While writing a wind quintet this week, the necessary downtime was spent finishing a pair of wonderful and wonderfully idiosyncratic books: Thomas Pynchon's new novel, Inherent Vice, which arrived on Monday and not only read as quickly and smoothly as a Wild Turkey chaser but also carried the same emotional after-burn, and Henry Brant's orchestration handbook, Textures and Timbres. I'll probably have more to say about the Pynchon later, but for now: damn! between it and Mieville's The City & The City, this has been the year for the weird-boiled detective novel. As for the Brant, he is refreshingly modest and upfront in explaining the goal of the project in hand, which is achieving homogenous orchestral textures from mixtures of instruments. His method, through defining broad timbral categories, is a sound one and always based on years of empirical study. Nevertheless, the real substance in the book comes in a number of verbal suggestions and in many of the Brant-composed musical examples, which, particularly regarding the differentiation of contrapuntal lines, go a great distance toward a more comprehensive theory of the role of orchestration in relationship to each of the other compositional parameters. This is reading a more abitious program between the lines of Brant's text, but I think that inventive orchestration is more or less always going to have to be read between the lines of existing practices, deconstructed even, and Brant's own music was never less than that.
Although Brant is most famous for his works with spatially separated instruments and voices, he does not explicitly discuss space as a parameter here, but it's abundantly clear that clear that Brant understood orchestration as a necessary prerequisite to making spatial music actually work. Brant's book thus follows his own insight: the recipe for a homogenous orchestral timbre also contains the recipe for its opposite and suggests the timbres which would maximum contrast for the projection of contrapuntal materials.
Brant's book got me to thinking about the teaching of orchestration in general and, more specifically, his use of his own musical examples made me wonder which pieces of historical repertoire I would choose if I were to teach orchestration. There are really three approaches possible to the topic, one — and the tactic taken by the textbooks now most widely used — attempts to teach an optimal "mainstream" orchestration aesthetic based upon standardized instrumentations and ideals of contrast and balance; a second approach is essentially through teaching the history of orchestration, particularly in its luxurious change of focus from function to color; and the third approach is more analytical in nature, taking advantage on the one hand of a more contemporary understanding of musical acoustics and, on the other hand, considering scoring patterns as integral aspects of compostional technique.
Although my own composition instincts are mostly those of the third approach, I think I would teach with an acoustically-informed historical approach. I would skip around rather than hew closely to chronology. I might start, say, with a Schumann Symphony or Overture, pieces supposedly suffering from poor orchestration, and try to make the case against that supposition: the continuous doubling of the violins by the flutes, for example, usually considered a textbook example of what not to do, may just reveal instead a unique timbral imagination and conception/ When one considers that the bowing technique of Schumann's time involved a shorter stroke, with less sustain and considerable decay, allowing the near-sine tone of the flute to emerge from the composite tone lent the prevailing melodic surface of the music an edge for which the adjective "silver" is spot-on. Then, as a counter-example, I would take the Berlioz Messe des Mortes, a piece I really treasure, to illustrate — if somewhat paradoxically — mastery in a less-is-more style of orchestral economy, in which a large reserve of forces is kept mostly in reserve and often, in fact, paired down to some strange and wonderful subsets (unaccompanied choir in the Quaerens me, the antiphonal brass, the mens' chorus, low reeds and low strings of the Quid sum miser, the tenor soli "solo" in the Sanctus...). As an additional counter-example, I might use some Liszt piano pieces, as examples of exploiting the full palate of an instrument's resources as "orchestration" without an orchestra.
Only then would begin the historical survey proper, paying special attention to the historical cul-de-sacs, as the orchestral styles not taken but well worth revisiting: whole and broken consorts (especially in Morley's Book of Consort Lessons, with its contrast between sustaining instruments and plucked metal and gut strings), polychoral or antiphonal ensembles, the five-part string texture in Lully, and then trace the mainstream of orchestral development from continuo-based ensembles to the classical four part texture with the gradual incorporation of winds and percussion (the brass, in particular, represent an interesting bit of musical sociology as the horns came from the hunt, the trumpets from the military (each bringing their own drums) and the trombones from the church tower; the origins of orchestral percussion are often even more exotic) . I would spend some time on the Harmonie (wind) ensemble and trace wind bands through their court, civic, military, and popular traditions. Although I would pay some serious attention to opera orchestration (especially Rossini, Weber and Verdi), I might just skip much about the "normative" large 19th century orchestra, under the motto "nice work, if you can find it (or if your name is Johnny Williams) and you can always read about it in Adler, Blatter or Kennan" and focus instead on the contemporary chamber ensemble (with its "classical" but one-on-a-part instrumentation) and the composer-led ensembles (with their frequenty amplified and electronically mixed instrumentation).