Partimento practice appeared, almost unnoticed, at the end of the seventeenth century, grew up steadily during the eighteenth century, and disappeared (though not entirely) during the course of the nineteenth century. When Gustav Fellerer discovered some partimento manuscripts in the Santini collection in Münster around 1930, nobody in Germany knew about them. But in a city in southern Italy, once among the capitals of European music and since the unification of Italy deprived of most of its importance – Naples – partimenti were still in use. Thirty years after Fellerer's ‘rediscovery’, Jacopo Napoli, the director of the local conservatory (heir to the four eighteenth-century conservatori) published the last partimento collection belonging to a living tradition (Bassi della Scuola Napoletana, con esempi realizzati (Milan: Ricordi, 1959)).
However, Napoli's and Fellerer's work was by and large inconsequential, for partimenti fell into oblivion again. Only very recently have partimenti been rediscovered, by a number of scholars researching independently in different countries. Now there is a growing and enthusiastic community of partimentisti in both the old and the new worlds, and more and more musicians have become fascinated by the practice. The recent interest in partimenti has been one of the most engaging developments in modern music theory, and I wonder about its causes.
One reason is that partimenti fill some voids and address some needs. In the wake of Alfred Mann's seminal works (in particular, Theory and Practice: The Great Composers as Teachers and Students (New York: Norton, 1987)) there is an awareness that the ‘official’ history of compositional theory, largely based on German and French treatises, cannot account for the transmission of compositional craft. In other words, you cannot teach how to compose music by writing a book. Consider the history of compositional pedagogy, which in the eighteenth century was largely based on strict counterpoint in its many facets: simple, with species, invertible and imitative. There is a huge gap between strict counterpoint and free composition, one which Heinrich Schenker was acutely aware of; finding the bridge between these two realms became the most serious theoretical problem he had to solve. Besides strict counterpoint, eighteenth-century musicians often referred to thoroughbass as the ‘true fundament’ of compositional craft. In J. S. Bach's words, ‘General Bass is the most complete foundation of music’ (Gründlicher Unterricht des General-Basses, chapter 2: Von der Definition, in Beiträge zur Generalbass- und Satzlehre, Kontrapunktstudien, Skizzen und Entwürfe, ed. Peter Wollny, Johann Sebastian Bach: Neue Ausgabe Sämtlicher Werke, Supplement (Kassel: Bärenreiter 2011), 5). Bach's opinion has always puzzled me. To be sure, thoroughbass is important; it is one of the hallmarks of what we call ‘baroque music’. But it was essentially an accompanimental technique, and ‘accompaniment’ means something that ‘comes with’ – a side order to the main course, so to speak. The rediscovery of partimento practice helped me to understand better the actual role of thoroughbass in eighteenth-century theory. Thoroughbass was not only a convenient accompanimental practice, but – under the clothing of partimento – was also the main conduit for the transmission of compositional craft.
The recent interest in partimenti may also be a symptom of a dramatic change in the field of music analysis since the mid–2000s. Until the last decade of the twentieth century, music analysis was primarily a North American discipline, focused on musical works belonging to the Austro-German canon and dominated by a mere handful of theoretical models. Most transatlantic exchanges in the field of music analysis flowed in one direction only: from America to Europe. The pre-twentieth-century European (and particularly German) theoretical legacies were revered by historians of music theory, but in a field separate from music analysis. A deeper integration of history and analysis (the much sought-after ‘native’ analytical methods) was invoked as a distant goal. The idea that European and American analysts could develop a new theory together, deeply rooted in history and with strong heuristic potential, could scarcely be imagined. And no one expected that this new theory would pose a challenge to the most foundational pillars of music analysis: the concepts of unity, coherence, authorship and the ‘work’ itself. In truth, this multifaceted ‘new theory’ is not a unified one and has taken on different names. Robert Gjerdingen's schemata (Music in the Galant Style (New York: Oxford University Press, 2007)), German Satzmodelle (see the special issue of the Zeitschrift der Gesellschaft für Musiktheorie, 4/1–2 (2007)), the Austrian partitura tradition (Felix Diergarten, ‘“The True Fundamentals of Composition”: Haydn's Partimento Counterpoint’, Eighteenth-Century Music 8/1 (2011), 53–75) and Italian partimenti (my The Art of Partimento: History, Theory, and Practice (New York: Oxford University Press, 2012)) are all aspects of this trend, and each one represents a part of the whole. In fact, schemata, Satzmodelle, partitura, partimento, cadential formulas, scales, the Rule of the Octave, bass motions and invertible-counterpoint patterns are all different aspects of the native eighteenth-century compositional craft.
The assumption of these theories is that during the eighteenth century a large number of pre-compositional schemata, comparable to building-blocks, were consciously used by most composers, and that many of them were passed down to the following generations, well into the nineteenth century. I say ‘consciously’ because young composers were taught schemata – particularly in Italy – although there is hardly a trace of them in the long and complex treatises that have so far been the favourite sources for historians of music theory. Rather, schemata in their purest form abound in manuscript collections of composition exercises, such as partimenti, regole and solfeggi developed at the four conservatories of Naples.
What does this kind of analysis tell us? If we turn our attention to what I have called ‘traditional’ analysis, we behold a very different landscape. The keyword – actually, the obsession – of late twentieth-century analysis was ‘unity’ (whether organic or not), a term sometimes combined with the less cogent word ‘coherence’. Unity was so important because there was a strong belief in the uniqueness and individuality of the piece of music under scrutiny. And, in turn, the clearly defined identity of a piece of music (or better, of a ‘musical work of art’) was a token of indisputable authorship.
Many of these beliefs are still operating today, even if most analysts are reluctant to raise the flag of organic unity (too much postmodernism, too much deconstructionism has passed under the bridge since the roaring seventies). Still, nobody would deny that, say, the Allemande of Bach's French Suite in C minor is a strongly coherent and meaningful piece of music, that it is a work of art endowed with personality and individuality and, finally, that it is a manifestation of the creative mind of Bach. But it uses the same building-blocks used by generations of composers, before and after Bach, who operated in exactly the same ways. Of course, there is nothing new in this. Leon Battista Alberti built the Tempio Malatestiano in Rimini using the same architectural elements described by Vitruvius some fourteen centuries before: the same arches, columns, capitals and pilasters. Yet the Tempio Malatestiano is one of the greatest masterpieces of the early Italian Renaissance.
We are also witnessing a general crisis in the performing model rooted in the ‘musical-work’ paradigm, based on the concept of Werktreue and operating according to the parameters established by Liszt during the mid-nineteenth century. Performers increasingly seek to escape from the confines of ritualized performance, as in the case of the Russian violinist Viktoria Mullova and her recent engagement with Gypsy music. However, the rigidity of classical training for performers – especially for those versed in classical and romantic repertory – creates an almost insurmountable barrier to improvisation and freedom. This last point, the challenge to the Werktreue model, seems to me the most intriguing parallel development to the partimento renaissance.
Challenging the Werktreue ideology has more far-reaching consequences than improvising a cadenza or supplying bel canto embellishments during a reprise. As a classical pianist myself, I have always been painfully aware of the restrictions imposed by the ‘musical-work’ paradigm on our performing approach. Some of us cannot even test an instrument in a piano showroom without playing other people's music; just improvising on a simple cadential progression seems out of the question. This problem has nothing to do with one's ability at the keyboard. In my partimento classes I have met brilliant concert pianists who could play Rachmaninoff's Third Piano Concerto with ease. But when asked to play the easiest of Fenaroli's partimenti they faltered in the realization, their fingers stiff with anxiety. Can we reasonably ask them to improvise a cadenza in a Mozart concerto?
In a recent video produced by the Department of Tourism in the Marche region of central Italy, American actor Dustin Hoffmann recited the poem L'infinito by the Italian poet Giacomo Leopardi, in the original language and from memory (the reader can watch the video at <www.youtube.com/watch?v=F6PNvmZs1gE>). Hoffman does not know Italian, and had to memorize the pronunciation and the meaning of every single word. The result had a certain alien quality, though not disagreable to those Italians familiar with the poem since childhood. But this was obviously an experiment, a deliberate provocation. Something similar happens in musical performances when one gets the impression that the performer's keyboard skills exceed her understanding of the musical language.
This problem is obviously well known, and was pointed out by the fathers of music analysis a century ago. Analysing the music we perform will no doubt deepen our understanding of it and strengthen our confidence in its delivery. Still, analysis is an intellectual activity; its benefits tend to remain at the surface of the performer's consciousness. In order to ‘speak’ an eighteenth-century musical language, we need something going much deeper, down to the level of what we call ‘musical instinct’. This was the exact purpose of the rigorous partimento training the Neapolitan maestros imposed on their novices: to make music a native language.
Improvising a complete piece of music using a partimento as a guide represents a middle ground between performing an existing piece and improvising ‘from scratch’. One does not have complete freedom, but one does not risk going astray either. However, the amount of freedom allotted to the performer is very high, and more importantly, one cannot play a partimento verbatim, as in a poor performance of a fully written piece. The performer's creative intervention is mandatory. The step from partimento to free improvisation is significant, but not huge. Once you have memorized a good number of schemata, you can recall them and create your own mental partimento. With time, the operation will become easier and almost unconscious (as far as I know, this is the method followed by the Forschungsgruppe Basel für Improvisation (FBI), led by Rudolf Lutz). I would say that the problem of improvisation is particularly acute for performers specializing in classical and early romantic repertories because – in contrast to their baroque colleagues – the idea of improvisation is much more alien to them. The importance for performers of improvisation in the classical repertory has been demonstrated by Robert Levin, among others. But it is not yet reflected in course syllabuses for the majority of music students, except in a few niches (particularly in fortepiano classes).
Now, what does a modern performer gain from partimento practice, and how can she use partimento fluency in her performance? First of all, practising partimenti is a way to return to performers what they have been robbed of: music. A glance at the history of piano treatises from the late eighteenth century to the early twentieth century is revelatory (compare, for example, those by C. P. E. Bach and Daniel Gottlob Türk with Attilio Brugnoli's Dinamica Pianistica (Milan: Ricordi, 1926)). The earlier treatises focus on music – expression, style, embellishments, fingerings and improvisation; the later ones are increasingly concerned with anatomy – bones, muscles, finger motions, wrist rotation and the like. Music was gradually removed from the pianist's syllabus, beginning with improvisation. When even fingering disappeared, the transformation of pianists into perinde ac cadaver executors was complete.
Very well, one might argue: let us teach pianists how to improvise a short prelude, a modulation or a sequence. But then? Audiences still want to listen to works of art such as Beethoven sonatas, not a performer's whimsical extemporization. Or should we contemplate introducing partimenti into the classical canon? Frankly, I think partimenti will remain what they were two centuries ago: a special kind of art reservata. But even if we are unlikely to find partimenti in a future concert programme, their introduction into performance classes will greatly enhance musical fluency.
Finally, growing awareness of the pervasiveness of musical building-blocks in eighteenth-century music poses a significant threat to the idealistic concept of authoriality. What is left of the musical-work paradigm when so many pieces of music share the same compositional elements, even if applied with varying degrees of skill and (fortunately) yielding varying aesthetic results? Finally, what happens to authorship when, as was the case in eighteenth-century Naples, a strong, supra-individual and long-lasting tradition led many composers to write works so similar that musicologists must analyse copyists' handwriting to distinguish a work by Pergolesi from one by Durante?
By way of example, let us consider a typical combination of schemata that was taught in Naples during the eighteenth century. It is a model of triple counterpoint that can also be used in double counterpoint, with the exclusion of one voice. Basically, the scheme comprises a do-re-mi (element B), a matching do-si-do with syncopation (element A) and a faster element (C) that, when appearing in the bass – as often happens – produces the cadential progression 
, perhaps the most common instance of what the Neapolitans called ‘cadenza lunga’. Example 1 shows all six possible permutations of this scheme, though in practice permutations e and f were seldom used.
Example 1 Three elements in triple invertible counterpoint, in six permutations
My next example shows how these schemata were taught. It is a selection of partimenti by different Neapolitan teachers, spanning from the early eighteenth century (Gaetano Greco) to the early nineteenth century (Fenaroli). Example 2a shows element C in isolation, repeated in sequence. In Example 2b, C appears after B, meaning that in the realization of this partimento the right hand will play B while the left hand plays C, and vice versa. The same thing happens in all the remaining examples except for 2f, where A and B join together to the exclusion of C.
Example 2 (a) Gaetano Greco, I-Nc 45.1.65 c. 100v (Gj deest); (b) Leonardo Leo, Gj 1700 (several sources, among them I-PESc Rari Ms.c.12); (c) Nicola Sala, Gj 2015 (several sources, among them I-Nc 46.1.34); (d) Nicola Sala, Gj 2005 (several sources, among them I-Nc 46.1.34); (e) Francesco Durante (?), No. 56 from I-Nc Roche A.5.6 (Gj deest); (f) Nicola Sala, Gj 2056 (several sources, among them I-Nc 46.1.34); (g) Fedele Fenaroli, No. 8 from book 4 of Partimenti ossia Basso Numerato (Florence: Gio. Canti, c1850; facsimile reprint Bologna: Forni, 1978). Gj numbers follow the cataloguing system used by Robert Gjerdingen on his website Monuments of Partimenti <http://faculty-web.at.northwestern.edu/music/gjerdingen/partimenti/index.htm>. Used by permission
With my third example we are in the realm of ‘real’ music. Example 3a is the beginning of Pergolesi's Salve regina (one of the works securely attributable to the composer). This music is made of the same contrapuntal fabric as my previous examples: a do-re-mi, a syncopated do-si-do and a faster ‘cadenza lunga’. The same combination, in the same order, is also used at the beginning of another sacred piece by Pergolesi, the Laudate pueri (Example 3b). Yet the two pieces could not be more different: one is a Largo in minor mode, and the other is an Allegro in major mode. Pergolesi was not the only composer to exploit this component, nor indeed was he the first. A celebrated composer of the previous generation, Arcangelo Corelli, established many schemata that would later become standards. The triple-counterpoint schema appears in several of his works, such as the Allemanda of the Sonata a tre Op. 4 No. 11 (Example 3c) and the first Allegro of the Concerto Grosso Op. 6 No. 8 (Example 3d: the celebrated ‘Christmas Concerto’).
Example 3 (a) Pergolesi, Salve regina (Vespro della Beata Vergine, ed. Malcolm Bruno (Kassel: Bärenreiter, 2007)); (b) Pergolesi, Laudate pueri (Vespro della Beata Vergine); (c) Corelli, Sonata a 3 Op. 4 No. 11, Allemanda (Corelli, Sonata da camera a tré … Opera quarta, ed. Joseph Joachim and Friedrich Chrysander (London: Augener, 1888–1891; reprinted New York: Dover, 1992)); (d) Corelli, Concerto Grosso Op. 6 No. 8, Allegro (Corelli, Concerto grosso, op. 6, Nr. 8 (Weinachtskonzert), ed. Waldemar Woehl (Leipzig: Peters, 1937)). Used by permission
The idea that a piece of music may grow out of the gradual refinement of a handful of partimento schemata is comparable to the way in which an orator works out his speech according to classical rhetoric. As Roman authors such as Cicero and Quintilianus argued, a successful oration should include five stages of elaboration: inventio, dispositio, elocutio, memoria and pronunciatio (invention, arrangement, style, memory and delivery). The choice of partimento patterns might be compared to invention, not in the sense that the composer creates the patterns but in the etymological meaning that he ‘finds’ them (inventio essentially means ‘discovery’). Dispositio corresponds to assembling the patterns, carefully connecting them so as to ensure continuity and coherence. Through elocutio the crude succession of patterns becomes a flowing bass line, the right-hand part begins to take a more definite shape and the piece assumes its final design through additional figurations (the two remaining stages, memoria and pronunciatio, concern delivery or performance).
Application to music of the five stages of elaboration has a long tradition, and it was well established in the eighteenth century. As Laurence Dreyfus has brilliantly argued in a captivating book (Bach and the Patterns of Invention (Cambridge, MA: Harvard University Press, 1996)), Bach's music can be understood using Mattheson's adaptation of the stages as analytical and critical tools. However, neither Mattheson nor Dreyfus considers bass schemata to be part of inventio; for them, the compositional process is based on distinctive motivic ideas, not on overarching and almost featureless bass schemata.
However evocative, parallels between rhetorical elaboration and the composition of music from a handful of partimento schemata need not be taken literally. Whether or not the rhetorical rubric can be applied successfully to all eighteenth-century compositions, in part or in whole, the period gave rise to what became an established tradition of developing complete compositions from underlying patterns. Simple schemata, such as those developed for accompanying a scale, served as points of departure for an impressively large number of compositions spanning more than two centuries.
In conclusion, the discovery of a repertoire of pre-compositional building-blocks should lead us to rethink traditional notions of organic unity and coherence, authorship and the identity of the ‘musical work of art’. But it does not rule out these concepts entirely. Coherence grows out of the sequence of components (inventio) and their elaboration; the composer's individual nature expresses itself in the compositional paths he decides to follow among a huge number of possibilities; and a work of art's identity lies less in a hypothetical Grundgestalt than in a coherent succession of compositional choices.
Every analytical category employed here has its roots in historical sources – not only treatises, but also (and especially) manuscripts of compositional exercises, partimenti and solfeggi. Collectively, schemata theories are perhaps the closest approximation to a ‘native’ theory we have had so far.
