Yale University Department of Music

Rosalia, Aloysius, and Arcangelo: A Genealogy of the Sequence Author(s): Daniel Harrison Source: Journal of Music Theory, Vol. 47, No. 2 (Fall, 2003), pp. 225-272 Published by: Duke University Press on behalf of the Yale University Department of Music Stable URL: http://www.jstor.org/stable/30041091 Accessed: 29/06/2009 07:02
Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you may use content in the JSTOR archive only for your personal, non-commercial use. Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at http://www.jstor.org/action/showPublisher?publisherCode=duke. Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed page of such transmission. JSTOR is a not-for-profit organization founded in 1995 to build trusted digital archives for scholarship. We work with the scholarly community to preserve their work and the materials they rely upon, and to build a common research platform that promotes the discovery and use of these resources. For more information about JSTOR, please contact support@jstor.org.

Duke University Press and Yale University Department of Music are collaborating with JSTOR to digitize, preserve and extend access to Journal of Music Theory.

http://www.jstor.org

ROSALIA, ALOYSIUS, AND ARCANGELO:

A GENEALOGY OF THE SEQUENCE

Daniel Harrison

Introduction Among the structuresof tonal composition, the sequence is perhaps the most distinctive. Others equally important,such as the harmonic triad,the cadence, and directedvoice leading, are found, by way of contrast, in both predecessor and successor compositional styles, making while the use of unprepared them generic ratherthandistinctive.Further, accenteddissonancesharplydistinguishescommon-practice tonal procedures from those of the Renaissancepolyphonic style, experimentation with this feature continues well beyond the boundariesof tonal technique; Schoenberg's"emancipation the dissonance"is, in this sense, of end the ultimateand fitting contrapuntal of the seconda prattica.Yet the of sequence is most pointedly not a characteristic the fully emancipated style, having fallen into disuse in the late nineteenthcenturyas obvious thematicrepetitionbecame undesirable;those composers working with tonalityin the twentiethcentury,such as Hindemith,Shostakovich,Britten, and others, eschew it nearly altogetherin favor of other kinds of transposedrepetition.At the other extreme, the sequence is hardlyused beforethe adventof common-practice proceduresin the laterseventeenth 225

century,and even then, as will be shown, more as a special rhetorical effect than as a regularcompositionaltechnique.It is not only a distinctive techniqueof the common practice,but also co-extensive with it. At this point, it is useful to define the term, even if it is clear that this paperis not concernedwith a category of medieval Latin chant.A good startingpoint is providedin the work of JairoMoreno, who defines it as "amusicalphenomenonconsisting of the successive repetitionof a musical object... at two or more levels of a pitch array"(1996, 6).1 Further refinementssuggest themselves, such as considerationof the natureand extent (size) of the "musicalobject,"the relationshipamong levels of a given pitch array,and the numberof repetitionsof the object. About the latter,for example, the general rule seems to be the largerthe musicalobject unit, the fewer repetitionsa compositioncan affordbefore breaking the bank with boredom. Further,music-theory lore identifies the number3 (object+ two repetitions)as ideal for the sequence,which constrainsthe unit size.2 Relying on Morenobut tighteninghis definitionsomewhat,this paper treatsa sequence along the following lines: 1. A pattern is a musical object constitutedby harmonic,melodic, rhythmic,and metricfeatures.The length of the patterncan be describedby its rhythmic/metric durationas well as by the number of its discreteharmonicstates (chordchanges).A standard pattern has one or two harmonicstates and lasts up to four beats; a long pattern has more harmonicstates, lasts longer, and can be heardas a small, self-containedphrase.The standard patternis typical for the common-practice period.A patternmay also contain motivic subpatterns. 2. The immediaterepetitionof the whole patternat a different,nonoctave, pitch level producesa pattern transposition.The features of the patternare preservedin this operation. 3. Anotherimmediaterepetitionof the patternat a pitch level "in line" with the patternand its first repetitionproducesa pattern sequence, or simply, sequence. By "in line,"I mean thatthe directional vector in pitch space establishedby a patterntransposition is continuedwithoutdeviation,subjectto a single-octaveregister change to accommodatecertainpitch-spacelimits. 4. A patterntranspositioncreates a musical line segmentbetween two points. A sequence, however,implies potentialopen-ended linearcontinuation. sequence is thus projective,in a sense conA sonantwith Christopher Hasty's usage (1997, 78-95).3 Being a tonal-rhythmic event ratherthan a purelyrhythmicone, however, sequentialcontinuationcreatesexpectationsfor eventualclosure

226

and change. Sequences can thus be used for a numberof rhetorical effects, such as transitionand climax, that are unavailableto patterntranspositions.Structurally, sequences (especially those with shortpatterns)createlinearprogressionsaccordingto Schenkeriananalysis. Some features in each of the claims above can be altered without changing underlyingproperties.For example, the main outlines of sequence structureremainclear even if the patternis varied or elaborated in repetition, and even if the directional vector is adjusted slightly to avoid voice-leading problemsor to create a rhetoricaleffect. Such altersequence, ations became prominentand useful in the nineteenth-century as RichardBass has documented(1996). We will take our compositionalbearingsfrom the works of Arcangelo Corelli (1653-1713), which are situated at the confluence of various of streamsof patternrepetitionand which formedthe background subsemusic. As we will see, these streams quentdevelopmentsin instrumental have headwatersin techniquesroughlya centuryold. Corelli andhis contemporariesused these techniquesto create new and powerful means of controllingtonal flow. But Corelli made such control, by sequence and other techniques,an explicit (perhapseven the main) featureof his handling of tonal materials, a characteristicthat led ManfredBukofzer to write that he "was the first to put the tonal formulasto systematic use" (1947, 222) and could "takethe credit for the full realizationof tonality music"(220).4This, combinedwith othercirin the field of instrumental cumstancesdescribedlaterin this paper,makesCorelli's music an appropriate startingpoint for what WalterPiston (writing in 1941, only a few In yearsbefore Bukofzer)called the "commonpractice." this light, examining Corelli's sequentialtechnique-including its origins as well as its immediatedisseminationand developmentby others-can shed light on the early history of tonal music. Corelli's sequence types have two immediatepredecessors, samples of which (excerptedfrom trio-sonatamovements)are shown in Example 1.5The two passages sharea common harmonicfeature-root motion by descending fifth-a point of comparison whose significance will be assessed later on. But on a strikinglylarge numberof counts, these two sequences are quite dissimilar.The passage at (a) is motivically concentrated,dissonantby virtue of its constantsevenths,and, as a result, contrapuntallyengaging. By comparison,(b) is motivically expansive unto limpid. Comparedto thematic,triadicallyconsonant,and contrapuntally (a), (b) seems the more typically sequentialof the two from a later,common-practiceperspective,and we will thereforebegin our genealogical investigationswith it and its ancestors.

227

Rosalia The sense of familiaritythatExample 1(b) offers is due not only to its largepatternsandleisurelypace of harmonicchangecomparedto (a), but also to its multi-layeredmetric and motivic effect. The "fast"level is beat (parsed according to the ascending located at the dotted-quarter the level, markedas a on the example, is eighth-notepattern); "medium" at the half measureand is delineatedby the registraltransferof the double eighth-notepattern; the "slow"level is at the measure,delineated and by transpositionof the patternmarkedR down a second. The movement of a between differentinstrumentsand/orregistersdemonstratesa technique of motivic alternationcharacteristicof Corelli's style and that of other trio-sonatacomposers. This effect is a descendent of polychoral and echo techniquesdeveloped at the turnof the seventeenthcenturyby, among others, GiovanniGabrieliin Venice. PeterAllsop has traced the lineage of these effects in Italianinstrumental music from Corelli's gen(a)
6
Violin 1

Violin I

Violone
e Cembalo 6 7 7 7 7 7 7 7 7 5 6 4 3

(b)
R 22 Violin I R R (alt)

Violin II

Violone e Cembalo

tt

Example 1 (a) Corelli 2.9.1, mm. 6-8 (b) Corelli 3.5.3, mm. 22-5 228

erationback to GabrielithroughotherVenetiancomposers such as Giovanni BattistaRicci and Dario Castello. In describingthe texturalideals of the Venetiansonata,Allsop points out that Muchof the rhetoric theVenetian of sonataa2 [i.e., two solo instruments andsupporting basso continuo]actuallyhas its originsin 'echo' compositionsratherthanvocal monody,and this colouredthe development of the mediumthrough remainder the century. of the This type of composithanpolyphonicandits transference rather tion is essentiallyantiphonal to the a2 mediumlent itself to the cultivation a limitedrangeof basic of textureswhich became the stock in tradeof the Venetianstil moderno sonata(Allsop 1992, 87). Among the textureshe cites are: * Motivic dialogue:A motivicfragment, often articulatedby a rest, is passed between the two instruments; * Antiphonalstatement:differingfrom the above in greater length of the subject; and * Parallelmovement:extensiveparallel thirdsbetweenthe treble instruments. In Example 1(b), motivic dialogue (betweenthe continuoandviolins) and parallelmovement(between the first and second violins) are clearly evident.6While dispositive evidence of the influence of Venetianecho techniquesmay be lacking in this case, examples of "echoed"or otherwise "motivicallydoubled"sequences of the Example 1(b) type abound in Corelli's works. Example 2 shows one instance in whichforte/piano quintessential the originalecho techniqueareused sequento alternations tially. A multi-layeredeffect is presenthere, as in the previousexample. The fast subpatterns markedby dynamic changes, and the slow patare terns are again markedby transpositiondown a second of the pattern, labeled once again as R. A more complex case, without the obvious forte/piano alternations but with "faux"antiphonaleffects from alternationsof concertino and concerto grosso, is shown in Example3. And finally,Example4 presents a sequence without echo or antiphonaleffects, but one having the characteristic motivic doubling associated with those techniques; note that afterthe mannerof the doubled material,a, could easily be orchestrated Example 3, with the first a given to the concerto grosso and the second to the concertino. These last threeexamples bring increasinglyto the fore a certaincrudity in the connectionof the patterns,which have been markedas R in the precedingexamples.The wealthof materialwithinthe R patternsfocuses attentionon the patternsthemselves at the expense of transitionsbetween them. The overalleffect is one of simplejoinery:each patternis buttedto 229

39 Violin I

Violin II

Basso

Continuo

R 42

Vln

Vln

Vc. 6 5 6 5 9 6

Example 2. Corelli 3.5.2, mm. 3944

its successor without any mediating (and strengthening) mortising, resulting in conspicuous seams in the local structure. Questions arose rather early in the history of music criticism whether this kind of motivic joinery was artless and to be disparaged. Early criticisms focused on contrapuntal legalities and licenses, as we will see below. These were inconclusive, but questions of artistic value continued to dog the technique well into the eighteenth century. For example, writing about the composer Franz Xaver Richter in 1772, Charles Burney noted that His detail and mannerof treating [his subjects] is frequently dry and steril, and he spins and repeats passages in different keys without end. The French and Italians have a term for this tediousness, which is wanting in our language, they call it Rosalie, or Rosalia. (The term is derived from the name of a female saint remarkablefor repeating her Pater noster, and stringing her beads more frequently than even St. Dominic himself, or than any other pious person than has merited a place in the Golden Legend.) An Italian cries out, upon hearing a string of repetitions, either a note higher, or a note lower, of the same passage or modulation, ah, santa Rosalia! Indeed this species of iteration indicates a want of invention in a composer, as much as stammering and hesitation imply a want of wit or memory in a storyteller.7(Scholes 1959, 240)

230

Burney's account was translated into German, rather freely by J. J. C. Bode, in the following year. Bode curiously ascribed the term Rosalia to the French alone and then omitted Burney's mention of St. Rosalia as well as of the ironic "ah, santa Rosalia" exclamation.Moreover, he added a footnote of his own: call The Germans it a "cobbler's patch"[Schusterfleck], one can easand humorously referred as a "cousin ily see why.I havealso heardit rather to [Vettermicheln]. the Michel" Anyonewho has everheard well-known song

Concertino

Concerto
Grosso

Example 3. Corelli 6.7.1, mm. 24-35 231

Concertino

Concerto Grosso

R 61

Vc.

Vc.

Example4. Corelli 6.4.4, mm. 56-67 "Lastnight cousin Michel was here" will know immediatelywhat is meant when someone says, "this or that composer 'cousin Michels'." (Burney1959, 267) Bode's footnotesuggests thatdiscomfortwith this kind of motivicbutting was not confined to the French(and Italians)alone, and, moreover,that by the eighteenth century critical reaction to them tended towards the humorouslycutting. Perhapsstimulatedby Bode's translation Burney,C. E D. Schubart of published a short article, "ConcerningRosalias,"in a 1774 issue of his His Deutsche Chronik. explanationfor the origin of the termdiffersfrom Burney's:"ARosalia is successive repetitionof the same motive without preparation. namecomes froman old secularfolksong, 'Rosalia,mia The cara,'the chorus of which presents a repetitioneven cruderthan in our German"CousinMichel" (Schubart[1774] 1839, 164).8He then quoted the passage in question, which is presentedin Example 5(a).9 For comparison, the chorus of "CousinMichel" is shown at (b). Unlike Burney and Bode, Schubartwas of two minds aboutthe aestheticvalue of Rosa232

lias. Although he viewed them as symptoms of a superficialand trendy RosaItalianstyle, he nonethelessdid not condemnthemoutright."Truly, lias alwaysbespeakof a spiritual poverty;in music the listenermustnever be able to guess what will follow,"he writes in summary;"Yetthere are cases in which Rosalias arepermitted-indeed, in which they createreal beauty,"going on to mention a "Halleluja" Caldarain which "a simby impressionupon my heart."Further ple Rosalia made an extraordinary along these conflicted lines, Schubartavers that "Rosaliasare insufferable on the piano, but bearableon the organ... [And] in sacred music they offend less than in opera"(ibid., 165-6).0o All of Schubart'smusical examples of Rosalias involve ascending despite both his silence on mattersof transpositional transposition, direction and Burney's explicit indicationabove that they involve repetitions "eithera note higher, or a note lower.""This happenstancemay be the cause for the currentunderstanding Rosalia as an "identicalrepetition of Further, Schubart'sexamples exhibit possiof a melody a step higher."'2 bilities for a single type of harmonicsupport.This has led one scholar furtherto restrictthe meaningof Rosaliato only those sequencesthatuse the harmonicpatternI-IV, II-V, etc. (Jersild 1982). In the following, I will resortto somethingcloser to Burney'soriginalsense of the termand use it to cover cases where prominentseams can be heardbetween transposed patterns.Those by step are the classic cases, but, as we shall see, the techniquecovers other transpositional intervalsas well. The prominence of the seams can rangefrom the subtle, as in Example 1(b), to the both the motivic obvious, as in Example4. But all involve foregrounding contentandthe transposition individualpatternsat the expense of tranof sitional techniqueslinking the patterns.

(a)
after Schubart 1774

(b)

Example 5 (a) "Rosalia,mia cara," chorus (b) "GesternAbend war Vetter Michel da," chorus 233

R I I I

R I

Example 6. Bull, "Dr.Bull's my selfe," mm. 1-4 (FitzwilliamVirginalBook #189) Both Bode and Schubartemphasizethe crudityof Rosalias by linking them to folk songs, and Bode's "cobbler'spatch"metaphor(via Riepel) furtherlikens the techniqueto that of a lowly artisan.This suggests that we might look to the schemas of popularmusic for otherancestorsof the sequence. To be sure, the lines of communicationbetween "high"and "low" musical art were not so constrainedthat the cultured Schubart, Bode and their critic friends did not know (or could not sing) the street ditties "VetterMichel" and "Rosalia,mia cara."In fact, Rosalias can be that found in hybridrepertories occupy the middlebetweenhigh and low, such as dances or variationsets on populartunes. Such is seen in Example 6, the first strainof an early seventeenth-century jig variation by John Bull in which the bracketedmotivic patternR is repeatedtwice, first a step higher and then back to the originallevel. This passage exhibits the characteristic motivic butt-jointsof the Rosalia; what it lacks, however, is the consistency of transpositionaldirection that is essential to the sequence. Thatis, the passage shows two patterntranspositions, first the upwards,the second downwards.13 It is an open questionwhetherthe closely transposedstepwise repetition typical of the Rosalia arose first in artor popularrepertories.It certainly found a congenial setting in popularrepertories,where repetition has always been a positive value, and Corelli's da camera works in particularhave a distinctpopulartouch. But opportunitywas also available upon pairedmotivic in echo compositions;these originallyconcentrated statementsat the unison, but they eventually used other transpositional intervals. In any case, the technique certainly shows up alongside the
234

extended echo and antiphonaltechniques that Allsop discusses in the repertoryof Venetian a2 sonatas. Example 7, an excerpt from a 1629 sonataby Dario Castello, is a case in point, with its figurationaldialogue extendedtowardsclimax (perhapsoverly so) by means of Rosalia. The Castello excerpt quite clearly exposes an importantcause of technical-aestheticdiscomfortwith Rosalia for contrapuntists, different perhapsfrom those affecting Schubartet al. For the series of ascending five-three chords exhibits a careless disregardfor cherished values of composition-namely, varietyin harmony,interval,and lineardirection. To be sure, consistency of linear direction is one of the features of sequence, and this consistency has potentialfor positive aestheticeffect, especially in the creationof registralclimaxes and nadirs.But, in Example 7 at least, the structural implicationof parallelperfect fifths at every
31 Treble I

Treble II

Trombone & B.C.

34 I

II

Trb./B.C.

36

II

Trb./B.C. 6 6 7 5 4

Example 7. Dario Castello, Sonata#12 for two treble instruments, trombone,and continuo.Allegro #3, mm. 31-38. 235

ViolinI

ViolinII Basso Continuo

Example 8. Corelli 2.5.4, mm. 17-20

chord change is difficult to ignore. Further, even if the skilled thoroughbassist avoids them in the accompaniment the usual contrary-motion by methods,prominentperfect fifths are projectedin the instrumental parts at every downbeat.14 To a greater or lesser extent, Rosalias give rise to faulty perfectintervalparallelmotion. In some cases, such as in Examples4 and 7, the impression is particularlystrong;in others, such as in Examples 2 and 1(b), it is softened somewhatby mediatingcontrapuntal events. Even so, the problemof contrapuntal continuityis always present.Loversof Rosalias could claim thatthe seams between motivic patternsconstitutewhat KnudJeppesen,following Riemann,called "deadintervals," that,in other words, the seams effectively seal off each motivic pattern as a selfcontained contrapuntal entity (1970, 160). Certainly, this explanation must be invoked for a situation such as the patterntransposition rests (attemptto) cut off commushown in Example8, where the quarter the nicationbetweenthe patternsof the Rosalia and "prevent" creationof parallelfifths. But even this idea-that rests in a Rosalia could inoculate against bad intervals-was subjectto dispute, as will be discussed later in this paper. Aloysius This discussion of the potentialcruditiesand contrapuntal inadequacies of Rosalia brings us now finally to consider,by way of contrast,the contrapuntally passage of Example 1(a). While it shares irreproachable root motion anddirectionalvector with (b), the passage at (a) has none of its butt-jointedseams, the patternsconnectedby the much strongeroverlappingjoint. Passage (a) also is made up of very brief motivic patterns device unknownto the ignorant involving suspensionchains, a "learned" Rosalia or Cousin Michel. It is useful to consider (a) as not directly relatedto Rosalia, but ratherfrom anotherpartof the family tree. 236

The generatingprincipleof Example 1(a), the chain of suspensionsin the uppervoices, referencesa compositionaltechniquefar older thanthe echo and antiphonaleffects of early seventeenth-century vogue. Suspension chains and their sequential treatmentcan be found as early as effective case. The excerpt Josquin,and Example 9 shows a particularly is takenfromthe most profoundpoint in Josquin'slamentfor Ockeghem, a passage in which he sounds the names of those who should be Ockeghem's principalmourners:Josquinhimself, Brumel, Pierchon (de La Rue), and Compere. This startling direct address to living people is expressedby means of a canonic device involving suspensions,producing a sequence.The motivic materialof the canonconsists basically of an unorderedintervallicmotion {+1,-2 }, where the numbersindicate diatonic steps andthe plus/minussigns indicatedirection.In the deployment of this basic intervalmotion, orderingconstraintscome into play, so that the uppertwo partsuse (+1,-2) in a canon at the lower fifth with a onemeasuretime interval.15 lower partsuse (-2,+1) also in canon at the The lower fifth, but with a one-half-measuretime interval;the lower canon begins a half-measureafterthe upper.The net resultis outervoices moving in paralleltenths, an alto moving in contrarymotion with the outer voices, and a tenor who, because of the half-measuretime intervalfrom his dux, presents with a series of ligatures.This remarkablepassage is clearly rhetoricalin intent; the unusually persistent syncopationof the

118 Jos quin, Bru mel, Pir

Jos Jos
-

- quin...

quin...

Jos

quin...

123 chon,

Com

p8

re

Example 9. JosquinDesprez, La Diploration de Johan. Ockeghem, mm. 118-26 237

tenorallows for a metricallyconsistentrepetitionof the syllables of each composer'sname, emphasizingthe direct address. These suspensions,however,are all consonant:5-3-5-6 as reckoned from the bass. And this points up the principaldifference between sixsuspensionchains, composed of consonantsyncopes, and teenth-century ones, in which dissonancesarechained,as in Examseventeenth-century ple 1(b). Whereasthe consonantsyncope-chaindevelops as a rhetorical device of the Renaissance polyphonic style, underscoring,as we have point or sentimentin the text,16 dissonant just witnessed, a particular the chain originatesfrom an early seventeenth-century technique of extension throughrepetitionof cadentialfigures-of what Rameauappropriately called "avoidingcadences while imitatingthem."17 This technique is illustratedin simple strict-contrapuntal terms in Example 10. The situation at (a) is a standardthree-partformula involving a cadential 2-3 suspensionbetween the cantusfirmusin the descantand a fourth-species line in the middle voice; a first-speciesline is in the bass. The prototype of cadential prefixion, shown at (b), projects the cadential suspension backwardsfor some measures,increasingthe power of the final cadence by startingthe cadentialdrive sooner.The rhetoricaleffect of this device can be increased furtherby coordinatingwith the third voice to create "deceptive"cadences, an effect shown in the example in the thirdmeasure of (b). There, the possibility of a C in the bass following the previous G sets up a cadence on C-which is, of course, avoided by the suspending of B. This effective intensificationof cadentialextension brings forth a sequence, as the bracketsshow. Before exploringthe sequentialpossibilities of cadentialprefixions,it (a)

(b)

Example 10. CadentialExtension

238

10

10

10

10

10

( 5

5 6

5)

so

(2

3)

Example 11. Frescobaldi,CanzonaPrima (Second Book of Toccatas, etc.), mm. 82-8 is worth noting that chains of both consonant syncopes and dissonant cadentialsuspensionswere used extensivelyin the early seventeenthcentury-that, in other words, both techniquesof suspension were merged technique.A characteristic, by no means and into a joint "fourth-species" isolatedexample,is shownin Example 11,the conclusionof Frescobaldi's CanzonaPrima (Second Book of Toccatas, 1627). The passage is in two sections, a climax basedupon a series of ascending5-6 syncopes, as indicated by thoroughbasssignatures(addedby the present author),and an extended cadence, using a descending chain of 2-3 suspensions in the uppervoices. While neitherpartis strictlysequential,both aremotivically concentrated,as in Example 1(a). In the first part, after the alto subject entrance is completed (at the second beat of the second measure), the uppervoice moves (with one displacementby octave) in paralleltenths with the bass, using, at first,a motivic rhythmicfigure.The tenthsandthe figure are shown with arrowsand brackets,respectively.In the second part,constantsixteenthsin the left-handpartare organizedinto a consistent motivic contour.Again, neitherpartis a sequencebecause of motivic variation within the patterns,but both have the motivic concentration characteristic the sequence. of The initial developmentof this kind of fourth-speciestechniqueinto works of the early the sequence is located by Allsop in the instrumental composerTarquinio seventeenth-century Merula: 239

More than any other composer of his generation, [Merula] utilizes consistent harmonizations of recurrentbass patterns as an importantcompositional device. The commonest are those formed upon scalic basses, typically chains of [X] ij chords in rising passages and, slightly less usual, a series of [Y] 7-6 suspensions on a descending bass... Another much used patternalternates a rising second with a descending skip of a thirdalmost invariably harmonized with a series of [Z] i chords. These formulas, which occur in both fast and slow movements, of course became the staple fare of the late seventeenth century, but before Merula they are remarkably rare, and their consistent use constitutes one of his most significant contributions. (Allsop 1992, 135) The boldface letters in the passage (added by the present author) refer to three different fourth-species types. The first, X, is the series of 5-6 syncopes already encountered in Example 11; the second, Y, is the dissonant cadential prefixion, with the counterpoint lines inverted so that the 2-3 suspension is made into the 7-6; and the third, Z, is based on a combined-species prototype of the cadential prefixion, sketched in Example 12. (a)

6
5

6
5

6
5

6
5

6
5

6
5

6
5

(b)

6
5

6
5

6
5

6
5

6
5

6
5

Example 12. Tarquinio Merula II quatro libro delle canzoni da suonare a 2-3. (a) La Loda, mm. 14-15 (b) L'Anselma, mm. 29-30

240

It might be thought that invoking fourth-speciescounterpointin the presentcontext is somewhatsimplistic and even anachronistic, considering thatMerula'smusic predatesJohannJosephFux's famoustreatiseby Of some seventyyears.18 course, this gap is the opposite of what it at first seems, for Fux's theory was, if not based on Palestrina'slate sixteenthcentury style, then conceived as an homage to that "celebratedlight of music"-who died the year Merulawas born.Further, appearsthat,for it Fux, composers after Palestrina'stime (likely including Merula and his contemporaries)often evinced an "unrestrained insanity"of technique not suitable,to say the least, for study and imitation (Fux [1725] 1943, 17-18). Despite this impasse, forcing Fux's alter ego Aloysius into dialogue with Merulaet al. resultsin the disclosurethatFux's fourth-species counterpoint ends up modeling Merula's sequential technique more closely thanit does Palestrina'ssuspensions. To arriveat this unexpectedconclusion, we must first get past Aloysius's own explanations,which amounts simply to the fact that fourth species "is called ligatureor syncope, and can be eitherconsonantor dissonant"(ibid., 55).19While this statementhides the separatestylistic origins of consonantanddissonantsuspensions,mergingthe two techniques into a single category,it also seems to reducefourth-speciescounterpoint to a possibly indiscriminateuse of suspensions.Dissonant suspensions, in particular, suffer from a lack of stylistic context in Aloysius's presentation, being explained in the most general terms and not as a special device for cadences: One can see thatit is easy to find the consonanceinto which any dissonancemustresolve;thatis to say, it mustbe resolvedto the consonance if whichwouldoccuron the downbeat the followingmeasure theretarof if dationwereremoved. Therefore, the cantusfirmusis in thelowervoice, of theinterval the secondmustbe resolvedto theunison,thatof thefourth to the third,thatof the seventhto the sixth, and thatof the ninthto the octave.(Ibid.,56-57) Fromthis list, Aloysius restrictsthe usage of 2-1 and9-8 suspensions, prohibitingthem in series on accountof illicit parallels;yet he still allows single cases, since there is nothing wrong with the underlying consonances they displace. Even so, there is but one illustrationof a two-part 2-1 or 9-8 suspensionin both the fourth-and fifth-specieschapters,with illustration the fourth-specieschapter, two and in one additionalthree-part more in the fifth.20 Clearly,while Aloysius allows the single 9-8 in theory, he gives little encouragementin practice.This subtle favoring and disfavoringof certaindissonantsuspensionssuggests thatAloysius's explanationof dissonantsuspensionas displacedconsonance,while covering the mechanics properly,allows for too many dissonant suspensions thathave no stylistic precedentor contexts.Simple contextscan be found, 241

(a)

(b)

Example 13 (a) Fux [1725] 1943, fig. 143 (b) Fux [1725] 1943, fig. 94 however,in Aloysius's own counterpointsand those of his student,Josephus, and they suggest a furtherrefinement of the concept of fourthspecies dissonantsuspension. of In counterpoints threeor moreparts,the single suspensions4-3 and 9-8 are always treatedas componentsof 53 and '9 respectively,meaning that a 2-3/7-6 relationshipexists between the upperparts. In other words, both the 9-8 and the 4-3 can be understoodas cadential2-3/7-6 types with an additionalthirdpartbelow. In the two-partcounterpoints, the 9-8 is so rareas to be an exception, and the 4-3 is, perhapssurprisingly, only somewhat less so, being used only twice.21In contrast,the 2-3/7-6 formationis used frequentlyin two-, three-, and four-voice settings, so much so thatit appearsto be the defaultdissonantsuspensionof fourth-speciescounterpoint. The otherdissonantsuspensions,as the previous discussion has suggested, can easily be understoodas versions of
the default with added bass parts.22

the Significantly,in the model counterpoints, 2-3/7-6 formationis the only suspension type used in series.23Aloysius pointedly discourages overuse of this formation;but once again his composed examples belie his own advice, this time in the other direction.24 the contest between In "harmony," where varietyof melodic and harthe ideals of contrapuntal 242

monic interval is prized, and the effective compositional rhetoric of cadentialprefixionillustrated the dissonantsuspension,the latteroften in prevailsin Aloysius's fourth-speciesenvironmentsand, as a result, prosequences.Indeed,Aloysius passes over withoutcomduces rudimentary ment the remarkable counterpointshown in Example 13(a), in three-part which the first-speciesline in the descantis entirelyin paralleltenthswith the cantus-firmus bass. While not strictlysequential(save for the cadential prefixion), the counterpointshows a telltale sign of sequentialtechnique-pattern transposition-in the strictdependenceof the first-species line upon the varying "transpositionalinterval" of the cantus-firmus for line.25The utility of fourth-speciescounterpointas preparation a sequentialtechniqueinvolvingboth consonantanddissonantcadentialsuspension chains, suggestedby circumstancesdiscussed above, is spectacularly confirmedin Example 13(b), which is taken from that portionof Gradusad Parnassumdealing with fugue (Mann 1958, 94). This counterpointobviously models the passage from Merula shown in Example 12, as well as the many similarpassages found in the work of latercomposers, including Corelli. It does not-pace Aloysius-model anything found in Palestrina. Arcangelo Although Merula and his contemporaries may have begun to restrict the cadentialprefixion motivically, creating thereby the suspension serefined it using comquence, ArcangeloCorelli and his contemporaries parativelysophisticatedcontrapuntal techniquesthat increasedthe number and type of suspension sequences. As we will see later, they also betrothedthese to Rosalia, an alliance thatwas eventuallyfully consummatedin the works of Vivaldi,whence it descendedto J. S. Bach, among others,to become a settled-and anonymous-technique of the common practice. The following discussion engages in a detailed examination of the conditions, and possibilities that affect Corelli's structures, contrapuntal sequentialtechnique.The narrative here perforcebracketshistoricalcontext-the genealogy-while examiningstructural properties-the genetic code-against a background of strict-counterpointprototypes. These prototypesareorganizedinto seven classes: four with descendingvectors andthreewith ascending.The ratesof descent and ascentcan varywithin each class dependingupon the use of additionalelaboration,which rules out the more usual sequence classificationby consistentpattern-transpositional interval(e.g., descendingthird,ascendingfifth). Descending prototypes, based on 2-3/7-6 cadential prefixions, are examined first. At the outset, we must make a distinctionbetween these two inversionally-related structures reservethe "2-3" label for those and 243

formationswhere the suspendedvoice is in the bass, retaining"7-6" situations where it is in an uppervoice.26 Example 14 presents a catalog of increasinglycomplex elaborations of the basic 7-6 formulathat is shown by the staff combinationAl (i.e., staves A and 1 sounding together). Examples of each elaboration in Corelli's works are shown in the boxed text, except for Al itself, the MeWhile rulanformulaso common that specific examples are redundant.27 the increasingly complex elaborations of this type in combinations A2-A5 are largely self-explanatory,we should take note of suggestive connectionsto less complex elaborations,ones in which the 7-6 suspension is impliedbutnonethelesssuppressed.The most important the sucis cession of descendingparallel6chords-as ubiquitousin Corelli's music as Al itself-which would appearinitially to be unrelatedto cadential prefixion. Example 15, however, suggests otherwise.There, two related

1 1.1.2,mm.32-3; 1.3.2,mm.38-9; 1.12.4,mm.26-8

7 3 2

5 3

1.4.3,mm.3-6; 1.5.2,mm.2-3

7 3 3

6 3

1.8.2,mm.4 and 14

7 3

5 3

3.3.4,mm. 19-20

6 5

6.7.5,mm.58-63; 1.12.1,mm.2-3

Example 14. Class A sequence types (descending):7-6 suspension chain in outervoices, inessential seventhchords 244

Violin1

Violin2

Violone e Organo
6 t 6 6 6 6 6

24
Violin1

Violin 2

Violone

e Organo
5 6 6 7-6 6 7-6 6 7
S4

Example 15. Corelli 1.2.3, mm. 1-8, 24-9

cadentialapproaches,mm. 5-8 and mm. 24-29, arepairedin such a way thatthe formercould profitablybe thoughtof as a deformationof the latter-that is, as a cadential-prefixion figurewith the suspensionsremoved. The approach the Phrygiancadence in mm. 5-8 uses parallel6 chords; to suspensions but in a similar approachat mm. 24-9, cadential-prefixion appear,resulting in combinationA3 of Example 14, a variantthat preserves the half-quarterrhythmcharacteristic the movement.Parallel| of chordsin Corelliaresometimesdescribedwith referenceto fauxbourdon, of vocal remnants a fifteenth-century as if theseformationswereunaltered technique.The alternativepresentedhere may be superior:as non-suspending variantsof cadentialprefixion. Anotherrelated elaborationis shown in the alternatedescant line of Example 14, A'. This line, eitherby itself or in combinationwith line A', is not found in Corelli's works. Examples can be found in the work of earlier composers, however, in contexts that apparentlydid not interest Corelli.28Example 16 contains two passages from a work by Samuel Scheidt commonly known as the CanzonaBergamasca.29 The first passage, mm. 29-34, containsa combinedAA'. Althoughthe fifth leaps to a third over the 6 chord instead of progressing to the sixth as in the A' model, this license frees the cadentialprefixionfrom potentiallyconfusing interferencefrom the A' line. The second passage, mm. 122-8, is the 245

Example 16. Scheidt, CanzonaBergamascaSSWV 64 (LudiMusici I, 1621) triple-metermotivic relativeof the first. Here is found only line A', without the 7-6 series. (The fifth can move to the sixth in this case without fear of encumberinga cadentialprefixion.)Even so, despite the absence of the 7-6 prefixion, its spirit seems nonetheless to pervadethe passage in the same way as it did in the Corelli excerpts in Example 15. In this way, both the 5-6-(8) motion of line A'andmotion by parallel6 chords of seem enabledby an underlyingstructure 7-6 cadentialprefixion. Returningnow to the more elaborate versions of class A shown in Example 14, we note thatby equalizingthe durationof ornamental tones in A2-5, another7-6 type results in which the resolutionof the 7-6 suspension occurs over a change of bass.30 Combinationsof this type belong to anotherclass and are shown in Example 17(a). Staff B contains the basic 7-6 suspension(cf. the treble and bass ofAl), and staff 1 contains the middle voice of staves 3 and 4 from Example 14, now made rhyththe mically even and placed underneath suspensionvoices.31Corelli frequently ornamentsthis Merula-stylesequence (B 1) as in B 1 instances la; are given in the boxed text. Subposingthe leaping bass from A4 and A5 results in combinationB 1+2.32 verCombinationB3 (the equal-duration sion of A5) producesa 7-7 series, althoughthe voicing of the chords is not consistent,as the thoroughbass indicates;the first chordlacks a fifth, and thus conforms to the traditional7-6 suspension-chord construction. The second chordhas a fifth but lacks an octave. Fully consistentvoicing is obtainedin versionB3b, which adds a line in parallelwith the 7-6 suspension. The differencesbetween B3 and B3b might seem slight, but by 246

voicing all the chords in the sequence consistently as seven-five-three chords,B3b completes the developmentof the sequence type as a series of essentialseventhchordsout of the originalinessential7-6 context.Line b adds not only texturalrichness to the sequence, but also an additional source of energy since, by doubling the bass on the second half of the measure,it dissonateswith the upperline of 3. One way to understand the contributionof line b is as a metrically shifted 2-3/7-6 formationwith the upperline of 3, illustratedin Example 17(b). In this way, the leaping notes of line 2 can be seen as alternating"castingsout" from the wholenote lines of (b), appearingat every change of whole note. In Corelli's (a)
B

1 6 5 5 3 1.12.1,mm. 11-12;3.11.2,mm.35-6; 4.2.1,mm.2-3

etc.

2 8 7 3 5
3

5.3.4,mm.20-2; 3.7.4,mm.3-6

etc.
3 8 7
3 8 7 3

b:

(b)
B line3 top

lineb 2

Example 17. Class B sequence types (descending):7-6 suspension chain in uppervoices, essential seventhchordswith bass in descendingfifths 247

1 6 6 5 (6) 5 B1 9 6 3.5.4,mm.30-2; 3.6.2,mm.5-6

3.3.3,mm.5-6; 6.5.3, mm9-12

6.8.5, mm. 15-18;2.10.1,mm. 10-11

Example 18. Class C sequence types (descending):7-6 suspension chain in uppervoices, double-lengthunit
5

Violin1

Violin2

Violone
e Cembalo 7 1t 9 6 5 4 3 9 6i 5 4b 3 6b 5 5 9 6 6 5 6 5 9 6 6 5

C2a

Cl

Example 19. Corelli 2.5.1, mm. 5-10 works,examplesof B3 andB3b cannotbe distinguishedby thoroughbass alone, but by foregroundfiguration(see Example25 below) or by differing voice-leadingcontextsto which a thoroughbassist would be sensitive. However, the basic type, with or without line b, is quite common; one instance-shown in Example 1(a)-inaugurated our entirediscussion of Corelli's sequences. The thirdclass of 7-6 formation,apparentlyan original contribution of Corelli, is shown in Example 18. Here, the unit of sequence pattern, shown by the dottedbracket,is twice as long as in the previoustypes. The firsthalf of the patternconsists of a B 1, andthe second of a cadential-resolution figure that seems to "resolve"the first half. (See the discussion above concerning the similar situation illustratedin Example 10.) The 248

second half of the patterncan receive therebya mild tonic accent, which can be strengthened alteringthe B 1 fragmentso that both chords are by in 6 position as well as by adding some or all of the parenthesizedacciversion from Cl thatCorelli seldom dentals.C2 extractsa "first-species" uses; more frequentis a "third-species" elaborationof C2, a typical verThatCorelli consideredC1 andC2 fundamentally sion of which is C2a.33 similar is shown in Example 19, a passage in which a 2-3 suspension chain in the violins is treatedsequentiallyfirst by C2a and then by Cl. Note thatthe change here is not a simple matterof figuration;the switch takes place after C2a moves from D to Bb, whereuponCl picks up Bb area of the passage twice as long as before moving to G, makingthe BL, the otherharmonicareas. So far,we have seen threeclasses of sequencesbasedon the 7-6 cadential prefixion. Using principles of invertiblecounterpoint,Corelli fashioned certain2-3 forms of these classes, althoughhis options were quite limited, as we will see. Some of the possibilities are shown in Example 20. The 2-3 series is on staff D, and above it are the admittedlyproblematic supplementary lines that forced Corelli to use Dl almost exclusively, with one importantexception to be discussed below. The contrapuntal problems of D1- if one were to be a stickler-are between the whole-noteline of D andthe suspendedline of 1, which moves by 5-4/4-5 progression(the difference dependingon whetherline 1 is between the elements of line D, which, if so, would have to be realized as 9-10); the issue is whetherthe progressionof fourthsandfifths legitimatelydestroys

1
6

2--23

2
2

5---6

3 4-----5 3-----4 4
52-3 6 4 2--3 5 5 2 6 3

Example 20. Class D sequence types (descending): 2-3 bass-suspensionchain 249

any consecutive fifths.34 The lower line of 1, which alternately converges

on the intervalin question and then leaps down to an imperfectconsonance, takes attentionaway from the questionableline. The problem of consecutive fifths is greaterin D2, where the fifths are formed with an outervoice and accentedto boot. These fifths are mitigated,however,by being brokenup by sixths. Unlike Dl, there is no flexibility of realizing the 2-3 motion of D as 9-10; were line 2 interposedbetween the voices of D consecutive fifths would result with the whole-note line (with no Corelli never used D2. Nor did interveningintervalto mitigate them).35 he use D4, a double-lengthpatterninvertedfrom type C, which offers a line in brevesto breakup consecutivefifths;the immobilityof line 4 surely posed compositionalproblemshe would have ratheravoided. versionthatbothmade But he did tryD3 once, in a slightly ornamented the fifths dangerouslyapparentand played into a weakness of thoroughbass notationthat, in this case, workedto suggest that consecutive fifths were a compositional intention. The passage, shown in Example 21, sparkedone of those ardentcontroversiesbetween learnedmusicians so characteristicof seventeenth-centuryItaly; Allsop has termed it the "Affairof the Fifths."36 reductionshows the underlyingbasis of the The passage in D3, a basis that Corelli explicitly claimed. But he both decoand ratedthe resolutionof the suspensionby anticipation put a rest where the momentof suspensionwas to occur.The anticipations placedthe fifths closer to the beat, makingthem more apparent,and the rests in the bass since rests him signatures voice prevented fromwritingthe4thoroughbass What is left in the thoroughbassis a normally do not take signatures.37 series of blatant5s. Interestingly,the argumentsin the controversyfocused on whetherrests in general,and the eighth rests here in particular, were sufficientto "save"the fifths. The fact that the implied suspension was a dubious 4-5 and that it was deployed in series seems not to have been pertinent.38 in all, Dl has the fewest contrapuntal problems of All the 2-3 types and thus was used extensively wheneverCorelli wantedto vary the basic 7-6 formulasthroughinvertiblecounterpoint. We turnnow to the ascendingprototypes.Example 22 shows a schematic of a techniquefor which Corelli becamejustly famous:the braiding of upper lines in a series of 2-3 suspensions to create an overall ascendingeffect. Note thatthe generalascentbelies the originof 2-3 suspension chains in cadentialprefixion;the inventionof this figure marks the independenceof the dissonant-suspensionsequence from its structural origins. Interestingly,while the 2-3 braid easily lends itself to sequential treatmentin the bass, Corelli chose such treatmentonly occasionally since impressionsof paralleloctaves were all too easy to create. El shows his typical treatment,breaking up the octaves with sixths, althoughin both instances from op. 4 Allemandes cited in the example, Corelli ornamentedthe bass as in la, which repairs(if that is the word) 250

Violin 1

Violin 2

Violone Cembalo 5 5 5

Reduction

Vln. 1

Vln. 2

B.C. 5
4

6
5

Red.
Example 21. Corelli 2.3.2, mm. 3-6 the broken-upoctaves. Of course, any paralleloctaves are really illusory from the perspectiveof contrapuntal theory,since they are not consecutive between the bass andthe same uppervoice. Corelli's interestin findlaw ing and exploiting loopholes in the contrapuntal met with success in this case, since no controversylike the "Affairof the Fifths" arose in response to treatmentslike Ela.39 CombinationE2 is rare in Corelli's works, althoughthree short instances are cited. Yet it is notable for the connectionof variant2b, via techniquesshownin Example 14, to ascending sequences involving 5-6 motions, which will be discussed shortly. CombinationE3, a double-patternsequence, has but one derivativein Corelli's works, as shown in the example. Corelli invented a variantof the 2-3 braid shown in Example 23 in which the uppertwo voices alternateas agent andpatientin a suspension series. By moving a voice up a fourth(the crossoverintervalof the orig251

4.1.4, mm.9-10; 4.11.3,mm.6-7

4.2.4,mm.44-6; 3.6.4,mm.28-9; 3.3.4,mm.47-9

cf. 4.11.3.mm. 1-3

Example 22. Class E sequence types (ascending):2-3 suspensionbraid in uppervoices

36

63

3.1.2,m. 31; 6.2.4,mm.44-6

Example 23. Class F sequence type (ascending):Alternating2-3/7-6 suspensionbraidwith bass in ascendingfifths

252

inal braid)at the resolutionof 2-3 and 7-6 suspensions,a sixth and third respectivelyare created,indicatedon the example as 2-[3]6 and 7-[6]3. Althoughneitherof the uppertwo voices actuallycrosses over the other, each takes a turnas a 4-3 suspensionwith the bass in combinationFl. In addition to the cadential prefixion, fourth-speciestechnique also encompasses consonant syncopes, as was discussed on page 239. Only one basic type was cultivated,which involves stepwise ascent using 5-6 motion. The passage from the Frescobalditoccata shown in Example 11 suggests the reason for this restriction:to provide an ascendingcounterpart for the descending types based on cadentialprefixions. Consonant suspensions are fundamentallyindifferentto direction, since no downward pressureis exertedby the requirements dissonance resolution; for they can supportgeneral ascent or, as was shown in the passage from Josquin(Example9), generaldescent. Havingboth a wealth of descendadapting sourcesfor sequence and a newly inventedbut not particularly able source for ascent in the 2-3 braid,Corelli focused attentionon the basic type illustrated Example24. Gl, invariably voiced as shownwith in
G

3.7.4,m. 16;3.12.5,mm.6-8; 5.8.2,mm. 13-14

1
5 6

2
5 6 2.7.2, mm.22-3 (withGa);4.1.4, mm.25-6

3
14.5.4,mm.8 10

G
|3.12.3, mm.9-11; 4.7.3, mm.9-14, 17-21

4
Example 24. Class G sequence type (ascending):Consonant5-6 syncopes in uppervoices 253

Direction Type A B Descending C

Basic suspension Characteristic feature Example seventh-chords 14 7-6 Inessential 17 7-6 Essential seventh-chords 7-6 Double-length unit 18
2-3 Bass suspension 20

Ascending

E F G

2-3 2-3 5-6

Braid Alternating braid Consonant syncope Table 1

22 23 24

the top voice in syncopes, was inheriteddirectly from Merula,who cultivatedthis combinationextensively (Allsop 1999, 85). A not infrequent single-line expression,often found in the violin sonatas,op. 5, is Gal, in which the top voice leaps from the thirdof a | chord to the sixth of a 6, occasionally filling in with a passing note the connection to the next third.CombinationG2 is Corelli's main embellishment,which afforded him more rhythmicactivity than Gl. Frequentlyin fast movements, the dotted-halfnote of the example was furtherembellishedwith a complete lower neighbor.Corelli did not use G3 as a rule, although one (highly embellished)exampleis provided.However,an effect very similarto G3 is found in G'4, in which the formerlywhole-note voice of G moves in parallelwith the 5-6 half-notevoice. Because this motionwould produce a fourthif the bass remainedas in line 1, the bass is constrainedto move in orderto provideconsonantsupport.The uppervoices aretypically disline posed as in line b, which producesa morefluid, contrary-motion with motion found in the bass while preservingthe upper-voiceparallel-third many of Corelli's Rosalia sequences. A tabularsummaryof the seven strict-counterpoint prototypeclasses is presentedin Table 1, keyed by direction,letteredtype, motivatingsuspension, and characteristic feature. Rosalia and Aloysius Despite the differentorigins and developmentalpaths of the Rosalia and the fourth-speciessequence--epitomized in Examples 1(a) and (b) andthe associateddiscussion-these two types beganto be stronglyallied northern-Italian instrumental music, with the in late seventeenth-century result that distinguishingfeaturesof the two were mergedto variousextents. Startingfromthe propertyof pattern commonto both, transposition we can note essential similaritiesbetween the prototypesdiscussedin the of of last section andvariousserialtranspositions thematicpatterns, which 254

thatby second constitutesthe classic Rosalia. Corelli and his contemporariesexpandedthe purviewof the Rosalia to encompassnot only transposition by second but also transpositionby third,which was comparatively rare before this time. Combined with the possibilities of serial transposition fifth or fourth,a featureof imitativewritingsince the sixby teenth century,40 these transpositiontypes constituted a repertory of sequential techniques paired by direction and transpositionalinterval, which is shown in Table2. Two examples of each Rosalia type are given those with accordingto the numberof repetitionsof the thematicpattern; transposition) remindus thatRosaliasneed not form two patterns(pattern sequencesthemselves-something alreadysuggestedby Examples6 and 8. Three-pattern Rosalias (i.e., sequences) are also shown in all but one prototype case.41Each Rosalia type is matched to a strict-counterpoint based on intervalof patterntransposition.Differentkinds of foreground intervalshownin the table. figurationaffectthe analysisof transpositional The situation is illustratedin the relationshipbetween "Descending
fifth" and "Descending second by descending fifth."42Example 25 shows

a typical situation in which the descending motion by second (shown with brackets)is crossed with the fasterdescending-fifthratemarkedout by the beamed arrows. The figuration is not quite consistent, which emphasizesthe descending-secondmotion at the expense of the other.As we shall see, in Vivaldi's style, the two are often fully equalized. Example26 presentsa more complicatedcase thatalso illustratesone way by which the Rosalia and fourth-speciestypes were combined into a single entity. Measures 26-9 contain Rosalias in descent, the patterns of which are markedby the "R"brackets.These measuresas a whole do not form a standardsequence, since the intervalsof transpositionare not
Rosalia type
Ascending 4th

Examples (# of units)
4.6.1, mm. 21-4 (2)

Fourth-species type
B3b

= Descending 5th Ascending 3rd Ascending 2nd Descending 2nd

6.7.1, mm. 24-30 (3) 3.5.3 mm. 1-4 (2)

4.5.2, mm.13-19 (3)

E3
Ascending 2nd by Ascending 5th

4.9.4, mm. 5-10 (3)

F
Descending2nd by Descending5th

Descending 3rd 4th Descending

2.2.2, mm. 25-30 (3) 3.1.3, mm. 32-5 92)

B, Dl

1.73.,mm.35-40 (3)

C, D4 F

4.6.3,mm.6-8 (2) Table2

255

Concertino

Concerto Grosso

Example25. Corelli 6.12.2, mm. 5-7

Violin

Violone e Cembalo
6 6 5 6 6 7 6

30

Vln.

B.C.
7 7

Vln.

B.C.

Example26. Corelli 5.1.5, mm. 26-36 256

"in line": they differ from the first to the second pattern(B-A, descending second) and the second to the third (A-Ft, descending third);a pattern on G is omittedbecause of the difficulties it would have createdfor the upcomingmodulationto Ft minor.In the following phrase,mm. 31-5, the thematicmaterialof the previous Rosalias is deployed as figuration overa class-B sequence,the bass/rootnotesbeing indicatedby the beamed arrows.As in the previous example, the violin figurationbrings out the slow transpositionalintervalby descending second, which the brackets delimit. But the beamed arrowsshow thatthe bass-line figurationand its fast chord-rootrhythmare consistent as well. One detail missing from the passage in mm. 31-5 is the presence of the type-B suspensions and seventh chords-which are, after all, defining features of all fourth-speciestypes. While the bass and underlying voice leadingcould easily supportthem,the figuration,which uses a tonicizing lower neighborat the beginningandendingof the figure,is incompatible with a suspension since it would be cross-relatedto the lowerneighbornote. TheA in m. 32, for example,cannotbe suspendedthrough the B-minorchordof the measurebecauseof theAts on the thirdandsixth eighth notes (and so on throughthe rest of the sequence).As a result,the signals an important suspensionshave to be suppressed.This adjustment prototypesin orderto accommocompromiseof the strict-counterpoint date the comparativelyexpansivefigurationalpatternsof the Rosalia. One motivationfor making this compromise comes from Corelli's interestin pairingsequencesby directionas well as transpositional interval, a signaturefeatureof his sequencetechnique.Example27 illustrates the use of contrastingsequences sharingcommon figuration.Passage (b) G'b4, the figurationof which hides the underlying5-6 ligis a standard This passage is an answer atureandbringsout its Rosalia characteristics. to passage (a) from earlierin the movement,constructedfrom a basis in C2 butwithoutthe 2-3 suspensions.(The putativesuspensionsareshown for comparison.) Unlike the previous example, where the suspensions had to be removedto avoid cross relation,here theirremovalis motivated entirely by figurationalconsistency with the other sequence at (b).43By itself, passage (a) sounds normal with the suspensions-like, in other words, many other passages using C2 in Corelli's works-but the point of the contrastwith passage (b) would be lost if they were not suppressed. Anotherillustrationof contrastingpairedsequences (Example28) shows thatCorelli could be sensitive to chord-rootmotion as well as to general direction.Both the Fl and B-type sequences move by slow second, the first up and the second down. But they also move by fast fifths-again, the first up and the second down. The result is a perfectlybalancedpassage in which both slow and fast motions are contrasted.

257

(a)
Violin 1

Violin

Violone e Cembalo

(b)
Violin 1

Violin
Violone e Cembalo

Gmb4

GCob4

Gb4

Example 27. Corelli 3.12.3, 1-4, 9-13

Violin 2

Violone e Cembalo
b 6 n Fl p 6 B

Vln. 1

Vln. 2

B.C.
4 5 7 , 5 4 p , 71 5
4

Example 28. Corelli 4.2.2, mm. 10-15 258

Rosalia, Aloysius, and Arcangelo: Their Descendents AlthoughCorelli's workshave been a focus of this paper,they are not sonataduringthe last isolated from the developmentof the instrumental quarter the seventeenthcenturyby composes such as G. M. Bononcini, of G. Torelli,and G. B. Basani. Sequences in the works of these composers are little differentfrom Corelli's, so Corelli cannotbe creditedwith "inventing"the techniquesingle-handedly, even if his imaginativeand thornoteworthy. oughgoing use of the sequence was particularly Yet, from anotherperspective,he might have well been so credited, since Corelli is the first composer to have experiencedthe benefits of a modernprintingand distributionsystem (albeit in its infancy), and, as a result,he is the first composerto have achievedcontinentalfame during his lifetime and to have his works cited as models both in Italy and abroad.Perhapsthe previous composer who came closest to the kind of international fame Corelli enjoyed was Josquin.Significantly,both composers enjoyed a mutuallybeneficial relationshipwith a pioneeringpublisher;just as Josquinis associatedwith OttavianoPetrucci,so Corelli is with EstienneRoger.To be sure,the excellence of Corelli's music was its own best selling point, but without the efforts of Roger, Corelli's fame may have been much more limited and local. The great interest in Corelli's works created additionaldemand for music thatRoger was delightedto meet, prosimilarItalianinstrumental moting therebya kind of "ItalianInvasion," with Corelli playing the role of the Beatles. Since Roger had agentsin manymajorEuropeancities, he was able to distributehis carefully edited and high-qualityengravings widely, with the result that Corelli's music became the touchstone for an internationalstyle, being copied, imitated, and studied throughout Europe.44The well-known examination of passages from Corelli by Rameauis but one example;45 even in far-awayScotland,AlexanderMalcolm could recommend(1721, 426) that studentsstudy Corelli's works to learnthe basics of composition. was a stimulationto build from While one result of "Corellimania" Corelli's models-meaning that they were instrumentsboth of stylistic consolidationand of controlledstylistic change-another was the wholesale adoptionof effects initiallyassociatedwith Corellibutthatsoon came to be regardedas partof the public domain.The famous "Corelliclash," the resolutionof a 2-3 cadentialsuspension simultaneouswith an anticipationof the cadentialtonic note, was, as its name suggests, too associBut ated with Corelli to be so regarded.46 other techniques were easily into incorporated new styles, idioms, andgenres.The 2-3 braid,for example, was still being used a centurylater by Mozartin the opening of the Recordarefrom the Requiem.47 the most far-reachingincorporation But was certainlythe combinedRosalia-suspensionsequence. 259

The most well-known and importantearly adopterof this sequence was Antonio Vivaldi-significantly, another composer in the Estienne Roger stable.While thereis still debateaboutthe extentof possible influence of Corelli upon Vivaldi's style, the use of Corelliansequence techVivaldi'ssequencesgenerallyachievea greaterbalniqueis undeniable.48 ance between Rosalia and suspensionelements thando Corelli's, which, as Examples 25 and 26 suggested, favored Rosalia in most mixtures of the two. In addition,Vivaldi's sequences consistentlyavoid the extremes representedin Example 1: the unadornedcadentialprefixion of (a) and Rosalia of (b). Vivaldi'stechniquebenefitedfrom cadenthe butt-jointed tial-suspensiontechniqueto soften the seam between local patternsbut used extensive figurationsreminiscentof the Rosalia that slowed down harmonic rhythm and encouraged expansive thematic constructions. One favorite figurationalpatternwas the simple quick repeatednoteVivaldi's adaptation of the stile concitato that Monteverdi made famous-which offeredvoice-leadingandcontrapuntal clarityand exciting motor-rhythmssimultaneously.Vivaldi used sequences hardly any less than did Corelli, but along with his narrowingof sequence style to the via media of Rosalia and Aloysius, Vivaldi also narrowedthe repertory of Corellian sequence types. The descending cadential-suspension types of A and B predominate,with types C (descending third) and F (descending fourth)being nearly altogetherabsent, leaving him the 5-6 ligaturesof type G as the only ascendingsequencetype. Because of these restrictions,and because Vivaldi deploys (generally slower-moving)sequencesjust as often as does Corelli, the sense that sequence dominates Vivaldi's textures,perhapsoverly so, is a typical reactionto his style. A characteristic of passages is shown in Example29, from the last set and by movementof Vivaldi'sop. 3, no 11 (RV 565, transcribed arranged J. S. Bach as BWV 596. The thoroughbasssignaturesare the present author's,based on Vivaldi's originals.) The opening is a 2-3 braid, but with an ingenious use of an incessant pickup note, A4, that gives an The unclear status of any impression of multiple points of imitation.49 bass line, anotherVivaldianstylistic signature,makes identifying a particularaccompanyingline-1, 2, or 3 from Example 22-a moot point. The following passage, mm. 4-6, is clearly type Dl embellished with chromaticismin the bass (not unknownin Corelli's technique)as well as in the upperparts (unusualin Corelli), leading to consistent augmented fourthsin the 4 chordsand minorsixths in the following 6s. Both this and the previous passage show eighth-note versions of the stile concitato repeated-notefiguration.Measures 7-10 is a passage using B3b (every chorda seven-five-three) which the periodof the chordchangesis fast, in at the half measure,while the middle parttakes on a sixteenth-notefiguration whose period is the slow whole measure.In this it resembles the passage from Corelli's op. 6.12.2 shown in Example 25, but the empha260

3,

j, L)1

4r

h~

10

B3b

(4)

6)

Example 29. Vivaldi, op. 3, no. 11, mvt. 4, mm. 1-14 sis in the formerwas on the slow, Rosalia period,while in the Vivaldithe emphasis is on the fasterone. The final sequence of the set, mm. 12-13, is also the one thatBach had to adaptsomewhatfreely. Insteadof the sixteenth-noteconcitato figurationof the original, which would have been impossible to play on the organ,Bach arpeggiatedthroughthe chordsin sixteenths with occasional thirty-second flourishes, maintaining the energy of the original,while reassemblingthe brokenconcitato bass line pedal line. More significantly,he added anothersusinto a quarter-note pension line beneaththe original7-6, figuredas 5-4, which resultsin the progression4o. This passage clearly pairswith the secthicker-textured 261

ond, mm. 4-6, in its use of a chromaticdescendingbass from D3 and in its foundationin cadential-extension technique. While Vivaldi frequently pairs sequences for varied repetition, he pairedthemfor oppositionor contrastfarless frequentlythandid Corelli, which is why he neededfewer sequencetypes;a balancedrosterof ascents and descents was unnecessary.Even so, one outstandingcase of contrastingpairingis shownin Example30, fromthe fugal second movement of op. 3, no. 11 (RV 565/BWV 596). The answerand motivically related countersubjectare shown at (a). The second half forms a type-D (2-3) suspensionsequence,with the substitutionof a sixth for a thirdat the resorganizes olution (cf. Example 23). The figurationof the countersubject the sequenceinto Rosaliasrelatedby descendingthird,shownby the dotted brackets.The episode at (b) has materialin the outer staves that is Because the motivically relatedto the tail of the subject/countersubject. passage uses B2, an alternation seventhand common chords,it is easy of to emphasize the slow descending-secondRosalia movementfrom seventh to seventh chord, which Vivaldi does throughthe figurationshown in the second staff. The dissonant sevenths, however,provide agreeable mortising that smoothes over the seams between patterns.Later in the fugue, Vivaldireversesthe figuration'sdirectionand places it in another episode, in which the countersubject sounds in the middle of the texture. In attemptingto preservethe suspension series, Vivaldi runs into some trouble contrapuntally.The top two lines contain the 2-3 cadentialprefixion series, leaving the countersubjectto shadow the lower of the two in thirds. The inverted figuration,located in the bass, is left with strong-beated dissonancesof a ninthandfourththatresolve over a change of bass. However,the bass is also left with perfect fifths on each strong beat. Given the inheritedfigurational patternthatemphasizesstrong-beat arrivals,it is impossible not to notice a series of parallelfifths between of top andbass voices. PerhapsbecauseCorelli's"Affair the Fifths"ended ultimatelyin favor of Corelli (the contretempswas never a factor in the subsequentreceptionandcriticismof his music), no one apparently bothered Vivaldi aboutthis infelicity.51 In presentingexamples of Vivaldi's works that were transcribedby J. S. Bach, I have wantedto show the directtransmissionof the combined source in Corelli's works Rosalia-suspensionsequencefrom its apparent throughearly adaptorslike Vivaldiand then into the worksof othercomposers, most effectively by those who learnedit copying and transcribing model pieces. Certainlyanothersource of disseminationwas the various partimentimanualsthatcirculatedas late as the early nineteenthcentury, triowhich preservedidiomatic figurationsof the seventeenth-century sonatasequenceas compositionalschemas--even those, such as manyof the ascendingtype using suspensionbraids-that were stylisticallyobsolete afterVivaldi'stime. Example 31 shows a typical case from Fenaroli 262

(a)

3]6

[3]6 2

[3]6

[3]6

(b)
21

B2

(c)
34 5

9 6

Example 30. Vivaldi, op. 3, no. 11, mvt. 2 [1800] 1978: compare (a) with Example 22, sequence E2; and (b) with Example 23. Both exercises sound like trio-sonatapassages transcribed for keyboard,and may in fact have gained a place in thoroughbass pedagogy originally as practicefor accompanyingsuch common passages.52 They may have remainedafter newer composers displaced Corelli and his cohort, for the valuable, concentratedpracticethey affordedpreparcomposers who had ing and resolving dissonances. Eighteenth-century partimentotrainingand relied on its schemas thus may have unwittingly a learnedand transmitted basic Corelli idiom withoutknowing a note of his music. After Vivaldi embracedthe combined Rosalia-suspensionsequence with enthusiasmand made it one of the many attractiveand influential features of his style, few refinementsof the techniquetook place afterBach's sequencetechniqueowes muchto whathe learned wards;certainly, in transcribinghis Vivaldi models. The contrapuntaltechnique having been perfected, the main development of the sequence thereafterwas 263

(a) A

46

6 5

8 3

(b)

3 4

85 4

Example 31. Sequentialpassage exercised in Fenaroli 1978 [1800] (a) Fenaroli,Partimenti,p. 45; Example [V] (b) Fenaroli,Partimenti,p. 53; ExampleA.u. rhetoricalratherthan technical: sequences were used less routinely and with purposefulrestraintas transitional ratherthanexpositionaldevices. Whereas both Corelli's and Vivaldi's quick-tempo music seems constantly either to be playing out cadential formulas or breakinginto sequence, Bach and other northerncomposers developed other thematic techniques, and were able therebyto use sequence strategicallyinstead of habitually. The Rosalia-suspensionsequence maintaineda place in nineteenthcentury compositional technique. Brahms used B3b (|:), for example, quite frequently(recall its memorableeffect in the "Hille, wo ist dein Sieg!" passages in the sixth movementof Ein deutschesRequiem).Even Bruckner the "progressive" used the same prototypein the last movement upped of his Fourth Symphony (mm. 282-6). Liszt, characteristically, the voltage of this effect by heapingminorninths onto it in mm. 34-8 in "IIPenseroso"(fromthe Italianvolume of Annees de Pelerinage), a passage thatis otherwisestrikinglysimilarto Example 1(a). But, at the same time, the combinationalso became undonein the worksof Wagner, Liszt, Rosalia for and others of the "New GermanSchool," who rehabilitated new expressive purposes;the opening of the Tristanpreludeis perhaps the most famous example, the patternsof which are separated-successfully this time-by rests.53 The cadentialprefixion-indeed, any kind of preparedaccented dissonance that resolved down by step-was, in this if light, an antiquated venerablegestureat a time in which an unprepared accented dissonance could leap away to resolution or even to another 264

dissonance. Further,Rosalia fell victim once again to the old "spiritual poverty"criticism that Schubartvoiced as premiumswere increasingly placed on concise, original, non-formulaic,and non-repeatingmusical ideas--on "musicalprose,"as Schoenbergtermed it.54Under these circumstances,the Rosalia-Aloysius line eventuallydied out at the turnof the twentiethcentury.Having endowed the artof music with many wonderful passages and effects for two hundredyears, it passed peacefully into the historyof compositionaltechnique.

265

NOTES
I would like to thank Elizabeth Bonheim Bodek for her assistance in the early stages of research for this article. For their comments and suggestions, I also thank William Rothstein and Jennifer Williams Brown. 1. Ricci 2004, 1.1, provides a detailed history of definitions and associated terminology. 2. See Ricci 2004, 1.1.1, for various authors' views on the amount of repetition in a sequence. 3. Bass 1996 uses "projection" as a technical term differently from Hasty, but the coincidence is suggestive. 4. By "tonal formulas," Bukofzer means sequences, cadences, and functional chord progressions. 5. A shorthand will be used for all citations from Corelli's works: dots will separate opus, number, and movement. Thus, op. 3.6.1 is the first movement of op. 3, no. 6. In Chrysander'sedition of Corelli's works (reprintedby Dover), movements are separated by concluding double bar lines. 6. Allsop 1992 makes a useful distinction between the a2 and a3 textures that is pertinent to the present discussion. The crucial difference between the two is ". .. the role of the continuo which in the sonata a2 acts largely as a non-thematic harmonic support above which the trebles indulge in an interplay of figurations, and the elaboration or extension of thematic material remains throughout strictly their domain. In the sonata a3 on the other hand, each of the three melodic voices participates equally in the imitative texture while the continuo merely doubles the lowest part, often simplifying it in the manner of the basso seguente of the previous century" (26). Corelli's "trio" sonatas are of the a3 type, but Allsop's observations concerning the echo and polychoral effects of the Venetian a2 sonata are equally valid for the sonata a3. 7. Scholes 1959, 240. Spelling and punctuation are Burney's original, which the editor, Percy Scholes, did not update. I have, however, placed within parentheses a passage that was originally a footnote and have slightly altered the punctuation before and after the passage in question. 8. Burney visited Schubart in Ludwigsburg in mid-August, 1772. See Scholes 1959, 39-40. 9. Schubart's version (mistakenly?) has a C5 on the second beat of m. 1 instead of a sequentially expected Es. I have not been able to locate a source for "Roslia, mia cara" outside of Schubart. "VetterMichel" is given in full in Erk and Friedlaender 1880, 50. 10. Schubart's views are discussed at length in Moreno 1996, 336-44. For a more easily available overview of eighteenth- and nineteenth-century sequential repetition, see Moreno 2000. 11. Pfannkuch 1994 notes that the use of "Rosalia" to cover both ascending and descending stepwise motion is found in Castil-Blaze, Dictionaire de musique moderne (1828) and persists even in Eberhard Thiel's Sachworterbuch der Musik (1962). Schusterfleck appears to have been coined by Joseph Riepel in connection with his Monte phrase stereotype. See Riepel [1752-86] 1996, 39, 150. Riepel is also uncomfortable with the obvious simplicity of the technique, and recommends to readers that they learn to hide them by varying surface figuration (p. 349). Hein-

266

rich Chritsoph to Kochtook overRiepel'susage (withoutreference Monte).See Koch 1802,s.v. "Schusterfleck."

13. This transpositional especiallythe initialacendingstep, is typicalof the pattern, Englishvirginalpieces basedon popular tunes. to up 14. Thesefifthsmightbe understood be broken by intervals the secondbeats, on yet these intervals perfectoctaves,the successionof whichperhaps are savesthe situation not withoutprotestfrompurists. but of In to 15. Theangledbrackets indicatespecificordering themelodicintervals. order delaysthe altoentrance half a by createa moreexcitingseriesof attacks, Josquin but the restoring one-meameasure thenshortens initialnoteby half,thereby the sureinterval. 16. Cf. Josquin's celebrated Maria... virgoserena,in whicha seriesof ascendAve up ing 6-5-6-5-6-5 syncopesin mm.44-50 is twice transposed a step. of 17. See Rameau [1722] 1971,83 ft. Thetechnique cadential extension be found can [1477] 1961).See, for example,the formations in theworkof Tinctoris (Tinctoris
marked 3p in Examples 8 and 9 in The New Grove Dictionary of Music and Musi-

cians (SecondEdition),s.v. Counterpoint. this 18. Barnett, 1997, 191,explicitlyendorses connection Fuxianfourthspeciesto of
contrapunto legato.

has whichcurrently pop-music sincopeas "syncopation," 19.Manntranslates rhythmic connotations I preferto avoid.I havesubstituted that "syncope." 20. See Fux [1725] 1943,p. 61, m. 7 of fig. 77, for the two-part example,andp. 101, fourth-species m. 4 of fig. 146 for the three-part (I illustration. am leavingaside anexampleillustrating theoretical a point,fig 136, sinceit is notfreelycomposed butrather For examples,see p. 104, fig. 154, m. manufactured.) the fifth-species 9, andfig 155, m. 9. Thesetwo arebasedon the sameDoriancantusfirmusand formation, only difference arethe samecadential the stemming fromthe relative lines.The samecadential formation used positionsof thefirst-andfifth-species is herein Example10(b),m. 3. 21. See Fux [1725] 1943,p. 61, m. 7 of fig. 75 andm. 8 of fig. 77. In bothcases, the for as 4-3 can be understood a replacement 2-3/7-6. In fig. 75, a counterpoint cantus firmusthat containsan ascending over an idiosyncratic Hypophrygian afterthis contrapuntally octaveleap,the 4-3 occursimmediately stressfulevent; werethe leapA-A avoidedandthe followingcompensating descending stepwise the motionto G madeunnecessary, cantusfirmusline couldeasilybe madeto pass for through as a replacement G, makingthe suspension 7-6. In fig. a D upwards discussed previously. con77, the4-3 followsuponthe9-8 hapaxlegomenon The text for both suspensionsis againinfluencedby stressfuldisjunctevents in the Lydiancantusfirmus,which is unusually disjunctas a whole. In mm. 7-10, the cantus firmus leaps three times in succession:from F up to C, followed by motionsthrough to F. The 9-8 occursoverthe firstF, andthe 4-3 descending A overthe C. The cantusfirmusthusbehavesin thesemeasures morelike an added first-speciesbass line, with the "real"cantus firmusbeing implied above the line. fourth-species Thisimpliedline hasA andG overF andC respectively--creatinga seriesof 2-3 formations withthe fourth-species line. with thatof Schenker's "In Kontrapunkt: spite of the three22. This view contrasts counterpoint conceivedonly in relation voice texture,syncopesin three-voice are

267

to thelowestvoice"(1987, 2:83).See also 4(a),wherehe emphasizes "when that the syncopated voice is an inneror uppervoice, it is notappropriate distinguish to or to conceptually verbally addition its relationship thebassvoice also some in to syncope-relationship thethirdvoice"(ibid.,2:100-1). Schenker to wishes, further of course,to emphasize structure, whichis an Ursatzin his theoryof outer-voice musicalstructure. Hanson 23. In a surveyof three-and four-part sectionsin the Massesof Palestrina, of the dissonances 1983 corroborates generalfrequencydistribution suspended suggestedin the previousparagraph. Because the authortakes a thoroughbass approach analyzing to suspensions, connections fromthe4-3 and9-8 formathe structure not explicit. tionsto the 2-3/7-6 cadential are at 24. See Fux [1725] 1943, 60-1, fig. 73, wherethe cadentialapproach mm. 8-10 in containsthree7-6 suspensions a row. 25. Thatis, if each measureof the cantusfirmusis considered motivicunitof one a line to note,thenthefirst-species transposes according themotionsof theunit,just Example13 with Fux [1725] 1943 (60, as in a sequence.It is usefulto compare line fig. 73), wherealmostthe same fourth-species over the same Doriancantus firmusis used.Theonly difference betweenthemis themotivicrepetition D-C in mm. 5-6 thatis allowedin the three-part version,butwhichJosephuswantedto avoidas a "badrepetition" the two-part in counterpoint. in it 26. Becausethe bass is unspecified the 7-6 situation, is possiblefor the upperto some of the 7-6 formations voice suspensions be expressedas 2-3. However, in to be examinedhavethe suspension an uppervoice againstthe bass, in which as 2-3. case therecan be no voicingof the suspension an upper-voice 27. Two passagesusing an unusualvoice-exchangetechnique,which Corelli used only in op. 1, are shownin the boxedtext at Ala.The boxedexamplesaredrawn froma catalogof sequences Corelli'soeuvre,opp. 1-6, madeby theauthor; in the interested readeris invitedto consultthese to examinethe types of light embelsequences. Corelliuses with suspension lishments on 28. In theirextensivesurveyof sequencetechnique pp. 246-67, 524-32, Aldwell 1989 do not considerthe line-A'possibility, whatthey mightterm andSchachter deriva"descending stepwith voice-leading chords." a likely chromatic (Cf. 6 by betive usingaugmented-sixth chordsshownon p. 530.) This sequence,perhaps in sequencauseof Corelli'slackof interest it, did notenterthe common-practice Vivaldidid use theAA'combination occasion.An tial technique. on Interestingly, of examplewill be seen in Example29 below.Also, see the last movement op. 3, no. 3 (RV310; transcribed J.S. Bachas BWV 978), mm. 9-13. by 29. FromtheLudiMusiciof 1621.Thetitlein thecompleteworksandin thethematic Bergamas Angl.The Englishconneccatalogis Canzond 5 voc.Ad imitationem tion to which the title alludesmay be Giles Farnaby, whose "Rosasolis" (Fitzas williamVirginal material the CanBook #143) sharesthe samebasicthematic zonaBergamasca. 30. Therelationship betweenthetwo kindsis explicitlynotedin SalzerandSchachter 1969,206, andGauldin [1922] 1987,2:85-6, 2:101-2, is very 1997,320. Schenker sensitive thisdifference itspossibilities development freecomposition. to and for in of the voice andmakingit a bassvoice 31. Note thatby equalizing rhythm the "filler" in B 1, a previously ornamental escapetone froma middlevoice is promoted a to

268

role structural greater note, whichreducesout to leave thanthatof the downbeat 6 is a series of parallel| chordsbrokenup by downbeat chords.This promotion shownto be confirmed B2 andB3 as the structural is "castout"to the new in note bass line. 32. In 1.4.3, mm. 25-6, Corellifiguresa seventhon every otherchord,giving the in sequence formof B 12;thesecondviolinpart,however, ornamented a way the is to suggestB3. Even so, the thoroughbassist well-advised avoidaddingsevto is the enthswhereCorelliexplicitlydoes not call for themin orderto maintain clarity of the secondviolin line. promotion the formerescape 33. Note thatbothC2 andC2a confirmthe structural of in toneof Example14, mentioned note 31. lenientviewon this 34. Schenker [1922] 1987,2:110-11, offersa perhaps surprisingly circumstances a withthe aid of a few mitigating [Milderungsgrunde,legal matter "compulsion" theligature series,theurgetowards term]: comcontinuing of triadic placement. pleteness,andafterbeat Muchof this leniencyis a resultof privileging outer-voice in relationships mentioned note 22. 35. See SalzerandSchachter 1969,96, for a discussionof thesedangers. in fromtheletters 36. Theaffair discussed detail,withlengthy is through extracts which the controversy raged,in Allsop 1999, 35-40. For anotheraccount,see Arnold 1965,901-2. see 37. However, Example25 belowfor a counter-example the op. 6 Concertos. from Butanother overstrucwhereCorelli'sthoroughbass instance habitsfavorsurface tureis 1.7.1, mm. 24-5, whereoccursa hybridof A4 andB2. A 7-6 in the first violinresolveswithan anticipation beforethe changeof bassmakestheinterval a 5-4 tenth,butthe secondviolin,movingin parallel progresses before underthirds, thebasschangecreatesanoctave.Corelliis forcedto figurethe anticipations, and the two chordsof the sequencearefigured7 . "Of course, it is not possible to 38. Salzer and Schachter1969, 97, admonishes: in employthe f__progressions becauseof the fifths." mentioned As consecutively note 34, Schenker wouldallowtheseprogressions. 2:111,Ex. 193. See legal victoryoccursin the openingphraseof 4.11.2, 39. A perhaps morespectacular parallel to themselves the ear.Again,becausedifwhereprominent fifthspresent for basedon theletferentuppervoices areresponsible the effect,no prosecution terof the law couldbe successful. 40. Forexample,in Scheidt'sCanzona discussedon page 23 (Example Bergamasca, fromC to G, backto C, andthenfinally 16) above,a thematic blockis transposed to F beforea cadenceis madebackin G. See mm. 108-12. Stein 1994provides an techniquemotivic transpositional excellent overviewof seventeenth-century albeitin the worksof a single composer, She GiacomoCarissimi. notes (p. 127) of by the prevalence transpositional intervals fourthand fifth, the possibilityof of thoseby second,andthe rarity others. three-unit Rosaliasin Corelli'sworks. 41. Thereare no examplesof ascending-fifth Successiverootmotionby ascendingfifth is not uncommon, Corellideploys but seconds. by this withinunitsthataretransposed ascending for secondby determinology sequences(e.g., "descending 42. The double-interval of scendingfifth")is suggestedby the practice Laitz2003. practice reading supof in 43. Thefictivesuspensions Ex. 27 bringto mindRameau's

269

in pressedsuspensions Corelli'swork.See Lester1992,305-19, for a translation with commentary the well-knownpassagein the Nouveausystemein which of Rameau analyzespassagesfromop. 5. 44. SeeAllsop 1999, 153ft. foranappraisal Corelli'sprofound of influenceandreach, includingthe surprising information that Corelli'sworks were known even in China.Besides Roger'seditions,the influenceof piratededitions of Corelli's freely,he was also worksshouldnot be discounted. WhileRogerhimselfpirated of a victim,frequently London's JohnWalsh. 45. See note43 above. 46. MichaelTalbotpoints out that this device was "... an asperityfashionablein dancemusicaround1680,buthardlyappears [Corelli's] worksoutsideop. 2." in
(The New Grove Dictionary of Music and Musicians. Second Edition. Vol. 6. S.v.

of 460.) While it may not havebeen an innovation Corelli's, Corelli,Arcangelo, the fact thatit is associatedwith his nametestifiesto the extensiveinfluenceof Corelli'stechniques, or by even thosenot invented developed him.AlthoughI do not know who coinedthe term,it seems likely thatit was someonefrom a time when the only music from around1680 still esteemedand studiedwas that of Corelli. was sequencetechniques, canbe seen 47. Mozart fondof thisandrelated Corellian as in thelastmovement his StringQuartet, 387. Inmm.31-7 is anE3 type(susof K. transposition ascending third,a veryrare pensionbraid), by whichyieldsa pattern effect (combined effectin the late eighteenth An with century. evenmoreantique ombraatmostphere) foundat the beginningof the development, is a prevailing mm. 125 ff., wherean F-typesequence(overlpapping braidsandpattern transpositionby ascendingfifth)is deployed.This sequenceeffects a large-scale modulationfromG majorto Bb minor,risingby perfectfifth through variouspassing keys. 48. Cf. Selfridge-Field 1975, 179 ff., 234;Allsop 1999, 166-7. 49. Selfridge-Field 1975,236, cites thispassageas an exampleof Vivaldi's"fondness imitationin which a fourthin one partis answered a fifth in by for quasi-fugal another." 50. Vivaldi'soriginalvoicing of the 7-6 called for a fifth along with the seventh, which thenresolveddownwards a third.Cf. the passagefromScheidt'sCanto
zona Bergamasca shown in Example 16.

Corelliin 51. Talbot1992, 84, notesthat"Thetype of passagewhichonce embroiled disputewithcriticsin Bolognaoccursagainandagainin Vivaldi's an acrimonious compositions." 52. The firstappearance sequential of is literature roughly passagesin thepartimenti with contemporary the perfection the Rosalia-suspension of in technique instruthrough octave is in the mentalmusic. For example,the ascending5-6 pattern use 7-6 Gasparini [1708] 1963, 26; the descending (Al) is on p. 36. Gasparini's of thesepatterns significant and becausehe was a pupilof CorelliandPasquini, is wereactive his manual widelyknown.(ThatCorelli,Gasparini, Pasquini was and of in Rome-midway betweenBologna,the epicenter the laterseventeenth-centuryinstrumental pedagogywas especially sonata,andNaples,wherepartimento cultivated,is a particularly circumstance. Grampp Cf. suggestivegeographical 2004.) Whetherthese sequencesare intendedto illustrateanythingotherthan trio and chord-connection finger techniquein preparation accompanying for

270

sonatas and concerti is an interesting, open question, as is the relation of pedagogical material to artistic composition. 53. See Rothfarb 1991, 135-43 for Ernst Kurth's discussion of Rosalia sequence in Wagner. 54. See Dahlhaus 1980, 40-64, for an overview of the issue.

WORKS CITED
Aldwell, Edward, and Carl Schachter. 1989. Harmony and Voice Leading. Second ed. Orlando, FL: Harcourt Brace Jovanovich. Allsop, Peter. 1992. The Italian 'Trio'Sonata: From its Origins until Corelli. Oxford: Clarendon. . 1999. Arcangelo Corelli: New Orpheus of Our Times. Oxford: Oxford University Press. Arnold, F. T. 1965. TheArt of Acommpanimentfrom a Thorough-bass as Practiced in the XVIIth & XVIIIth Centuries. New York: Dover Publications. Barnett, Gregory Richard. 1997. "Musical Issues of the Late Seicento: Style, Social Function, and Theory in Emilian InstrumentalMusic." Ph.D. dissertation, Princeton University. Bass, Richard. 1996. "From Gretchen to Tristan: The Changing Role of Harmonic Sequences in the Nineteenth Century."19th Century Music 19: 263-85. Bukofzer, Manfred. 1947. Music in the Baroque Era: From Monteverdi to Bach. New York: Norton. Bumey, Charles. 1959. Tagebuch einer musikalischen Reise: durch Frankreich, Italien, Flandern, die Niederlande und am Rhein bis Wien, durch Bohmen, Sachsen, Brandenburg, Hamburg und Holland 1772-1773. Translatedby C. D. Ebeling and J. J. C. Bode. Kassel: Birenreiter. Dahlhaus, Carl. 1980. Between Romanticism and Modernism: Four Studies in the Music of the Later Nineteenth Century. Translatedby M. Whitall. Berkeley: University of California Press. Erk, Ludwig, and Max Friedlaender. 1880. Deutscher Liederschatz. Eine Auswahl der beliebtesten Volks-, Vaterlands-, Soldaten-, Jiger-, Studenten- und Weihnachts Liederfiir eine Singstimme mit Pianofortebegleitung. Leipzig: Peters. Fenaroli, Fedele. [1800] 1978. Partimenti ossia Basso numerato. Bologna : A. Forni. Fux, Johann Joseph. [1725] 1943. The Study of Counterpoint. Translatedby A. Mann. New York: Norton, 1943. Gasparini, Francesco. [1708] 1963. The Practical Harmonist at the Harpsichord. Translatedby F. S. Stillings. Edited by D. L. Burroughs. New Haven: Yale School of Music. Gauldin, Robert. 1997. Harmonic Practice in Tonal Music. New York: Norton. Grampp, Florian. 2004. "Partimenti: Musik fir Generalbass solo, II: Romische Quellen zur Partimento-Praxis."Concerto: Das Magazin fiir Alte Musik 21.194: 23-27. Hanson, John. 1983. "Enumerationof Dissonance in the Masses of Palestrina." College Music Symposium 23: 50-64. Hasty, Christopher. 1997. Meter as Rhythm. New York: Oxford University Press.

271

Jeppesen, Knud. 1970. The Style of Palestrina and the Dissonance. Translated by M. Hamerik. New York: Dover Publications. Jersild, Jorgen. 1982. "Harmonikse Sekvenser i Dur- og Mol-tidens funktionelle Harmonik." Dansk Arbog for Musikforskning. 13: 73-108. Koch, Heinrich Christoph. 1802. Musikalisches Lexikon, welches die Theoretische und praktische Tonkunst,encyclopddisch bearbeitet, alle alten und neuen Kiinstwiirter erkliirt, und die alten und neuen Instrumente beschrieben, enthalt. Offenbach am Mein: J. Andr6. Laitz, Steven. 2003. The Complete Musician: An Integrated Approach to Tonal Theory, Analysis, and Listening. New York: Oxford University Press. Lester, Joel. 1992. Compositional Theory in the Eighteenth Century. Cambridge, Mass.: HarvardUniversity Press. Malcolm, Alexander. 1721. A Treatise of Musick: Speculative, Practical, and Historical. Edinburgh:printed for the author. Mann, Alfred. 1958. The Study of Fugue. New Brunswick, N.J.: Rutgers University Press. Moreno, Jairo. 1996. "Theoretical Reception of the Sequence and Its Conceptual Implications." Ph.D. dissertation, Yale University. . 2000. "Challenging Views of Repetition: From Satzlehre to Melodielehre." Journal of Music Theory 44: 127-70. Pfannkuch, Wilhelm. 1994. Sequenz (Satztechnischer Begriff). In Die Musik in Geschichte und Gegenwart, edited by E Blume and L. Finscher. Kassel: Barenreiter. Piston, Walter. 1941. Harmony. New York: Norton. Rameau, Jean Philippe. [1722] 1971. Treatise on Harmony. Translatedby P. Gossett. New York: Dover Publications. Ricci, Adam. 2004. "A Theory of the Harmonic Sequence." Ph.D. dissertation, University of Rochester. Riepel, Joseph. [1752-86] 1996. Anfangsgriinde zur musikalischen Setzkunst. Edited by T. Emmerig. Vienna: Bohlau Verlag. Salzer, Felix, and Carl Schachter. 1969. Counterpoint in Composition: The Study of Voice Leading. New York: McGraw-Hill Book Co. Schenker, Heinrich. [1922] 1987. Counterpoint.Translatedby J. Rothgeb and J. Thym. New York: Schirmer Books. Scholes, Percy A., ed. 1959. An Eighteenth-Century Musical Tour in Central Europe and the Netherlands. Vol. 2, Dr Burney's Musical Tours in Europe. London: Oxford University Press. Schubart, Christian Friedrich Daniel. [1774] 1839. "Von den Rosalien." In Gesammelte Schriften und Schicksale. Stuttgart:I. Scheible. Selfridge-Field, Eleanor. 1975. VenetianInstrumentalMusic from Gabrieli to Vivaldi. New York: Praeger. Stein, Beverly Ann. 1994. "Between Key and Mode: Tonal Practice in the Music of Giacomo Carissimi." Ph.D. dissertation, Brandeis University. Talbot, Michael. 1992. Vivaldi. New York: Schirmer Books. Tinctoris, Johannes. [1477] 1961. The Art of Counterpoint (Liber De Arte Contrapuncti). Translatedby A. Seay. Rome: American Institute of Musicology.

272