article 23:   The Earliest Monothematic Variation Ricercar

Manfred F. Bukofzer’s “Music in the Baroque Era” says the variation ricercar was “established by Frescobaldi”. Indeed, the terms “variation ricercar” and “variation canzona” have usually been associated with his works in these genres where each variation of a single theme has its own separate section; the Capricci carry the technique even farther. But his first publication, the outrageously virtuosic (in the contrapuntal sense) Fantasias of 1608 are a throwback to an earlier type which is at least equally deserving of the name “variation ricercar”.

These are through-composed pieces where individual subjects are varied or developed – melodically, or rhythmically, or by using various devices of counterpoint (including inganno). Each internal thematic variation usually gets its own treatment as the piece goes along. Credit for the first such pieces is generally given to Girolamo Cavazzoni for the four ricercars in his 1543 print. The form received its classic manifestation in the ricercars of Andrea Gabrieli.

Acclaim for another signal development in the ricercar, the use of a sole theme (which is usually seen as the first step in an ineluctable evolution, via Sweelinck’s Fantasias, to a grand climax: Bach’s fugues), is handed to the ineffable Jacques Buus for the monstrous Ricercar IV of his Libro Primo (1547). But a vastly more important personage beat both Cavazzoni and Buus to their supposed innovations.

Adrian Willaert may be the best example of a composer religiously cited as “historically important”, and then either ignored or misunderstood, or both. His counterpoint is so rigorous, so complex and so devoid of superficiality that it requires long, patient, penetrating analysis before it reveals its profundities. He almost single-handedly founded the Venetian School which culminated in the Gabrielis and Monteverdi, and was the most revered teacher of the early 16th century. Gioseffo Zarlino, who codified his doctrines influential theoretical treatises, was his adoring student, as was one of the very finest mid-century composers, Cipriano de Rore.

Willaert and the younger Giulio Segni da Modena were the prime movers behind the stupendous Musica Nova (Venice, 1540 – not to be confused with Willaert’s 1559 set of motets published under the same title), which definitively converted the ricercar from a rambling free-form prelude into a vehicle for advanced contrapuntal experimentation.* Willaert’s three contributions to the volume already show significant reduction in the number of themes used relative to the number of bars, as opposed to the still-novel “motet style” of continuous, overlapping imitation of new themes for each fragment of text. (Willaert had a major hand in that development as well, along with Gombert and Clemens non Papa, who was even younger than Segni.)

Those three four-part masterpieces of 1540 sound almost symphonic compared with the products of the even greater challenge the “eccelentissimo Adriano” (as Zarlino repeatedly refers to him in Le Istitutioni Harmoniche) then set himself: composing ricercars in three parts. Four such works, one on each of the authentic modes, appear as a kind of appendix to a publication of three-part motets by various composers entitled Motetta Trium Vocum which appeared in 1543. It might seem that juggling three parts would be easier than juggling four, but we are not talking about tennis balls. Since full harmony requires at least three voices, the difficulty of achieving euphony with what Zarlino calls “stylish melodies”, while sticking to the rules of strict counterpoint and endeavoring to weave in clever imitations and canons withal – the difficulty of that complesso is beyond any but the greatest masters. Bach rose to a similar challenge in his most gemlike compositions, the so-called Three-Part Inventions; but he had all the freedoms of late figural counterpoint at his fingertips. Compared to Willaert’s Spartan fare, they are Torte with whipped cream.

But the point of this little essay is that the first of these four 1543 ricercars is both a monothematic and a variation ricercar – the first of either to my knowledge. The other three point in the same directions (nr. 3 is virtually monothematic), as do a few in Musica Nova, but the 1543 Ricercar in re is the genuine original item. The four relatively brief works appeared in the same year as Girolamo Cavazzoni’s first print, mentioned above as the putative fons et origo of the variation ricercar; but Marc’Antonio’s son’s name appears in small print in Musica Nova – one of two indicating apprentice status. Girolamo says he was “hardly more than a boy” when he composed the ricercars in his 1543 book, and they still show traces of the earlier, improvisatory type. He was cooking in the same kitchen as Willaert, and Adriano’s masterworks à tre of the same year put the fanciullo in his place as a sous-chef. Buus’ endless, dull monothematic work only appeared four years later.

One could carry on endlessly about the contrapuntal splendors of Willaert’s ricercars. Those of 1543 were followed by another nine in three parts some years later (seven in 1549, two added in a 1551 reprint**), which show the aging master’s continued fascination with the form. Canons within canons, pre-imitations, inversions, ingenious thematic development and variation, inganni... But the main thing, which cannot be shown by analysis, but only sensed through direct engagement by a performer or an intelligent listener, is Willaert’s unfailing use of such devices to create those intangible entities which distinguish real music from the sham of a Buus — or a Philip Glass.


It is baffling to note that eminent musicologists continue to echo Willy Apel‘s misguided assertion that Musica Nova, in spite of its having been explicitly published ACCOMODATA PER CANTAR ET SONAR SOPRA ORGANI; ET ALTRI STRUMENTI, was intended primarily for instrumental ensemble. The reason given is the concrete form this and similar works took: separate part-books. But intabulation, preferably in open score, was the most common thing in the world for keyboardists of the time, and Willaert and Segni undoubtedly knew how error-ridden and unclear previous attempts at publishing keyboard music on two staves had been. The misery would continue until high-quality engraving on copper, an expensive and difficult procedure, became the norm. Until then, a composer of keyboard polyphony who wanted his voice-leading to remain legible in print chose open score. The first nearly comprehensive list of such publications was provided by Gustav Leonhardt in1952, when he made the epochal connection to Bach’s Kunst der Fuge, itself the biggest of all variation ricercars. Leonhardt’s proof that it is a harpsichord (or clavichord) work still goes largely unnoticed.

This is not to say that a consort of viols or flutes (or the quartet of singers ohne Worte which occupies first place on the title page) had no business using the Musica Nova part-books. But it is worth noting that 16th-century Spanish theoreticians required organists to master the feat of accompanying from choir-book format — four separate parts on a single page. All but the most experienced and mnemonically gifted would surely have preferred to write out their spartitura (or some kind of short score), just as they would have done from the Musica Nova part-books.

*The pre-eminent lutenist of the era, the “divino” Francesco da Milano, had given an important impulse in the same direction with his Fantasias of 1536. Until Musica Nova, “fantasia” was the term used for more or less pure counterpoint either composed for or improvised on an instrument.)

**The first ricercars in both the 1543 and the 1551 prints take the tetrachord as their opening themes (and in 1543 Willaert sticks with it to the end; the tetrachord and the hexachord are the foundation of his grandest ricercar, Musica Nova XIV). The final ricercars of both sets use the opening notes of the Marian hymn Ave maris stella as a theme and an ostinato. In the latter, it appears nine times (= 3x3), in ricercar nr. 9. A long theme in Musica Nova X follows the hymn’s contours closely.

February 1, 2021

- back -