The relationships between music history, music theory, and composition at times seem so tenuous today that it is easy to forget that those three areas of specialization have split off from one another only recently, and that the practice of one can hardly be carried out successfully without some interpenetration with the other two. The discipline of music theory as we now understand it arose in the eighteenth century in a number of treatises that served both as composition manuals and as vehicles of theoretical speculation and, on occasion, polemic. The most famous of the eighteenth-century theorists is, of course, Jean-Philippe Rameau, who also—a fact sometimes overlooked by historians of theory—was one of the best composers of his generation and the long-awaited successor to Lully as master of French opera and ballet. Rameau's conception of the fundamental bass as the generator of harmony and harmonic progression seems to have been a genuine innovation, and one which furnished the basis for most of the important theory, above all that of harmonic function, which followed in the course of the next century and a half.1 Yet as Heinrich Schenker realized, Rameau represents only one side, and in Schenker's view the less incisive side, of eighteenth-century theory.2 Representing the other side are the German writers on figured bass and related topics, above all C.P.E. Bach, whose writings on continuo realization and improvisation, though only marginally dealing with composition as such, were admired by Haydn, Mozart and Beethoven, and were regarded by Schenker as symbolic of the purest and most profound strands in eighteenth-century compositional thought.
Bach's Versuch über die wahre Art das Clavier zu spielen, particularly the final chapter on the improvisation of keyboard fantasias, represents an approach to composition and improvisation which Bach seems to have shared with other members of the Berlin and North German schools of composers, including J.S. Bach and his other sons and students.3 Johann Philipp Kirnberger presented a similar conception of composition in his Kunst des reinen Satzes, which also includes theoretical discussions that seem to have been intended to adapt Rameau's theory of harmony to the more contrapuntally conceived harmony of the Bach circles.4 In their practical precepts, Bach's Versuch and Kirnberger's Kunst belong to a tradition that seems to have first been expressed in the Handleitung zur Variation of Friedrich Erhardt Niedt, the second volume of a three-part composition manual.5 It was here that the idea of composition as the elaboration of a figured bass line—the basis of both C.P.E. Bach's essay on improvisation and the practical treatment of composition in Kirnberger—seems first to have been explicitly stated. While variation sets and even entire suites composed over ground basses had been known for well over one hundred years, Niedt seems to have been the first theorist to teach composition specifically as the Veränderung of a bass line, and his demonstration of a complete suite of dances as the composing-out of an original line of figured bass has no exact parallel in earlier theory or practice.
Thus two theories of composition, representing two distinctive attitudes toward harmony, were first articulated in the early decades of the eighteenth century. Both assume the priority of harmony in a general sense to other elements of music, but for Rameau this was harmony in a sense somewhat like the modern one of functional chord progression, while for the Germans it was harmony as the harmonic/contrapuntal Satz that Schenker would later rediscover as the basis of composition (and analysis). Rameau was quite conscious of what he saw as the revolutionary, scientific nature of his theory, and his Traité is organized along the lines of the great theoretical compendia of the past, particularly that of Zarlino,6 with an initial philosophical and mathematical section followed by separate sections on practical matters such as composition and figured bass realization, each understood as independent arts arising as corollaries of the first principles demonstrated in the initial section. Niedt also writes within an established encyclopedic tradition, but it is a more pragmatic one best known from the writings of his immediate predecessor Printz and his famous successor Johann Mattheson.7 This is a tradition of practical manuals written in an engaging, even colloquial, style, designed as introductions to the skills required of church musicians, especially organists, who necessarily doubled as composers. Niedt's importance seems not to have been previously noticed; I have argued elsewhere that his manner of teaching composition seems to have resembled that of J.S. Bach and his followers.8 My concern here, however, is to investigate what preceded the apparent innovations of the eighteenth century; what were the theoretical conceptions and the actual practices of composers prior to the adoption in the eighteenth century of theory and pedagogy in which harmony (of one sort or another) precedes melody and counterpoint? This is a question that bears not only on the history of theory but on the history of compositional practice and its relationship to musical style. Indeed, as I will suggest, consideration of how composition may have been understood in the seventeenth century raises problems with generally held views on stylistic history in music and, in particular, on the definition of the musical Baroque.
In the treatises of the seventeenth century, harmony—that is, figured bass realization—seems never to have been understood as the basis of composition. In apparent disregard for the fundamental changes in musical style that occurred around 1600, theorists throughout the seventeenth century continued to take two-part species counterpoint as the basis of composition. Chromaticism, florid embellishment, the musico-rhetorical figures, and other typical characteristics of seventeenth-century music, were treated as so many embellishments to simple lines composed one-by-one against a cantus firmus in the tenor or bass.9 Figured bass realization, if considered at all, was treated as an ancillary skill, though of course an important one, particularly for keyboard players. Yet not only in its almost universal employment of a basso continuo part, but also in such characteristic genres as the strophic aria, variation-suite, and in recitative, a great deal of seventeenth-century music seems to proceed from the conception of musical structure represented by figured bass realization, not species counterpoint. Were the theorists of the seventeenth century simply one hundred years behind practical musicians in their understanding of how music was conceived and played? Is the new-found basis of composition in harmony in eighteenth-century writings symptomatic of a sudden change in compositional practice and musical style, or was the delay in recognizing the priority of harmony a product of the reluctance of teachers to accept current practice and abandon the venerable doctrines of sixteenth-century counterpoint? The latter continued to survive, though generally in a somewhat moribund form, in the stile antico, to which composers as late as J.S. Bach and Handel dedicated important efforts. Yet the use of independent basso continuo parts in virtually all seventeenth-century ensemble music—including many works in stile antico—suggests that in practice music was being understood by composers and performers alike in ways different from what contemporary writings would suggest. To consider the actual state of things will require a review of how figured bass was used and what it stood for in seventeenth-century music.
Early continuo parts fall into at least two categories. On the one hand are the so-called bassi seguente of works in the stile antico, which evolved as a shorthand employed in place of the tablatures or scores which keyboard players used in accompanying sixteenth-century polyphony. On the other hand are the continuo parts found in the new genres of the stile moderno, the independent, compositionally essential bass lines of recitative, sonata, and other forms. The ideal accompaniment in stile antico seems to have been the verbatim doubling of all the vocal parts; scoring the work not only assured that all the voices would be accurately doubled but supplied the missing parts that might not be sung when, for example, a polyphonic work was used as a vehicle for solo display by a virtuoso singer or player embellishing one of the original parts.
Example 1. Diego Ortiz, Recercada prima (with intabulation of Pierre Sandrin, Doulce memoire), from Trattado de glosas (Rome, 1553)
This sort of performance was the object of treatises by Ortiz, in the mid-sixteenth century,10 and by later writers; it furnished an opportunity for solo performance long before composers seem to have thought of adopting such a texture as the basis of an original composition—an idea to which I shall return.
Quite clearly, the doubling of the parts in a polyphonic composition would have produced a continuo realization, if one can call it that, which followed the same rules of counterpoint as the original. And in principle this seems to have been what keyboard players were expected to do well into the seventeenth century, at least when accompanying polyphonic church music at the organ. Early seventeenth-century sets of parts often include an organ part ranging in form from an unfigured basso seguente to a full score; most common, however, is a reduced score showing the bass and one upper part. This tradition originated in Germany but quickly spread throughout European regions under Italian influence. Even in England such reduced scores may have been added to Elizabethan consort music and used customarily in seventeenth-century consort performances, but the parts were not always prepared by the original composer and tend to show much wider variants between sources than do the other parts.11 The English practice may represent a special case, but there can be little doubt that the trend was universally away from the ideal of the continuo as the literal doubling of the parts. Such doubling, however, could not have been the actual accompaniment even in most sixteenth-century music; the crossing of voices, spacings encompassing more than an octave, and similar exigencies of Renaissance style would have obligated the organist to depart more or less radically from the literal doubling. Few sources on early continuo realization provide even a suggestion as to how organists understood what they were actually doing when accompanying in this manner. Certainly there are few theoretical sources (if any) before Rameau that articulate the type of harmonic patterns and relationships which organists and other players of chordal instruments must have been perceiving, if only intuitively, even in the sixteenth century. Yet from certain sources we can deduce the existence of traditions of continuo realization that owed very little to the strict style, even prior to the adoption of figured bass in the seventeenth century.
The independent basso continuo parts of the new Baroque genres have been regarded as a development from the basso seguente, but the full explanation may be more complicated. There seems to have been a long established tradition of improvised monody in Italy, in which a singer would accompany himself on the lira da braccio or other chordal string instrument. This practice evidently resembled falsobordone and other types of accompanied plainsong, including varieties for solo voice with lute or keyboard.12 In this music the realization of the accompaniment—not necessarily notated, of course—must have been relatively free; rather than including the doubling of the solo part, it presumably furnished a discreet accompaniment idiomatic to the instruments on which it was played. It may well have been the merger of this tradition with that of the organ bass that led to the nuove musiche of Caccini and other monodists, as well as subsequent instrumental types that mimic the declamatory style of monody. As is well known, Caccini and his contemporaries often seem to indicate the realization of the continuo in specific registers through the use of so-called compound figures (e.g. 10 for 3, 11 for 4).
Example 2. Giulio Caccini, Aria di Romanesca: Ahi dispietato Amor, from Le nuove musiche (Florence, 1601)
Particularly at cadences, the figured bass serves as a genuine tablature, indicating specific pitches and specific voice-leading, not pitch-classes that in principle could be placed in any register above the bass. That this is a borrowing from the tradition of the continuo as a doubling of the upper parts becomes evident in the choruses of early opera; in some cases the continuo in these choruses uses the same type of figured bass to prescribe literal doubling of the vocal parts.13
Regardless of its function or type, however, the role of the bass line in directing harmonic progressions or establishing a harmonic rhythm is at best ambiguous; the bass is not a generating part, as in later Baroque music, nor does it articulate a metrical rhythm defined by the simultaneous motion of all the voices in recurring patterns—that is, accented chords—except in special genres derived from dances.14 The latter, of course, grow in importance over the course of the seventeenth century, but at first, and particularly in stile antico, the bass is more important as a melodic line in its own right than as a generator of chord progressions. There are some exceptions, apart from the simple dance songs which have long been recognized as foreshadowing later types of harmony. The bass functions in a genuinely harmonic way in the strophic arie and instrumental variations over ground basses of these genres is if anything clearer in the sixteenth-century settings; the very florid embellishment and contrapuntal elaboration which they sometimes receive in the Baroque, particularly in keyboard settings, often seem to be the products of a conscious effort to avoid the more obvious harmonic implications of the ground bass. In other genres, in both stile antico and stile moderno, one finds: linear basses, each note of which bears a triad in root position; modal chord progressions, that is, successions of sonorities that are not built upon any of the recurring formulas characteristic of later basses; and long passages or even entire pieces for bass alone, which, unlike continuo solos in the later Baroque, seem to have been composed with relatively little thought to the harmonies that such a line might bear.
Example 3. Girolamo Frescobaldi, Canzon I a Basso solo, from Il primo libro delle canzoni (Venice, 1634)
Each of the preceding cases poses special problems for the continuo player—at least for the modern one, trained in eighteenth-century techniques and accustomed to eighteenth-century bass lines. In the absence of a tradition of improvisation centered specifically on the strict realization of figured bass, such as existed in Germany in the eighteenth century, players in the early Baroque seem to have had two choices: to improvise in a contrapuntal fashion dictated strictly by consideration of voice-leading, or to ignore the rules of counterpoint and play an idiomatic accompaniment suited to one's instrument. From the sources—which will be discussed presently—one might conclude that two such mutually exclusive approaches to figured bass realization existed throughout the seventeenth century. One was a continuation of the ancient tradition of improvised counterpoint, the other being based on the intuitive recognition of chords and harmonic progressions. The former must have been the one considered appropriate to church music, although there are signs that in practice it was not always the one actually employed.
But if musicians in Italy and France were customarily applying harmonic intuitions in their improvisations, seventeenth-century writers on composition and figured bass realization give little suggestion of it. If they treat of basso continuo at all, it is as a practical application of the principles of strict counterpoint, as it was for Lorenzo Penna in the middle of the seventeenth century: "Counterpoint is the theory of music; figured bass realization (suonare sù la parte) at the organ is its practice."15 Such writers evidently devoted little thought to composition or performance in the free genres, even though the most learned of them all, Anathasius Kircher, gives extensive discussion and quotation of recitative when his attention turns to musical expression.16 Adriano Banchieri, whose famous treatise on organ playing L'organo suonarino went through several editions in the earlier part of the seventeenth century, used figured bass in teaching "an easy method of improvising (sonar di fantasia)," printing an organ mass and several other pieces in a partial score which the player was to complete by realizing the figured bass.17
Example 4. Adriano Banchieri, Qui tollis peccata mundi Suscipe con Tremolo, from L'organo suonarino (Appendice, Venice, 1638)
The same pedagogic device occurs as late as 1697 in a treatise by Speer, who—like Banchieri—continues to apply this method to pieces that would seem to call for a free style of playing: toccatas, bizarrie, etc.18 Yet, to the extent that either writer describes the actual technique of figured bass realization, neither goes beyond the description of single consonant sonorities or the addition of voices one at a time above the bass—the approach to counterpoint found in the treatises of the sixteenth century.19
Could this really have been how continuo players conceived their realizations? From the few surviving practical documents concerning continuo playing in early monody and other theatrical forms, it seems clear that the strict contrapuntal style of realization was, as I suggested earlier, out of the question. A few sources, for example, give versions of Caccini's solo arie with written-out lute parts full of the forbidden progressions that one might, in fact, have expected in such settings.20
Example 5. a) Caccini, Udite, udite, amante, from Le nuove musiche; b) Accompaniment for lute, from B Br Cod. 704
Similar progressions occur in brief keyboard realizations offered by two early writers, Agazzari and Praetorius, although these make reference, at least, to a more genuinely polyphonic style of playing.21
Example 6. Michael Praetorius, example of continuo realization for organ, from Syntagma musicum (Wolfenbüttel, 1619)
The same sorts of "crudities" would, in any case, have arisen when a number of players simultaneously embellished the bass or one of the upper parts, or added improvised counterpoint to the notated texture. Such practices seem to have been common in early opera performances.22 While there seem to be no surviving examples of fully "realized" opera scores from the period—not that one would expect to find them—a peripheral source such as Thomas Morley's Consort Lessons may reveal something of the nature of contemporary ensemble performances of the theatrical type. The voice-leading in these settings, which like other publications of Morley's may have been intended to educate the English in Italian practices, is far from strict, both within and between the parts played by the chiefly chordal instruments of the ensemble.23
From such evidence it seems that a distinction was made at the very beginning of the Baroque to correspond with the distinction noted by Heinichen early in the eighteenth century between strict and "theatrical" styles of continuo realization. For Heinichen the theatrical style was essentially a modification of strict style through the granting of certain licenses, for example in the maintenance of a given number of voices or in the resolution of certain dissonances.24 Perhaps something similar held true in the earlier period, with players in the stile moderno abandoning some of the strictures that applied in old-fashioned church music. Yet the boundaries between the styles could not have been maintained without some intermingling. Penna complains about organists who accompany church music without knowing counterpoint.25 His complaint echos one made by Frescobaldi, who had deplored the tendency of keyboard players to avoid use of the four-part open score notation which Frescobaldi had employed in his Capricci and other contrapuntal keyboard works.26 The consequences of this tendency can be seen in the repertory of anonymous keyboard pieces preserved in seventeenth-century Italian manuscripts. These anonymous canzone and capricci, while imitating the outward forms of Frescobaldi's works in the same genres, are written in two-staff keyboard score and rarely achieve genuine four-part counterpoint, often settling for the mere suggestion of polyphony and occasionally omitting inner voices.27 One can imagine how the users of these manuscripts would have realized the exercises found in Banchieri's treatise. Even Penna, however, gives most of his examples of continuo realization in just three parts; thus it seems likely that the light textures of seventeenth-century keyboard manuscripts come closer to representing usual practice than the few prints from the period, which tend to be stern exercises in stile antico. Particularly in Italy, the chief function of the organist appears to have been to grace the liturgy with improvised or quasi-improvised works of the sort illustrated in the organ manuals; in such a tradition it would appear that habits applied in improvising figured bass would carry over into the improvisation and composition of keyboard music.28
Later continuo treatises from France and Italy suggest how musicians of those countries may have actually understood figured bass realization through much of the Baroque. The traditions preserved in these sources, rather than the stricter approach favored by the Germans of the eighteenth century, were quite possibly the more common in actual practice;29 if so, the implications go beyond a few details in the performance practice of the continuo part. The treatises of the French and Italians, unlike those of the Germans, usually consist largely of musical examples illustrating the realization of typical chords and progressions. While the examples avoid outright parallelisms and other forbidden progressions, the manner of notation often suggests a careless attitude toward voice-leading, and the verbal descriptions, if any, pay greater attention to the motion of the hands than that of the parts. For example, French and Italian sources in the eighteenth century regularly indicate use of what we would call the second inversion of the dominant-seven chord (V
). In similar passages few of the Germans would countenance the inclusion of the fourth of the chord, which strictly speaking is an unprepared dissonance when used in this way.30
Example 7a. J.F. Dandrieu, Table pour s'éxercer sur l'Acord de la Petite Sixte, From Principes de l'accompagnement (Paris, c. 1719). In the original the realization is indicated solely by the arrangement of the figures in the chord signatures.
Example 7b. C.P.E. Bach, realization of the six-chord with major sixth and minor third, from Versuch, II (Berlin, 1762)
In France this chord went by the name petite sixte,31 a reference apparently not to the size of the interval, which is a major sixth, but to the fact that the hand is relatively unextended in playing this chord when the sixth is placed in the upper voice. Given the differences in approach between the Germans and their southern counterparts in the eighteenth century, it is not surprising that the modern theory of harmony emerged in France, with Rameau, and met considerable resistance from some of the northerners. The Ramellian theory of harmony could have arisen only within a tradition of figured bass realization which depended already on the intuitive recognition of chords and chord progressions, which at times were emphasized over lines and intervals. Indeed, Rameau's doctrines of chord inversion and supposition are implicit in the rules given by Michel de Saint-Lambert, the most important French writer on figured bass realization before Rameau, for recognizing and fingering the chords of the seventh and ninth32—chords which, by contrast with the German tradition, seem to have been taken normally by French and Italian musicians as stacks of thirds, rather than as carefully prepared and resolved dissonances. One effect of this is perhaps the interest of French composers such as Rameau and François Couperin in textures and sonorities that are unusually rich in just these harmonies.
Yet, if musicians clearly anticipated the theorists and discovered both harmonically conceived chord progressions and harmonically generated melodic material, the conceptual framework for the composer or improviser, at least through the seventeenth century, seems to have continued to be species counterpoint and the melodic embellishment of the voices in the resulting polyphonic texture. What marked the Baroque, especially the early Baroque, as a distinct and integrated style was its eclecticism in practice: the acceptance of genres and techniques previously regarded as strictly improvisational—impossible, unnecessary, or undeserving of notation—as legitimate objects of the composer's concern. Hence the appearance of dances and similar music as something more than occasional "local color" in a madrigal or keyboard fantasia; hence also the importance of ornamental melodic figures in theoretical discussions of composition as well as of performance practice. In examining Baroque music, particularly that of the early seventeenth century, I would emphasize its points of contact with the Renaissance tradition, of which it is in many respects a continuation. For example, in the early Baroque solo sonatas of G.B. Fontana, the bass frequently introduces points of imitation in long note-values that are embellished when the soloist plays the fugal answer. The result, when the continuo is realized in a reasonably strict, straightforward manner, is a texture not terribly different from that which arose in certain sixteenth-century embellishing practices mentioned earlier.
Example 8. G.B. Fontana, Sonata II, from Sonate a 1, 2, 3 . . . (Venice, 1641)
Beneath the Baroque surface of such works is a style of counterpoint, as well as an approach to rhythm and phrasing, which remains close to that of the previous style.
The idea of a Baroque style in music stems from Hugo Riemann's discovery of a "thoroughbass period," lasting from roughly 1600 to 1750, in which virtually all ensemble works employ a basso continuo.33 The difficulties with this view of the style have long been recognized; in an attempt to rectify it Manfred Bukofzer and others pointed to other important defining characteristics, such as the stylistic heterogeneity—the consciousness of separate strict and free styles alluded to in the preceding—which also distinguish the Baroque from the previous style.34 But apart from the reliance on a basso continuo and on a highly stylized form of recitative in some vocal genres, the music of the early eighteenth century has practically no technical elements of great importance that would link it uniquely with the music of one hundred years earlier. This is particularly true with regard to musical form: by the early eighteenth century the form of most pieces is dictated by modulations through various keys and back to the tonic, in short, by a full-fledged system of tonality which, in the absence of lasting modulations, can hardly be said to have existed in the early Baroque. Hence late Baroque music has a hierarchic structure in which various levels of the work are governed by the same harmonic/contrapuntal relationships: a momentous development which far overshadows the common use of a figured bass.
Recognition of this dilemma has led at least one observer, Claude Palisca, to abandon the search for a definition of "Baroque" on technical grounds.35 Yet the alternative favored by him, to define the style on the basis of its expressive aspirations and devices, leads to equally daunting problems. A solution might begin with the recognition that our current habit of defining "style periods" on the basis of a few putatively universal "style characteristics," such as use or non-use of a basso continuo, is not a useful way of understanding music history, particularly when the divisions between periods have been arrived at by reference to the history of the fine arts rather than that of music itself. Indeed, the term "Baroque" seems first to have been applied to music after the realization that a certain musical repertory, like certain repertories of painting or architecture, depends on the ornamentation and the exaggerated emphasis of certain elements of late sixteenth-century style. In each case the special features of the "Baroque" style were employed in order to fulfill aspirations toward a heightened expressivity.
With this definition one might come up with a relatively concise list of musical devices—all employed toward fairly specific expressive ends—which characterize "Baroque" music. Yet it would be necessary to specify as well certain additional features of the "Baroque" music, for such devices as virtuoso embellishment, symbolic or iconic text expression, and free chromaticism, occur not only in the music of the seventeenth century but through the eighteenth and well into the nineteenth century. Only that music composed through the diminution of species counterpoint, without a more deeply hierarchic, that is, tonal, large-scale structure might be accepted as "Baroque." The music that meets these criteria—the toccatas of Frescobaldi, the violin sonatas of Marini and Biber, the operas of Monteverdi and his immediate successors—does seem to represent a parallel, coinciding roughly in technique, in expressive aspiration, and (not the least) in chronology with the "Baroque" works of other art forms.
But to understand what we mean when we extend the label "Baroque" beyond this relatively limited repertory, we might incorporate into our history of music insights from twentieth-century theory about the hierarchic nature of musical structure in general. I do not mean to suggest that we should accept Schenker's naively historicist view of stylistic evolution, according to which all music before Bach and Handel was wandering in search of the Ursatz.36 Yet there is at least an inkling of truth in Schenker's intuition—disregarded by some of his followers—that prior to those composers music lacked the deeply hierarchic structure which makes strictly reductive analysis of eighteenth- and nineteenth-century music seem attractive to some.37 Schenker's basic insight—that musical structures are composed of interrelated levels—suggests that definitions of style must take note of the depth and degree of consistency in the hierarchic structures characteristic of works in a given style. Moreover, a particular period might be characterized more by a processive change in its typical structure—for example, a gradually deepening structural hierarchy—rather than by fixed stylistic quantities.38
As a designation of style, the word "Baroque" might be restricted to the ornate, rhetorical, occasionally bizarre surfaces of many seventeenth- and eighteenth-century works; or it might be taken as referring to those works of this period that, while showing these characteristics, lack any structure deeper than that of the sixteenth-century contrapuntal framework that they embellish: there are no specifically "Baroque" principles governing large musical structure. On the other hand, as a designation of a period of music history, the term might be used with special reference to the period during which the familiar hierarchic conception of musical structure emerged—a process which continued through the course of the eighteenth century and which had among its products the articulation by theorists of the principle of the harmonic generation of melody, counterpoint, and even musical form. Of course, this raises the question of where the "Baroque" ends and the "Classical" begins, or whether such a distinction should be drawn at all. I would argue that the tonally organized structures of the late Baroque are manifestations of the same impulse that eventually produced what we call the Classical style; indeed, that Vivaldi, Bach, Handel, and their contemporaries were in fact employing a traditional vocabulary of Baroque affective devices within an essentially classical—that is, tonal and hierarchic—conception of form. Palisca sees their language as a formalization of the original Baroque style—affective gestures have become stylized cliches39—but I would rather understand this as the incorporation of Baroque gestures into a new type of structure, one that has more to do with Haydn and Mozart than with Monteverdi and Frescobaldi. Of course, this process alters the expressive value of the rhetorical figures, word-painting, and other Baroque expressive devices. They become subsumed into architectural forms and ultimately (in the Viennese Classical style) into dramatic ones.40 By then, however, the Baroque rhetorical habit of embodying expressive representation in iconic figures has largely given way to representation through musical processes, such as the extended crescendo or prolonged dominant preparation, operating at deeper structural levels.41
If we are to continue to understand the music history of the seventeenth and eighteenth centuries as a succession from Baroque to Classical style, we must recognize that these terms do not represent unified styles, that is, integrated conceptions of compositional technique and expressive aspiration. Contradictory conceptions of compositional theory may have existed simultaneously in both periods; and to a certain degree there is a period, roughly from 1700 to 1775, during which individual works can quite properly be understood as manifesting characteristics of both styles. Jan LaRue has used the concept of "style stratification" for some instances in which works manifest this situation.42 But his "strata," as I understand them, are very different from the levels of structure to which I have referred here; in any case, LaRue seems to regard style stratification as a somewhat regrettable characteristic of music composed in so-called transition periods, not a normative aspect of music history as I am suggesting. The very notion of style itself obviously needs some focussing; perhaps we should become comfortable with the idea of transition as a common, if not normative, state for musical style viewed as a historical phenomenon. This might confuse the simple categories which we have used to understand composers, compositions, history, and even theories of music, but it might help meet some of the objections to which traditional music history is increasingly vulnerable.
1Rameau's central work, the Traité de l'harmonie réduite a ses principes naturels (Paris, 1722), is published in facsimile in Jean-Philippe Rameau: Complete Theoretical Works, ed. Erwin R. Jacobi, vol. 1 (American Institute of Musicology, 1967), and has been translated, with introduction and notes, by Philip Gossett (New York: Dover, 1971).
2Schenker attacks Rameau, in the process presenting a brief treatise on the history of counterpoint, in "Rameau oder Beethoven? Erstarrung oder geistiges Leben in der Musik," Die Meisterwerk in der Musik, 3 vols. (Munich: Drei Masken, 1925-30; facs. ed., Hildesheim: Olms, 1974), III, 11-18. The essay is translated by Sylvan Kalib in "Thirteen Essays from the Three Yearbooks Die Meisterwerk in der Musik by Heinrich Schenker: An Annotated Translation" (Ph.D. diss., 3 vols., Northwestern University, 1973), II, 492-507.
3Bach's Versuch, 2 vols. (Berlin, 1753-62) is published in a facsimile ed. by Lothar Hoffmann-Erbrecht (Leipzig: Breitkopf und Härtel, 1976), and in translation by William J. Mitchell (New York: Norton, 1949).
4The Kunst des reinen Satzes, 4 vols. (Berlin and Königsberg, 1776-79) is available in facsimile (Hildesheim: Olms, 1968). The first two volumes have been translated as The Art of Strict Musical Composition by David Beach and Jurgen Thym, with introduction and explanatory notes by David Beach (New Haven: Yale University Press, 1982).
5The second volume of Niedt's Musicalische Handleitung, 3 vols. (Hamburg, 1700, 1706, 1717) seemed sufficiently important to Johann Mattheson that the latter re-issued it in an expanded and "verbessert" edition (Hamburg, 1721).
6I.e. Le istitutioni harmoniche (Venice, 1558; facs. ed., New York: Broude Bros., 1965).
7Wolfgang Caspar Printz, Phrynis Mitilenaeus, oder Satyrischer Componist (Quedlinburg, 1676-79), and Johann Mattheson, Der vollkommene Capellmeister (Hamburg, 1739; facs. ed., Kassel: Bärenreiter, 1954). The latter has been translated by Ernest C. Harriss as Johann Mattheson's vollkommene Capellmeister: a Revised Translation with Critical Commentary (Ann Arbor: UMI Research Press, 1981).
8I have discussed Niedt's work and that of his successors in "Composition as Variation: Inquiries into the Compositional Procedures of the Bach Circle of Composers," Current Musicology No. 33 (1982), 57-87.
9This is perhaps most clearly illustrated in Christoph Bernhard's Tractatus compositionis augmentatus (ms., c. l650-64), which has been translated by Walter Hisle in The Music Forum III (1973), 31-179. Bernhard discusses "variations"—the ornamental melodic motives basic to Baroque style and recognized as such by Niedt—among the special devices of the second practice (stylus luxurians communis), p. 96 in the Hisle translation, even though by Bernhard's day even strict church music depended heavily on passaggi and other Baroque elements.
10Trattado de glosas (Rome, 1553), facs. ed. by Max Schneider, 3d ed. (Kassel: Bärenreiter, 1961). Howard Mayer Brown describes such practices in Sixteenth-Century Instrumentation: The Music for the Florentine Intermedii (American Institute of Musicology, 1973), pp. 22-23.
11See, for example, the introductory comments by Donald Peart to his edition of the six-part consorts of John Jenkins (Musica britannica, XXXIX, 1977), pp. xix-xx, and by Richard Charteris in his edition of John Coprario's "fantasia-suites" (Musica britannica, XLVI, 1980), pp. xx-xxii.
12See John Bettley, "North Italian Falsobordone and its Relevence to the early Stile Recitativo," Proceedings of the Royal Musical Association 103 (1976), 1-18, and Murray Bradshaw, The Falsobordone (American Institute of Musicology, 1978). Bradshaw is perhaps too eager to see many early Baroque genres, especially the keyboard intonazione and toccata, as outgrowths specifically of the falsobordone tradition (see pp. 77-78). But it can be no accident that monodic psalmody showing varying degrees of melodic embellishment of the psalm tones figures prominently in such early Baroque collections as Viadana's Cento concerti ecclesiastici (Venice, 1602; works for one voice and continuo ed. Claudio Gallico [Kassel: Bärenreiter, 1964]), and Adriano Banchieri's L'organo suonarino (Venice, 1605; for modern reprint, see note 17 below).
13F.T. Arnold gives a telling example from Cavalieri's Rappresentatione di Anima e di Corpo (1600), showing the successive figures 
and 
over the repeated note G in the continuo. The two figures reflect the exchange of the notes B-flat and D in different registers between soprano and tenor. See The Art of Accompaniment from a Thorough-Bass As Practised in the XVIIth and XVIIIth Centuries (London: Oxford University Press, 1931), p. 62.
14The relationship between the ascendancy of accentual rhythm, as in the frottola and other dance-songs of the sixteenth century, and the advent of tonality has been suggested by several writers, including Edward A. Lowinsky, Tonality and Atonality in Sixteenth-Century Music (Berkeley: University of California Press, 1962), and Carl Dahlhaus, Untersuchungen über die Entstehung der harmonischen Tonalität (Kassel: Bärenreiter, 1968), p. 93.
15Li primi albori musicali (Bologna, 1679), pp. 134-35.
16Athanasii Kircheri . . . Musurgia universalis, 2 vols. (Rome, 1650), I, 660-65.
17This is found chiefly in the posthumously published Appendice . . . dove il novello Organista sopra due parte Acute, e grave pratica un modo facile in sonar di fantasia (Venice, 1638), one of several supplements added to the original treatise in later editions. All are included in the facsimile edition with introduction by Giulio Cattin, translated by Peter Williams (Bologna: Forni, 1969). Williams unfortunately renders part of the title of the appendix as "in two parts" (my emphasis), thus obscuring the key element in the title. The exercises in the appendix are clearly meant to supplement the four capricci that appear on pp. 106-09 of the original text, where they are preceded by this remark on p. 105: "Gli seguenti quatro Capricci servono per suonare doppò il Magnificat, spartito [sic] sopra due parti con gli riempimenti a beneplacito de gl'Organisti." The famous Battaglia, first printed in the supplement to the 1611 edition, is presented like the capricci on two staves, but presumably demands a more homophonic filling-out of the inner parts. Neither the capricci nor the Battaglia contains figures, although there are one or two accidentals over the bass of the capricci referring to inner voices.
18Daniel Speer, Grund-richtiger kurtz- leicht- und nothiger jetzt wol-vermehrter Unterricht der musicalishen Kunst . . . (Ulm, 1697). The earlier, shorter, and more crudely printed 1687 version of this work includes only one complete piece, a prelude notated in two parts with a very few indications of inner voices through figured bass. Perhaps representing a continuation of this tradition are two fantasias and fughettas of uncertain attribution, listed as BWV 907 and 908, notated in a comparable incomplete fashion.
19Banchieri describes figured bass realization as the addition of real voices one at a time over the bass, in the Dialogo musicale, the 1611 supplement to L'organo suonarino. Speer merely describes the notes called for by certain figured bass signatures and illustrates a few problems of voice-leading in two parts (1697 ed., pp. 180-87).
20See John Walter Hill, "Realized continuo accompaniments from Florence c1600," Early Music XI (1983), 194-208. Hill's characterization of the manuscripts discussed in the article is open to question; they might represent free entabulations of the lower parts of a polyphonic setting of certain arias, though a far more straightforward one than the polyphonic re-workings found in the Secondo libro de'madrigali a quattro voci, opera X (Venice, 1610) of Pietro Maria Marsolo, ed. Lorenzo Bianconi (Musiche rinascimentali siciliani, IV, 1973). The nature of the sources certainly points to their being hastily written-out realizations of Caccini's basses; moreover, Caccini, unlike John Dowland, is not known to have left any polyphonic settings of his own solo songs, as H. Wiley Hitchcock notes in his article "Caccini," The New Grove Dictionary of Music and Musicians (1980), III, p. 579.
21See Agostino Agazzari, Del sonare sopra'l basso con tutti li stromenti e dell'uso loro nel conserto (Siena, 1607; partially translated in Arnold, pp. 67-74), and Michael Praetorius, Syntagma musicum, 3 vols. (Wolfenbüttel, 1619; facs. ed., Kassel: Bärenreiter, 1958), III, 144. Arnold, p. 85, reminds the reader that early rules for thoroughbass realization, such as those in Viadana's Cento concerti ecclesiastici, expressly permit such parallelisms as occur in Example 6.
22See Agazzari and the article on his treatise by Gloria Rose, "Agazzari and the Improvising Orchestra," Journal of the American Musicological Society 18 (1965), 382, as well as Howard M. Brown's edition of Peri's Euridice (Madison: A-R Editions, 1981).
23Morley's First Book of Consort Lessons (London, 1599) has been edited by Sidney Beck, using various procedures to replace the lost lute part of the original (New York: C.F. Peters, 1959).
24Heinichen's Der General-Bass in der Composition (Hamburg, 1728) includes chapters on the "theatrical" treatment of dissonances and other liberties allowed in continuo realization outside of the church style.
25Li primi albori, p. 134.
26Frescobaldi's comments are in his preface to the Capricci . . . published in 1624 and available in several modern editions. On other early objections to both figured and unfigured basso continuo, see Arnold, pp. 80-81.
27I refer here to the works published in Girolamo Frescobaldi: Keyboard Compositions preserved in Manuscripts, ed. W.R. Shindle (Corpus of Early Keyboard Music, XXX, 1968). According to Alexander Silbiger, these are virtually all misattributed; see "Italian Manuscript Sources of Seventeenth Century Keyboard Music" (Ph.D. diss., Brandeis University, 1976), p. 254.
28Silbiger, p. 50, observes that the keyboard genres such as the toccata and intonazione probably drew on "a common fund of improvisational techniques and strategies at the surface level."
29The very strict realization, preserved by Kirnberger, for the Trio-Sonata from Bach's Musicalische Opfer (in the Bach-Gesellschaft edition, vol. 31/2, pp. 52-58), is probably more a harmony exercise than a documentation of actual practice. But there can be little question that the detailed discussions of niceties of voice-leading found in the eighteenth-century German treatises indicate a special attitude toward continuo realization in practice.
30The chord is realized as V in Jean-François Dandrieu, Principes de l'accompagnement du clavecin (Paris, c. 1719; facs. ed., Geneva: Minkoff Reprint, 1972) and in Rameau's treatises; both realizations are mentioned in Couperin's manuscript Regle pour l'accompagnement, ed. Maurice Cauchie (François Couperin: Ouevres complétes, I, 1933), p. 13. The Austrian organist Georg Muffat indicates the realization without the fourth as the usual one, but includes among his examples of chords of the fourth the "Quarta italica, die welsche oder irregular Quart," which is the same as the petite sixte, although evidently limited to use in certain contexts. See his Regulae concentum partiturae (ms., 1699), ed. by Helmut Federhofer as An Essay on Thoroughbass (American Institute of Musicology, 1961), pp. 41, 48.
31The term was apparently applied to all 
chords, but in practice the chord with dominant function was by far the most common. In the Code de musique practique (Paris, 1760; facs. ed. by Erwin R. Jacobi in Complete Theoretical Works, IV, 1969), Rameau writes of this and other names for the V7 and its inversions: ". . . ne nous occupons de ces différens nams que lorsqu'il s'agira des chiffres, encore même la méchanique les fait-elle trouver presque toûjours sous les doigts sans qu'on y pense" (p. 27).
32Michel de Saint-Lambert, Nouveau traité de l'accompagnement . . . (Paris, 1707; facs. ed., Geneva: Minkoff, 1972). Philip Gossett called attention to Rameau's borrowings, especially from Saint-Lambert, in the preface to his translation of the Traité (pp. xiii-xv). Though Rameau in effect plagiarized substantial passages from Saint-Lambert's work, it seems unlikely that Rameau did so because he was "struck by part of Saint-Lambert's discussion" of what was probably a generally received way of understanding the chords. I would imagine that the explanation was that the topic of figured bass realization—in France, at least, an essentially intuitive process, as the previous note suggests—gave Rameau little inspiration for exercising any original thought.
33As Claude Palisca notes in his article on "Baroque" in The New Grove, II, p. 173, Riemann avoided use of the term "baroque" for works of the thoroughbass period, but subsequent writers drew analogies between that music and the art and architecture of the same period, particularly as described by Heinrich Wölfflin.
34See, for example, Bukofzer's tabulation of opposed characteristics of Renaissance and Baroque style, respectively, in Music in the Baroque Era (New York: Norton, 1947), p. 16.
35Palisca, in New Grove, II, p. 175, rejects the possibility of defining the Baroque on the basis of technical features, although he accepts their usefulness in distinguishing the two styles that meet chronologically around the year 1600. He holds, however, that the only stylistic characteristic found throughout the Baroque period as usually understood is a common "attitude toward affective expression."
36Schenker's view of history is presented in outline in "Geschichte der Tonkunst," Der Tonwille II (1922), 3-4.
37The analyses of medieval and Renaissance music published in various issues of The Music Forum or in Leonel Salzer's Structural Hearing (New York: Dover, 1952) seem too ready to find Ursätze in music that lacks functional harmonic progressions and true bass lines, among other necessary criteria.
38Leonard Meyer, in Music, The Arts and Ideas (Chicago: University of Chicago Press, 1967), especially in the chapter on style changes, pp. 104-33, seems to suggest a relation between the hierarchic nature of art music and its capacity for development. But he presents a general hypothesis about style change (based on information theory), making few references to any particular stylistic development in music history. Meyer's book is sympathetically criticized by Leo Treitler in "The Present as History," Perspectives of New Music VII/2 (1969), 42-57.
39New Grove, II, p. 178.
40The familiar dichotomy between the architectural "Baroque" and the dramatic "Classical" conceptions of musical form seems to stem from Tovey, although a cursory examination of his writings revealed no explicit discussion on these lines; see Joseph Kerman, "Tovey's Beethoven," in Beethoven Studies 2, ed. Alan Tyson (London: Oxford University Press, 1977), pp. 182-84. As Kerman suggests, the description is valid only if one attaches a certain limited meaning to "architecture" and "drama."
41I cannot agree with Bukofzer, who first raised this issue, that "modern music," that is the music of the Classical and Romantic styles, achieves its expressive effects directly, "without the mediation of the intellect," as he states in "Allegory in Baroque Music," Journal of the Warburg and Courtauld Institutes III (1939-40), 21. The intellect is as much (or as little) involved in the perception of a musical process as of a concise motive, and in either case the listener more or less consciously translates the perceived sign into an extra-musical referent.
42Guidelines for Style Analysis (New York: Norton, 1970), p. 121.