Stay informed: Sign up for eNews Subscribe
Read Chapter 1

Part I

The Hydraulics of Musical Desire

Chapter 1

The Expansion Principle

The shaping of time counted among the highest priorities for seventeenth-century musicians. Of course, temporality always qualifies as a fundamental dimension of music making. But in the 1600s, composers sought to produce radically new, frequently extravagant experiences of time, alternately expanding and contracting, rushing impetuously forward only to hover in a state of apparent motionlessness. The arrangement of elements we recognize as tonality figured among these, but it often operated within contexts that also encouraged erratic fluctuations or nearly flat, virtually minimalist options. In this chapter, I wish to ask not why musicians persisted in using perverse procedures (the focus of several subsequent chapters), but rather why they occasionally found what we might regard as "tonal" arrangement advantageous. Along the way I will attempt to explain the mechanism that transformed particular modal patterns into tonal configurations.

Present-day discussions of early modern music too often bracket off as "tonal" those elements that seem familiar, leaving as "modal" vestiges those passages that do not work according to later premises. As a consequence, many of these compositions appear incoherent-as odd jumbles of progressive and reactionary features. Bear in mind, however, the fact that seventeenth-century musicians continued to make full use of other options long after they had "discovered" the one traced in this chapter; from their vantage point, the resources deployed in sixteenth-century modality and those characteristic of tonality were not mutually exclusive.

As always, I proceed with the assumption that changes in style and syntax are driven by expressive demands. If there exists no abstract reason why tonality should have developed, plenty of historical ones do present themselves, which is why cultural contexts matter even to questions of musical process. Over the course of the seventeenth century, composers assembled the devices at hand in many different ways, only some of which resulted in patterns that sound familiar to us today. Over the course of this book, I will introduce models that allow for cogent, internally consistent accounts of these repertories within their own contexts, paying attention to the aesthetic reasons why the pieces that move in the direction of tonality do so within the framework of their own range of choices.

Before we can plunge into what appears to be more familiar territory, we need to ground ourselves in the grammar that musicians around 1600 would have understood as transparent. By tracing transformations in a small sample of modal compositional strategies, I hope in this chapter to throw into relief the changes that occurred within the first decades of the seventeenth century.

I should warn the reader that this process will require a considerable amount of rewiring. As I have learned through experience in graduate seminars, anyone who simply plugs in Roman numerals during these discussions, dismissing the modal parts as "yadda, yadda, yadda," will not be able to follow the arguments. Try not to succumb, in other words, to the temptation to read everything as always already (sort of) tonal.

* * *

I shall begin with a composition that operates entirely according to sixteenth-century modal premises. Giulio Caccini's "Amarilli, mia bella" would scarcely seem to need an introduction (ex. 1.1a). Initially published in Caccini's celebrated Le Nuove Musiche of 1601, it quickly became an international hit; a keyboard arrangement by Peter Philips appears, for instance, in the Fitzwilliam Virginal Book. Moreover, it still occupies a place of honor within vocal pedagogy as one of the very first songs typically assigned to voice students. We are so accustomed to seeing it in anthologies for beginners-and to hearing their still-wobbly voices negotiating its melody-that we may think of it as a baby piece. Few other pieces designed for babies, however, include a graphic image of sexual penetration.

Example 1.1a, b, and c. Caccini, "Amarilli, mia bella"

"Amarilli, mia bella" operates within the Hypodorian mode, transposed-as it usually is-to G, with B? in the key signature. In a plagal mode such as Hypodorian, the octave stretches from the fifth degree of the mode down to the octave below, with the final (marked here as a double whole note) located in the middle of the terrain (see ex. 1.1b). More so even than is typical of pieces in this mode, the melody of "Amarilli" stays almost exclusively within the diapente from D to G; only once does it descend into its diatessaron (the fourth reaching from the lower D up to the final, G), but it does so in a most dramatic way. Not only does this move produce a temporary modulation suggesting D as a rival final and A rather than G as the proper divisor of the D octave, but it also produces the effect of penetration mentioned above and discussed in greater detail below.

Amarilli, mia bella, Amarillis, my fair one,

Non credi, o del mio cor dolce desio, Do you not believe, o sweet desire of my heart,

D'esser tu l'amor mio? That you are my love?

Credilo pur, e se timor t'assale, Believe it, though, and if doubt assails you

Prendi questo mio strale, Take this, my arrow,

Aprim'il petto, e vedrai scritto in core: Open my breast, and you'll see written on my heart:

Amarilli è'l mio amore. Amarillis is my love.

I want to concentrate on the unfolding of Caccini's melody in "Amarilli," for that is where his grammar resides. The opening section (ex. 1.1a) presents a series of interrupted descents from D toward G: some of the gestures halt at {3^}, others at {2^}, but within a framework that makes its orientation toward G abundantly clear from the outset. The withheld final, G, appears finally only on "mio," thus matching the rhetorical conclusion of the lyric statement.

We are accustomed to tracking melodic trajectories in tonal compositions, of course: think, for instance, of the chorale in the final movement of Beethoven's Ninth Symphony or nearly any hymn or folk tune. But Caccini's modus operandi differs from these in the relationship between its melody and harmony. From the point of view of functional tonality (the harmony-oriented syntax of later music), "Amarilli" plops into a modal pothole on the word "credi," and it also seems to vacillate indecisively between G and B? as potential key centers.

Caccini's contemporaries, however, would have regarded his harmonies as immediately comprehensible, even transparent in their implications-especially so in the context of the diapente-oriented melody to which they lend their inflections. For nearly every pitch in Caccini's tune qualifies as a node of crucial modal information, each one confirmed on a one-to-one basis by harmonic support. Thus the move to F? on "credi" (m. 4) serves within the conventions of the day to put special stress on the descent to scale degree 4 (melodic C), which might otherwise escape notice as a mere passing tone. Moreover, the swerve to B? in m. 5 counts among the most powerful means of articulating the top pitch of the G diapente-so common as to sustain a vast body of improvisations (see below, ex. 1.2).

Because of this one-to-one relationship between fundamental pitches and harmonic changes, the music remains tethered temporally to the exigencies of the poetic phrases, allowing for an expressive effect Caccini (following Castiglione) called sprezzatura, an attitude of nonchalance or unstudied grace. Each tiny lyrical phrase points in the same direction as the others, but the harmonic choices offer various shades of coloring, a spectrum of accents, before the inevitable final appears. A singer who focuses on that very low level of activity and nuance may hope to pull off that quality of aristocratic ease so valued in the Renaissance courts.

To be sure, Caccini makes liberal use of leading tones, though he does so intermittently and always for local purposes rather than as a given of harmonic syntax. For instance, the harmonic F? marking the very first move ensures that we hear everything at the outset within the context of G, as does the melodic F? that follows. Indeed, we could consider this first three-bar phrase, with its consistent leading tones, as tonal, though if we were to do so, we would fail to grasp Caccini's choices as significant. From Caccini's vantage point, those F?s freeze us into a holding position, as does his choice of weakly voiced chords with F? and E? in the bass, for both of which he designates a sixth (not a solid fifth) above the bass. The opening three bars consequently serve to outline the terrain of the diapente, always pointing through F? to G as the final yet also delaying any "real" move from the initial modal function, D, the fifth degree. Within the context of "Amarilli," this strategy counts merely as a short-term special effect, though it is precisely this ability of leading tones to prolong that will open up the new world.

And thus the significance of that modal pothole-the emphatic move in the bass to F? on "credi." If the first phrase sustains D (despite the apparent mobility of the melodic voice), the appearance of F? in the bass forcibly pulls the controlling modal line down from D to C, thence to B? and the still-unresolved A, all articulated as genuine syntactical moments. Stopping tantalizingly short of the final, G, the melody returns to its opening position in m. 5, now harmonized with a powerful B? in the bass. But this time the melodic line halts lovingly ("dolce desio") on B?. At last, with the poetic punch line ("d'esser tu l'amor mio"), Caccini allows for a direct diapente descent all the way from D to G. Note, however, that the quick reference to F? in the bass in m. 7 holds us up on D, while weak harmonies permit the melody to slide unimpeded down to A, which accumulates considerable gravity before it finally resolves to G in m. 10 (see the reduction in ex. 1.1c).

I have trudged laboriously through these few bars in order to tease out how Caccini wields the leading tone-and, more important, his other harmonic options-for the purpose of inflecting his melody, most of which remains identical with the generating modal line. He did not write "Amarilli, mia bella" as a theory exercise, however. His melody line flirts and teases, always stopping short just before divulging its secret, each time starting all over again at D but shading its approach differently. If the tortured melodic and harmonic contour of the first phrase underscores the singer's pathos, the move to C on "credi" insists on his sincerity, and the brief arrival on B? pauses to savor Amarilli's beauty. Only the last phrase completes the message delivered so haltingly with all those fits and starts. Understanding how all these minute details signify can help the performer make this song something other than just a repetitive melody with a modal pothole in the middle. Imagine the late Marlon Brando reciting the opening terzet with his usual self-indulgence, inserting pregnant pauses between each phrase, putting mannered emphasis on the odd word here or there. Caccini's heavily weighted modal line invites just that kind of rendition.

Now for the middle section, beginning in m. 11. Here our speaker becomes more ardent, pleading his case to the point where he offers up an image of masochistic submission. For the purposes of this argument, Caccini alternates between melodic 3 and 2, with leading tones appearing consistently under the second scale degree, heating up the need for some kind of resolution. Yet regardless of the pressure, the second degree remains in place, creating a kind of membrane that resists further action.

At last, recalling the opening section's success with descents from the top of the octave, the melody commences an approach from D on "Aprim'il petto" in m. 17. But when it comes into the vicinity of the barrier pitch A in m. 19, the harmonic F? (reliable thus far at moments of would-be cadence) suddenly gives away to F?, thereby reinterpreting A as the fifth degree of the lower D (confirmed by a C? in the bass). All the urgency aimed at transcending A suddenly breaks through to a terrain of interiority not yet even hinted at. That A becomes the surface of the body opened up for a moment of profound erotic surrender.

This moment lasts for only for the duration of three melodic pitches: F?-E-D. Almost immediately, the speaker seizes onto his beloved's name and hauls himself hand over fist back to the outside world of the diapente. The ascent requires the assistance of leading tones for what we would call secondary dominants (B?, C?), all perfectly available and comprehensible within sixteenth-century modal practice; within this context, they contribute to the impression of intense physical effort. And just in case you thought you had imagined that moment of penetration, Caccini lets us hear the entire sequence again, note for note. A brief coda elaborates a major-key plagal ("Amen") effect with E? blossoming out on top, before the voice concludes with a chain of ornaments to be executed deep in the back of the throat-a wordless orgasm of sorts.

Caccini composed his songs in Le Nuove Musiche within a court context. He had first attracted the attention of patrons as a solo singer in Rome; when he put this collection together, he was affiliated with the Medici cultural establishment in Florence. Just the previous year he had engaged in a sordid squabble over the invention of opera, and he had rushed his own setting of Ottavio Rinuccini's Euridice into print after he had recognized in Jacopo Peri's original the wave of the future. In his preface to that publication, Caccini set forth the now-familiar story of the Florentine Camerata as a way of backing up his own claim. Le Nuove Musiche, with its detailed account of Caccini's celebrated performance style, meant similarly to nail down his right to having developed the new style long before Peri.

But Le Nuove Musiche principally showcases Caccini's talents at setting lyrical verse, for which he had few peers. His songs served as vehicles for his own chamber performances, as well as for those of his wife and daughters (including Francesca, a prolific composer in her own right). By publishing Le Nuove Musiche he also entered into the burgeoning commercial market, and his songs were evidently sung in households eager to emulate aristocratic culture. Caccini's versions of temporality, subjectivity, and syntax, however, are identical to those of the court madrigal. As exquisite as these songs certainly are, they do not engage the expansion techniques Peri and others brought to the table.

* * *

In order to track the prehistory of what gave Peri the edge in this competition, we have to turn to improvisatory practices-practices also fundamental to court entertainments, such as instrumental dances or the recitation of epic poems such as Ariosto's Orlando Furioso. These practices date back to periods during which no one felt the need to notate them, but they begin to appear written out in the "teach yourself to improvise" manuals that proliferated in the second half of the sixteenth century. Much like the blues progression of a later time, the formulas that served as the basis for elaboration rely on the most fundamental of patterns: in this case, the diapente descent harmonized in its most powerful and straightforward ways.

We know the most popular of these patterns best from "Greensleeves," sometimes attributed to Henry VIII and sometimes to the unfortunate Anne Boleyn. Like most other aristocratic amateurs at court, Henry and Anne would have known how to ring changes on these familiar formulas, whether or not he or she actually "composed" this particular song. The first part of "Greensleeves" unfolds over a passamezzo antico harmonization of the diapente descent, the second over the romanesca. In each half, the generating descent pauses on 2, then repeats the pattern for a full cadential arrival on the final (ex. 1.2).

Example 1.2. "Greensleeves" (with generating modal line in top staff)

In the first strain of "Greensleeves," the passamezzo antico formula harmonizes the fifth degree, D, with the final in the bass, making the mode's identity fully audible right from the outset. To our ears, the first move in the progression may sound abrupt and archaic: indeed, it presents an instance of parallel fifths. Nonetheless, this progression occurs very frequently in modal music (recall the weighted arrival on scale degree 4 discussed above in "Amarilli, mia bella"), even if composers usually worked to hide the baldness of the parallels.

The romanesca-represented here by the second strain of "Greensleeves"-differs from the passamezzo antico only in that it leads off by harmonizing the fifth degree with a mediant rather than the final in the bass. When the bass moves to F? this time (in support of the descent to 4), it sounds as if it is establishing B? as a tonic-and, indeed, it would have had that implication in the Renaissance as well. The arrival on the third function, however, clarifies the situation by pointing to G as the final. It is crucial to realize that this formula counts as perhaps the most powerful available harmonization of the G Hypodorian diapente descent, and it should not be regarded as simply flipping from one key (B?) to another (G). Composers chose one or the other of these bass lines on the basis of expressive strategy; but with respect to modal syntax, which is located in the diapente descent, they are virtually interchangeable.

What differentiates "Greensleeves" and pieces of this sort from "Amarilli, mia bella" with respect to musical procedure, however, is a new level of melodic activity. Whereas nearly every pitch of Caccini's tune operates as a full-fledged function in the generating modal line, each move articulated by a new harmony, romanesca formulas present the corresponding functions in the background; over that background progression, the tune we sing as "Greensleeves" flows with rhythmic ease and melodic freedom: it does not, in other words, just hammer down four times through the diapente (as in my reduction line above), but rather spins a graceful, imaginative web.

Yet we do not get lost syntactically, for even if we do not hear the generating modal pitches directly, we can rely on the long-standard harmonic pattern to keep us oriented. Note that the harmonies in "Greensleeves" are not, except for the standard cadence formula, "tonal." What I referred to in the Caccini example as a modal pothole occurs right on schedule here and for exactly the same purpose: to project strongly the descent to scale degree 4 in the generating line. The harmonies still qualify, in other words, as a secondary parameter, as elements that inflect the steps along the modal line.

The guidance of the harmonic pattern absolves the singer from having to stick to the literal presentation of the modal line, as in the Caccini. With the more widely spaced harmonic moves, this new melodic line moves around in a relatively relaxed manner; the melodic pitches that emerge to configure the sung tune qualify as ornamental with respect to syntax. The downside (for we always lose something for every advantage in a style change) is that this kind of music making relies heavily on formulas to make sense. Moreover, it depends-like the 12-bar blues-on a pre-set rate of harmonic rhythm: in this case, one change per measure except in the approach to the cadence. But by means of those formulas, the melody of "Greensleeves" can range from its lower D in the first strain to the F a tenth above in the second. Note that we are no longer grammatically required to specify a plagal or authentic arrangement of species, for this mechanism operates solely on the basis of the generating diapente.

* * *

Jacopo Peri excelled as an improviser within such formulas. Even his contribution to the 1589 Medici wedding festivities-the lavishly decorated "Dunque fra torbide onde"-operates according to such models, as does the celebrated "Possente spirto" in Monteverdi's Orfeo. When Peri turned his hand to writing the speechlike recitations for Euridice, he brought with him, as though by second nature, this double-level sense of melodic activity, and he used it to produce the highly directed yet easygoing style of recitation that made Euridice a historically significant event. As Howard Mayer Brown has demonstrated, Caccini failed to grasp the foundation of Peri's technique, and his resettings of Rinuccini's speeches stagnate in rudderless monotony.

But Monteverdi, an artist then celebrated for his polyphonic madrigals, more than met Peri's challenge in his Orfeo (1607). A veritable compendium of compositional possibilities available at the moment, this favola per musica boasts every device Monteverdi had ever employed, plus those Peri had introduced onto the stage. In contrast to Caccini, Monteverdi managed to figure out how Peri had accomplished his trick. After a prologue, in which he displays his ability to turn out variations over a set formula, he opens his first act with a dazzling instance of the new stile recitativo as a shepherd exhorts his companions to join in celebrating Orfeo's marriage.

In questo lieto e fortunato giorno On this happy, fortunate day

ch'ha posto fine a gl'amorosi affani that puts an end to the amorous longings

del nostro semideo, of our demigod,

cantiam, Pastori, in sì soavi accenti let's sing, shepherds, in such sweet accents

che sian degni d'Orfeo nostri concenti. as will make our strains worthy of Orfeo.

Our shepherd's speech derives its directionality from Monteverdi's use of a single diapente descent in the background: the entire statement occurs under the umbrella of that powerful trajectory. In this sense, it merely harnesses the energies available within the familiar romanesca-type formulas to perform its rhetorical task. In contrast to a standard, evenly paced romanesca, however, Monteverdi dictates the rate of change in the background. The kind of melodic flexibility already noted in "Greensleeves" here becomes even more elastic, as the composer determines how quickly to proceed through the standard progression, and the degree of expansion increases enormously.

For instance, Monteverdi prolongs the initial fifth degree, A, for more half the length of the entire speech (ex. 1.3). He begins by harmonizing it through alternations between D and A in the bass, thus producing a holding position. No leading tones occur here, for this is not a cadential situation. But that very indeterminacy allows the melodic line to move sometimes through B? to circumscribe A as the upper boundary of D (m. 2), at other times through B? to imply A as a potential final (m. 3). Monteverdi's holding position matches, of course, librettist Alessandro Striggio's rhetorical ploy, which gets our attention but then defers delivering its message until the arrival of the words "cantiam, Pastori." At this point Monteverdi reharmonizes his still-reigning A with the mediant in the bass, and with this romanesca rendering of the fifth degree he commences his more or less straightforward diapente descent.

Example 1.3. Monteverdi, Orfeo, Act I, scene 1 (with modal line in top staff)

But, of course, we do not hear this display as compositional experimentation but rather as the speech delivered by a character in a dramma per musica. By means of this quite audacious expansion effect, the shepherd reveals himself as an orator quite worthy of belonging to Orfeo's entourage. He has not been hanging around with this rhetorical demigod, son of Apollo, for nothing; he has picked up some of the boss's techniques along the way.

Note that I have indicated a potential return to the fifth degree in parentheses in mm. 9-13. And indeed, this monody has a middle section (not shown) that reinterprets the apparent closure of m. 13 as temporary, and it proceeds to explore extensively a much longer series of inflections. Following this section, the portion of the monody displayed as example 1.3 returns for full closure. In other words, the arrival on D in m. 10 can be heard either as a step on its way toward the final or as the reestablishment of the fifth degree, depending on the larger context. But in either case, the fact that the singer returns to a melodic A in mm. 9-10 is not the deciding factor any more than his opening pitch on scale degree 8 determines the grammar of this little piece (recall the similar disengagement between the melodic line of "Greensleeves" and its syntax). Whereas the tune in Caccini's "Amarilli" also functions as its governing modal line, the shepherd's tune in "In questo lieto e fortunato giorno" qualifies as a new level of melodic activity over and above the diapente descent that generates it.

Orfeo's wedding song, "Rosa del ciel" (not shown here), begins with an even riskier maneuver in which he addresses an extended apostrophe to the sun over a single sustained pitch in the bass. As I have argued elsewhere, Orfeo thereby commands the very sun (that is, Apollo) to stand still until he releases it with the verb of his main clause. The rhetorical control required for meeting this challenge would have translated into personal power in the Renaissance courts that so treasured oratorical skill. We do not have to be told explicitly why Orfeo has acquired so much personal clout: we hear it with our own ears as he sweeps us into his thrall during his wedding vows.

Clearly, Orfeo's deputy-our shepherd who opens act 1 with "In questo lieto e fortunato giorno"-cannot be allowed truly to rival his mentor, even if his own exhibit of rhetorical prowess identifies him as the spokesperson of the shepherds, as a character with a degree of social prestige second only to Orfeo's. As Suzanne Cusick and Nina Treadwell have demonstrated, the Medicis were so jealous of the simulation of authority afforded by the stile recitativo that they wanted to be recognized as the source of the energies generated by these devices. Consequently, much of the monodic music produced in their court remains anonymous, making it difficult to establish with certainty who actually composed it. The widespread perception of these expansion devices as performances of power explains in large part why such apparently simple mechanisms suddenly took over the whole field.

Although they enable a radically new temporality, the harmonic changes in "In questo lieto e fortunato giorno" operate according to modal convention: they still serve to articulate the moves-or to delay motion-in the modal line. Those two opening harmonies that rock back and forth under the sustained modal A do not qualify as tonic and dominant: they are but harmonizations of the fifth scale degree with the final and the fifth degree in the bass. Indeed, the basso continuo refuses to include a leading tone until the penultimate chord for a standard cadential formula.

* * *

But Orfeo also offers up some genuine instances of what I will count as tonal strategy. Orfeo himself opens act 2 with a little song in which he celebrates his marriage. A dancelike lyrical piece, "Ecco pur" unfolds over an active, rhythmically shifting bass line rather than the sustained-note bass of the shepherd's speech-oriented "In questo." Moreover, the harmonies all work to circumscribe a succession of local key areas, each with its requisite leading tone. As it turns out, the succession itself traces the familiar romanesca background. But now each node along the way becomes a tiny key area in its own right (see ex. 1.4).

Example 1.4. Monteverdi, Orfeo, Act II, scene 1 (with background progression in outer staves)

Ecco pur ch'a voi ritorno, Behold, I return to you,

care selve e piagge amate, dear woods and beloved hills,

da quel sol fatte beate made blissful by that sun through which

per cui sol mie notti han giorno. alone my nights have turned to day.

If "Greensleeves" and "In questo" offered an added melodic level to produce their expansions, "Ecco pur" also presents another level of harmonic activity: one of them still marking the modal progression in the background, the other sustaining each point in the background through cadential harmonies. I will define this particular hierarchical configuration as tonal, insofar as it makes use of circumscriptive chords to sustain each moment along the way in a single, nonredundant teleological progression. In "Ecco pur," all parameters point forward to implied conclusions: the background to its arrival on the final, the surface harmonies to cadential confirmation of each immediate tonic. However brief this song may be, it operates according to fully tonal premises.

Because the background still lies very close to the surface, however, we ought to be able to hear the effort exerted in keeping each moment going until the eventuality of the next. For example, both melody and harmonies strongly suggest closure at the beginning of m. 2. Only the infusion of additional energy produced by the voice's sudden leap from the final, G, up to D prevents the piece from ending almost before it has begun. Against what Monteverdi stages as great effort, Orfeo's melody delays the arrival until m. 4: the friction between harmony and melody in mm. 2-3 derives from Orfeo's resistance to what would otherwise be just a chain of parallel sixths (C/E?, B?/D, A/C) pulling down forcibly down through the diapente to the cadence on G. None of these would qualify as "real" moves in modality, however; they participate as middle-level devices in the prolongation. Yet rhetorically this passage offers the impression of Orfeo defying even the gravitational pull of cadential harmonies, as he puts a drag on each step, finally allowing the cadence to occur only when he deems fit.

A four-measure-long prolongation may not impress us much, given the epic expansions of, say, Bruckner. But it all starts here-with the pitting of cadential harmonies against their own fundamental tendencies. A pattern that arouses the expectation for immediate closure produces a spark of energy that a composer can harness for ever-greater extension. That desire for closure-indeed, the assumption that it always lies nearly within our grasp-cannot be allowed to dissipate; it has to be channeled by means of middle-level tactics such as the ones just discussed until closure is truly granted.

Note that the cadential premise always announces itself with the appearance of the leading tone, signaling the arrival on melodic {2^} that is poised to descend to {1^}. Consequently, the double levels already discussed in the previous example appear here as well: although my top line traces the diapente descent that constitutes the backbone of the piece, most of the activity within the closed key areas concentrates on the vacillation between {3^} and {2^}, with the tonic or final withheld until the point of closure. Accordingly, a Schenkerian analyst might well read the opening key area of "Ecco pur" as sustaining {3^} rather than {5^}, as in my reduction-or, indeed, as I have chosen to indicate when this same four bars recurs at the end.

But however much my reductions may resemble Schenker graphs, they have a significantly different purpose: namely, to demonstrate the gradual emergence of tonal procedures from units of meaning well established in earlier repertories. In 1607, a descent from 3 in G Dorian would have made little grammatical sense; a modal analyst would always seek the generating 5. Moreover, the B? area within a G Dorian context always signals D, the fifth degree of G. As the expanded steps of the diapente descent become standard modulatory schemata, however, the interest for composer and analyst alike properly gets transferred to the middle ground-to strategies for simultaneously maintaining and postponing the inevitable point of arrival.

The question of whether "Ecco pur" qualifies as tonal really depends upon the degree of prolongation one requires. Like the other pieces just discussed, Orfeo's song is based syntactically on a descent through a minor-mode diapente. I use the term "minor mode" here because the diapente of Dorian and Aeolian have identical interval structures; the same is true of Mixolydian and Ionian. So long as all activity occurs only within the species of fifth (as in "Ecco pur"), we have only two procedures: major and minor. Moreover, each point along the diapente descent in "Ecco pur" features nothing but cadential prolongations. None of the ambiguities or exploitations of modally sensitive pitches beloved of Monteverdi and his colleagues in other contexts (including the rest of Orfeo) arise here. Still, each of these "keys" is so brief that it hardly warrants our opting for an entirely different brand of analysis, especially when "Ecco pur" can be understood more fruitfully from a vantage point that makes it consistent with other sections of the opera.

Consequently, I prefer to hear this little aria as performing a particular rhetorical role, which returns us to the issue of power. Orfeo's fake-out at the beginning of m. 2 (where he suddenly pulls away from the cadence in which he seems to have acquiesced) and his series of dragged resistances against the bass in the next two measures contribute immeasurably to his aura of personal charisma. Those cadential harmonies, which would have sounded definitive to Renaissance listeners, want to slam the door shut, but he inserts himself as a sonorous wedge, preventing-or at least postponing-the gratification of closure until he can do it his own way. "Ecco pur" demonstrates, in other words, a particular way of deploying the techniques at hand for expressive purposes.

* * *

I will flash forward about forty years for my next example: the opening section of Antonio Cesti's chamber cantata "Pose in fronte." In this aria Cesti manages to extend his diapente descent for a full forty-four measures, as each step of the descent balloons up to the length of a full-fledged key area by means of cadential harmonies. Yet the dynamic tension between surface and background remains quite palpable: repeatedly the harmonies attempt to cadence, only to have the melody refuse immediate gratification, demanding a renewed pursuit of the goal.

Like many midcentury pieces trafficking so obviously with dammed-up desire, Cesti's cantata presents an explicitly masochistic scenario. The cantata's final aria (taken up in chapter 2) simply murmurs over and over, "I love the arrows and adore the chains." This opening segment, however, has the task of whipping up (as it were) the energies that will carry the entire multisection composite to its languid, satiated conclusion. And for this purpose Cesti employs the most teleological devices he knows: the expanded diapente descent.

Pose in fronte al mio tiranno Proud Love sets chains and arrows

fiero Amor catene e strali; down in front of my tyrant;

queste schiavo il cor mi fanno, the former enslave my heart,

quei mi dan piaghe mortali. The latter give me mortal pangs.

The cantata starts with a brief continuo introduction, a slightly elaborated tetrachord descent from B to F? in the bass (ex. 1.5). When the voice enters, the bass tries to repeat its easy confirmation of tonic, and at m. 8 both parts appear poised for cadence; one has only to allow the voice to descend to B and we'd be off and running toward the next goal. But instead the bass veers off to D? (leading tone to an implied E), driving the voice up to cry out "proud Love" in its high register. With the reappearance of A? in the voice and F? in the bass in m. 12, we seem once again on the brink of closure on B. This time, however, the melody refuses to comply with the powerful arrival in the continuo. Marshalling all its energies, however, the voice pulls itself up-apparently dragging the bass up with it-to a definitive cadence only in m. 17.

Example 1.5. Cesti, "Pose in fronte," section 1

Just as Orfeo in "Ecco pur" interposed melodic leaps, refusals to follow the lead of the continuo, and even grinding dissonances to perpetuate his expanded first key area (if only for four bars), so Cesti's melodic line must suggest immanent closure while doing everything in its power to postpone that inevitability. Cesti's interpolation of the secondary dominant implying E as a possible final, the willful avoidance of the potential cadence in m. 13, and the deliberate annexation of the bass in mm. 14-15 all qualify as strategies for expansion: once the desire for a cadence on B is whetted, the voice works to postpone the gratification of that event, thereby luxuriating in the mixture of pleasure and pain so loved by the libertines.

In order to mark m. 17 as an arrival, Cesti disrupts the moto perpetuo of his bel canto triple meter with a hemiola pattern. Still, the voice has virtually no time to revel in that accomplishment, for the following beat announces the move forward to the next stage in the background-D major-and the first of two settings of the second two lines of text. Like many arias of this period, "Pose in fronte" unfolds in an ABB? formal schema, whereby the initial section (A) sets out the tonic key area and the opening lines of text while the second moves twice through the remaining text (B and B?), accomplishing whatever secondary key areas the composer wishes to touch upon before returning to tonic for ultimate closure. In this particular aria, Cesti chooses to remain within a single harmonically stable (though rhythmically turbulent) key the first time through. What might have produced an arrival in D major in m. 20 gets finessed, as the voice pushes onward, extending this key area forcibly until m. 26.

The reiteration of these same two lines proves rather more intricate, however, even though it does eventually work its way back to the closing tonic. For fast on the heels of the D major cadence in m. 26 comes what at first seems a return to the tonic, which could materialize as soon as m. 31. But the bass delivers a deceptive sixth degree in place of the expected B. Moreover, the melody flips up to E, which maneuvers the piece into a circumscribed area on A, the typical romanesca harmonization of scale degree 4 with the subtonic in the bass; as in our previous examples, this move announces a genuine descent from the background fifth degree, which has been sustained by both B minor and D major. The descent continues to 3 in the modal line, harmonized first as if returning to D, then with G, as the voice's abject C? seems almost to concede to yet another key area before a last-ditch effort pulls us back to the safety of C? (2) and a final, hemiola-marked cadence on B. The continuo rounds out the aria by repeating its ritornello.

To sing this aria one is forced to become breathless, to pant urgently after every designated goal, only to be goaded mercilessly on to the next. Both performer and listener are hurled forward in time, the present moment never anything more than a prod toward an ever-retreating future. Because of this breathless quality, whereby the singer marks each arrival with a gasp, we can still hear the generating background progression relatively easily and also the drastic means necessary for maintaining each area. As in the previous examples, the descent to 4 in the background sounds disruptive, even as it serves to pull us on toward closure.

Only when the final itself appears as the goal does the trajectory come to some sort of halt-a halt that within this style always sounds provisional and even artificial, despite the broadening effect of the hemiola. For once the desire machine gets wound up, it proves very hard to shut it down, to make any particular prize sound commensurate with what had compelled us there. In this case, the prize sought after so fervently and with so much genuine heavy breathing is a B. No big deal-until it becomes grist for the mill of tonality.

* * *

Everything necessary for the standardization of tonality exists at this point. In their respective pieces (or at least the segments discussed here), Monteverdi and Cesti not only make use of functional harmony, but they also reveal that they grasp the concept of hierarchical nesting, the process that would enable the repertories of the next three hundred years. If only seventeenth-century composers had recognized the potential of "Ecco pur" or "Pose in fronte" immediately!

Alas, historical events rarely unfold in such a straightforward fashion. In fact, the middle sections of Cesti's cantata exploit the erratic moves of his still fully operative Aeolian (not tonal minor) mode: they do not begin and end in the same key but simulate instead the ambivalences of a tormented, self-divided subject. Only this opening section, which sets the stage, qualifies as tonal-and even here the Neapolitan in m. 37 undermines the strength of this temporary tonic arrival, pointing forward to subsequent sections.

In other words, Monteverdi, Cesti, and their colleagues understood perfectly well what to do with the expansion procedures just discussed. Yet they persisted in treating them merely as tools they could implement when they deemed them appropriate for their particular rhetorical purposes-or push to the side when they were not necessary. As it turns out, composers continued for several decades to work within the structures already familiar to them from the madrigal, expanding some moments melodically, others by means of circumscriptive harmonic devices. And sometimes, even in music as late as Bach or Vivaldi, a passage of intensive expansion will give way to part of the background line that presents itself as the surface, with no elaboration whatsoever.

Seventeenth-century composers used expansion devices much as a photographer might a zoom lens. The performer, listener, or analyst must be prepared for a radical change of temporal orientation at any moment. Thus in order to follow their strategies, we need to have at hand-as did they-the potential complexities of modal structures, as well as a number of special operations that allowed for various degrees of expansion.

In short, tonality does not actually supplant modality. Mode continues to provide the background of successive modulations that lends a sense of inevitable telos to formally closed sections, such as the ones examined thus far; mode also supplies the logical connections among the series of units in multisectional compositions. Functional tonality then qualifies as but one particular way of construing the available resources. Moreover, for much of the seventeenth century it fails even to count as the most popular of the options. The following chapters examine a wide range of those options and attempt to understand why composers chose to operate as they did in a variety of different cultural and expressive contexts.