowledge of the musical structures contained in the melodies sung, an ability to perform at a relatively high level in a high-stakes, one-on-one assessment environment, and an ability to strategize appropriately to ensure success in that high-stakes assessment. When a student fails a sight-singing assessment, the failure can be caused by any one of these things. All are important, but good teaching requires that we identify which is in need of improvement, and fair assessment requires that we identify which have been mastered by the student. That's difficult to do if the only assessment is a brief task that combines all three.
It is also worth pointing out that strategizing can be learned in far less time than we spend practicing sight-singing. If indeed, as my experience suggests, concepts can be learned and performance consistency achieved in other ways as well, a lot of class time can be freed for other purposes, or potentially used to achieve a higher standard of performance from sight involving a broader spectrum of structures and concepts.
I hope to share more examples of this kind of success in future posts. However, this example illustrates that my hypothesis is viable and worth more rigorous exploration: sight-singing is a by-product of musical fluency, and tasks other than sight-singing can develop that fluency—and thus the ability to sing from sight—with at least as much efficacy, and possibly in less class time.
What are the real elements of musical fluency?
I don't have a set list outlining the elements of musical fluency. For one, it would not be the same for every program, career aspiration, student, or instructor. And in any case, I'm still in the process of reworking my current lists of course objectives into a description of conceptual fluency as I would articulate it. However, I think the key in making such a change is to differentiate tasks from concepts. We often construct tasks that include concepts, and then organize class around the tasks—both class activities and assessments. I think that teasing out the concepts from the tasks, orienting class activities around assimilating those concepts, and assessing the concepts as evidenced by the tasks rather than assessing the tasks themselves, will make great strides towards promoting and accurately assessing musical fluency.
I'm eager to hear if others have tried similar class arrangements and how successful they were (or were not). I'm also eager to hear what others think about the appropriate goals of these kinds of courses: specific skills, or conceptual assimilation evidenced in application. Please comment below if you have thoughts or experiences to share.