Toward the end of the 15th century, when long canons over a cantus firmus still reigned, some French composers began using folksong-like melodic fragments to compose chansons with multiple sections of quick contrapuntal imitation, a technique which came to be called fuga. Josquin des Prez brought the procedure to its first peak of mastery and expressivity. Jean Mouton, Adrian Willaert‘s teacher in Paris (whom Glareanus placed second on his list of great composers after Josquin), was one of the first to employ it in Masses and motets, which became the main vehicle for contrapuntal display in vocal music. They were soon rivaled by the madrigal. Clemens non Papa, Gombert and Willaert definitively established constant, overlapping imitation of brief themes (or “subjects”), through all the voices and in succession for every text fragment as the norm for serious composition. The Council of Trent demurred, and for a while Western polyphony looked into the abyss; but in 1540, five years before the council began, four part-books were published in Venice that provided sensational new impetus for instrumental music, which until then had fed to a large extent upon works for voices. Keyboard arrangements of these in a variety of notations were called intabulations.

Musica Nova, as this 1540 collection was appropriately entitled, applied the technique of through-imitation to long, strictly contrapuntal pieces which the initiators, Willaert and the younger Giulio Segni, called ricercari. The term, cognate to “research” and meaning “to seek out”, had previously been applied to preluding pieces for lute, the Spanish vihuela and keyboard which were sometimes written-out quasi-Improvisations, sometimes partially in strict counterpoint, and, most notably in the works of the most famous of Italian lutenists, “il divino” Francesco da Milano, sometimes even completely composed in the new motet style. Fantasia was another term indiscriminately applied to all these types, meaning simply a piece not tied to any text, whether improvised or composed. What Musica Nova did was to codify the ricercar as a piece in strict counterpoint for a fixed number of voices, and to lift the form to the highest level of compositional competence and freedom. Themes were no longer bound to an evolving text and could be treated at length. Nor did composers have to concern themselves with vocal limitations, so that bigger, odder leaps, long notes and wider ranges were possible.

There may have been earlier efforts in this direction for keyboard which have been lost. Printing lute and vihuela tablature was relatively easy, which could be why so many more ricercars for those instruments survive. Even Musica Nova is transmitted directly only in a single part-book. Without a Lyon pirate print called Musique de joye we would have no knowledge of it.

The fact that Musica Nova and similar later publications appeared in part-books contributed to the pernicious, persistent myth that the pieces were intended for ensemble, in spite of their titles usually stating that they were for voices, organs and “other similar [keyboard] instruments”. Earlier attempts at printing keyboard music, with its large number of notes to be played simultaneously by one performer, had been plagued by error in degrees ranging from annoying to catastrophic. The easy solution for contrapuntal works was to print part-books (which could also be used by ensembles), and let keyboardists intabulate them – i.e., put them into whatever form of keyboard score they liked. That was often simply spartitura – open score, with each of the voices visible on its own staff. It was a solution which placed high demands on the performer, but which because of its clarity was often chosen for later prints of keyboard music. One of the last such was J. S. Bach’s Die Kunst der Fuge, which I learned at age 22, laboriously and much to my edification, from the old Graeser edition in open score and C-clefs.

The proof that such early part-book prints were intended for the keyboard – if not necessarily only, or even primarily, then at least with that medium always in mind – is their playability. Try to play most motets on the keyboard and you will wind up with your fingers in knots and your brain short-circuited. (But here is a question that would bear examination: to what extent were composers working at their clavichords, and to what degree did keyboard textures therefore unconsciously find their way into true vocal music?) The previous connection of the ricercar to the lute – another solo-instrument – should also be considered. There was a commercial aspect, too: keyboards were ubiquitous in educated Italian homes and in countless churches, and an audience of one is always easy to reach. (So was an audience of two in the 19th century, as witness the flood of printed arrangements for four hands which still choke the stacks of music antiquarians.) The ricercar’s descendant, the fugue, largely remained in the province of keyboard music.


Here is a recipe (1588, Raggionamento di Musica) for a ricercar from the Reverend Don Pietro Pontio of Parma:

“The way to make a Recercar’ is this, that the themes want to be long, and the parts somewhat distant from one another, so that when they are played, the listeners can better hear the themes. And now one part, now another should be in motion, even if there are only two; and it is not allowed for the parts to stand still together in figures of semibreves like lessons of Holy Week. And it is also not allowed to begin with two parts at the same time, unless they are two different themes, in which case it can be done...One is allowed to repeat the same theme two, three, four and more times, as you wish and in various modes. It is also permitted to make your theme like a cantus firmus in semibreves, breves, longas and maximas. It is also permitted to continue from beginning to end with one theme, and if you are not pleased to continue that way, you can find a new subject, and repeat it as often as you like, as I just said; and that is the way to make a Recercario.”

If it only were that easy. The reverend father must be presupposing basic competence in counterpoint, in the way that continuo treatises assume a keyboardist can find his notes, can play chords and understands intervals. Anyone who has taken Palestrina Counterpoint 101, starting with the species in two parts of Fux, knows how hard it is to take the first steps. Bach was said to be able to “make the notes do what he wanted them to do, instead of what they wanted”. Reaching even the distant outskirts of that kind of mastery, following numerous rules of voice-leading and consonance/dissonance, is arduous in the extreme. That so many masterpieces of the 16th and 17th centuries have come down to us, with Willaert’s in the very first rank, is testimony to human diligence in times past.

But just composing counterpoint according to the rules is of itself useless. The 17th-century Jesuit polymath Anathasius Kircher created a “machine” (a little box filled with wooden sticks) which could do that. Even the most clever interweaving of themes can turn out dry as dust. Creating densely thematic counterpoint which is music – a quality standing in relation to “analysis” as a great meal to a printed menu – is the ultimate in musical craftsmanship, and in that respect, Willaert is a consummate master. His ricercars cohere and build; they have what the Germans later called Affekt; his melodies are what Giuseppe Zarlino, his student and author of the most important treatise on music of the 16th century, called “stylish”. They even retain much of the freshness and simplicity of the chansons of Willaert’s Paris years.


Willaert, who after serving the Ferrarese house of Este for some years was maestro di cappella at San Marco in Venice from 1527 until his death in 1562, is a textbook example of a composer religiously cited as “influential”, and then ignored in practice. And it must be said that he reveals his secrets reluctantly. I have found that getting past the initial stages of analysis and learning the notes from the score to a state of real understanding takes considerable time and patience, but the effort is always richly rewarded. What at first seems dry and barren turns out to be an endless source of astonishment and warmth. The listener will need the same degree of engagement. The old term musica reservata could be applied, for this, too, is music for a small circle of knowledgeable participants. It is open to all, but eternally rejected by the harried mass of humanity. And as I write, that circle seems to be shrinking by the hour.

Willaert’s longer canonic subjects are especially difficult to perceive by ear alone. These themes, often around 16 semibreves long and buried deep in the contrapuntal web, form the backbone of his ricercars. He almost always prepares them with their first notes, passed around the voices. Dux and comes (the “leader” and the “follower” – and there is often more than one of the latter) don’t usually overlap. Willaert is sparing with such effects, as with all his means. Dissonance, for example, is often absent for long stretches. But when Willaert strikes with an overlapping stretto, or a harsh discord, he strikes hard. Inversion is also rare, and is only found in the incipits just mentioned. He never uses augmentation or diminution – the lengthening or shortening of a subject’s note values; it was left to Andrea Gabrieli to explore those devices.

One practice Willaert frequently does indulge in is called inganno – literally, “deception”. He may have even invented it; the first description is found in Zarlino. It changes the intervals in themes by reassigning them the same note name in one of the other hexachords. These are three archaic six-note scales beginning on C, F and G – the original ut (later do ), re, mi, fa, sol, and la. This may seem odd...because it is. A theme which began with, for example, “C/D” can have those notes transformed into G/D, or G/G, or F/D, changing its shape and making it more or less unrecognizable, but still retaining its status an an entry of the subject. Any or all its notes can be thus altered. I have never been able to see the point of the device. I imagine it was a kind of trade secret or in-joke, in an era when mysteries and puzzles were highly prized.

But Zarlino also counts something else as inganno: snippets of counterpoint around a subject, not with the same intervals, but the same rhythm. These are very recognizable, and a useful means of unification. What throws one off the scent almost as much as the hexachord-inganno flummery is a theme which is shifted in time by one minim, which alters the accentuation; arsis and thesis, “good” and “bad” notes switch positions. Willaert does that quite often, and it always upsets me, even while I admire his ingenuity.

In the two years which I have spent until now (the end of the disastrous year 2021) studying these 85 minutes of music, I have had ample time to reflect on just what it is that makes densely thematic polyphony so deeply satisfying. I still haven’t found a fully satisfactory answer. The practice had its earliest roots in single notes added to plainchant by monks bored with the same old same old. It evolved into splendid, geometrical layers of gothic tracery (isorhythms, endless canons) before becoming the highly-wrought system of Willaert’s time. The earlier manifestations had little regard for anything besides the worship of God and his natural order as reflected in the order manifested in the works themselves. That aspect lived on in Josquin and his successors like WIllaert, but they were more interested in expressing the meaning of a text in flowing musical rhetoric. They certainly didn’t neglect canon, that most abstract of musical devices. (At least it was until the advent of Schönberg and serialism).

But is a nod to universal, divine law the sum of all contrapuntal delight? There must be something more earthbound and visceral. The following doesn’t add up to much, but is the best I can do for now:

An element of pleasant surprise plays a role, a sharing of pleasure with the composer’s recombinant genius. We like rhyme in poetry (or we did when we were little); does that not correspond to the shifting repetition of subjects and rhythms? And I think there is an undefinable human need to perceive something organic and unified. It is this need that partly led to the Gothic works just mentioned, and to the great cathedrals themselves. Perception itself is a pleasure to an intelligent mind. A kind of auditory fusion or catalyzation takes place when themes pass by the ear successively, the same as they were and yet different because of their context. Call it the transmutation of musical alchemy. Aborigines in Australia made mental maps of half the continent to guide their migrations; is there not an evolutionary need for that to give a feeling of satisfaction? I have seldom enjoyed a sight more than that of the Great Wall of China snaking over the contours of the landscape; it keeps its basic dimensions while constantly changing shape. In a similar way, I love to watch a steam locomotive’s parts moving in complex, perfect coordination to a useful purpose. A ricercar or motet follows its myriad laws while weaving its thousands of moving parts together in an ever-changing fourth-dimensional landscape.

And surely the urge to build, and to understand that which has been built, is deep within the human psyche. Willaert’s ricercars are nothing if not impressively constructed. It was said of him that he was great because he devoted all the time necessary to perfect his works before releasing them for performance or publication. Here are some of the building blocks he uses:

* Duets, repeated in voices previously silent – Josquin’s favorite device. Sometimes canonic. An obvious choice for variety of texture in four-voice pieces; more difficult in three voices, where one participant has to be both an ending and new beginning. A duet is also sometimes embedded in other voices.

* A countersubject, introduced alongside the first theme (as allowed by Rev. Pontio); it can offer material for later independent development.

* Themes introduced immediately in canon, either partially or wholly.

A section in triple time at the end of a piece, raising the work to the next level of rejoicing.

* A long section in two parts, to which the addition of the third creates a spurious feeling of voluptuousness.

* A ripresa – a repeat of a final section when Willaert thinks his final, most complex development needs another hearing. This is a characteristic borrowed from the French chanson, where it serves to emphasize a punchline or dénouement.

* Very rarely a first or earlier theme will reappear, later or at the end of a ricercar, lending it a cyclical quality.

* Sometimes the voices do in fact “stand still together in figures of semibreves [minims in Willaert’s case] like lessons of Holy Week” in spite of Rev. Pontio’s prohibition – but unlike Holy Week lessons, they are engaged in intricate thematic interweaving, and the contrast with the usual complementary motion is intentional.

* An ostinato – a motif or subject which is repeated more often than such an element would normally be – sometimes appears. In his longest and most intricate ricercar (Musica Nova XIV) Willaert takes up Josquin’s la-sol-fa-re-mi (lascia fare mi) as the work intensifies, keeps it for some 70 semibreves, and develops it for his penultimate long canon.

* Tessitura gets manipulated for variety. Willaert will bunch voices together high or low, or split them to maximum distance.

* Tetrachords and hexachords are often used as subjects. Musica Nova XIV is a massive fantasia mainly based on these scales, with the full octave thrown in for good measure. Its final theme begins with the “natural” hexachord on C and ends triumphantly in triple time.

These are all large molecules, built up from the atoms that actually comprise the DNA of Willaert’s ricercars. Those long strands, which defy academic analysis, determine what the works really are, and how they grow. The invisible aural entities that emerge bear a family resemblance to each other, but are all unique. They can only be understood by those who speak their common language, which is now as moribund as Latin or Homeric Greek, but just as worthy of study.

One major element of Willaert’s compositional process needs special mention. In his search for means to unify his ricercars, he seized on the principle of variation. In the two later sets his themes are almost always a chain of developments: sections of intervals from one are taken over in the next, shapes are expanded or contracted, rhythms more or less subtly altered, a rhythm taken from the end of a subject for the beginning of the following.

He shies away from the final unifying step: a single, unaltered theme. That must have seemed to him a too radical departure from the prevailing norm, and for a long form like the ricercar it could lead to boredom, as we see in Ricercar IV Libro Primo by Jacques Buus, who is usually credited with the innovation. (Andrea Gabrieli’s ingenious solution – multiple countersubjects and use of augmentation in the main theme – was brilliantly expanded upon by Sweelinck, but was ultimately a dead end.)

And yet there is one ricercar by Willaert, the first of his 1543 set, that comes so close to having one subject that I have claimed it elsewhere (article 23) as the first monothematic variation ricercar – an apparent contradictio in terminis. It opens with a tetrachord; this leads on to another tetrachord, which then curves into the shape of the favorite hymn of the seagoing Venetians, Ave maris stella. This theme is repeated twice like an invocation; then the composer takes the ascending and descending tetrachord as his sole subject, with only slight variation to the end.

Girolamo Cavazzoni, present as an apprentice in Musica Nova, gets musicological credit for publishing the first variation ricercars in 1543. They are delightful, half-improvisatory pieces, but far inferior to the ones by Willaert printed in the same year. Cavazzoni himself admits that his were youthful works. Did the younger man have an interesting idea which his teacher took up and perfected, or were there variation ricercars by Willaert of an earlier date which served as inspiration for the up-and-coming lad?

The answer may lie in Willaert’s third and final set of ricercars. Whereas the four pieces of 1543 form a masterly small cycle on the finals of the “authentic” modes, the set of seven first published in 1549, and then rearranged and expanded to nine in 1551 (with several later reprints) shows a few signs of Willaert wrestling with the extremely difficult three-voice format. Satisfying harmonic progressions are sometimes sacrificed to contrapuntal expediency. Some of them may therefore date from much earlier, to be assembled later for publication; but all employ thematic variation extensively.

Musica Nova is in four voices, which allows far more leeway towards achieving satisfactory harmonies. With only three voices, the basic triad is hard to form when the composer is bound, not only by so many rules of strict counterpoint, but especially by the demands of stylishness and musicality. The two ricercars Willaert added in 1549 are superb. I think they were composed last, to give the set a new dimension. They occupy first and last place, and the meditative ricercar number nine invokes Ave maris stella again – a total of nine times. Students of numerology will recognize the possible significance. Number one starts with an ascending and descending tetrachord. The four-note scale figure is so prominent in the other seven pieces that one is tempted to think of the cycle as “a giant variation ricercar”. Die Kunst der Fuge has been thusly described.

The composer of that latter epic had an easier time writing in three parts for his gemlike Sinfonias, the so-called “Three-Part Inventions”, than Willaert did. Rules had become lax, and figuration luxuriant. But all those older composition rules were not formulated for the sole purpose of making life difficult for composers and students of counterpoint. They had a logic behind them based on an extremely refined ear for perfect consonance and dissonance; they had their foundation in the laws of acoustics and the physics of the natural world. As for figuration: the simplicity of 16th-century themes and countermelodies is not the result of naivety or primitivism. It is, rather, a quality that enables easier perception of how the lines are interlaced, and which also corresponds with the demands of choral practice. On a broader plane, it reflects the spirit of clarity and serene rationality which is the defining characteristic of the Renaissance. It has the disadvantage, relative to Figuralmusik, that a certain sameness can seem to prevail in all pieces. But what is wrong with...let us call it a stylistic unity, which also prevails, say, among the frescoes on the walls of the Sistine Chapel by Rosselli, Ghirlandaio, Perugino, Piero di Cosimo, Botticelli and others?


A word now on the characteristics of the individual modes, the old scales which had largely collapsed by Willaert’s day, but still nominally functioned like keys. Tinctoris had already determined at the end of the previous century that they only work for a single vocal line and are useless for polyphony, but Zarlino painstakingly quotes all the classical authorities. He remarks that the evidence is sometimes contradictory, often confusing, and that not too much should be made of the matter. And yet, there is enough concurrence among the ancient Greeks and Romans that at least an approximation can be made. The Dorian (on D; all the modes use only the piano’s white keys) is serious and grave; the Phrygian (on E) is quick, warlike and aggressive – in contrast to the lower-range Hypolydian, which was used to help soldiers re-attain calm after the rage of battle; Lydian (on F, which Zarlino tells us “nowadays” usually alters B to B-flat to avoid the tritone problem) is light, happy and loving; and Mixolydian (on G) is relaxed, forgiving, peaceful or plaintive.

These characteristics carried down at least to some extent as far as the late 18th century and even further. D minor is often what Zarlino calls “material for a serious conversation”; E minor is often grim; F major open and gay; and the number of pastorales in G major is beyond counting. These vague guidelines have been of some help in varying the tempi of Willaert’s ricercars, and the musical material in them actually seems to agree with Zarlino’s summary. The subject certainly would have been discussed in the humanist circles around Willaert.

But his true attitude towards the modes is best indicated by the titles of the 1543 set. They are simply “on re”, “on mi”, “on fa” (with B-flat, hence F major) and “on sol”. (The last one ends in lilting, pastoral triple time.) The subject of added accidentals was controversial at the time. “Mixing the modes” was considered dangerous by conservatives; liberals used the concept to excuse a practice which had, as usual, left theory behind. Willaert shows himself to be a radical. His use of accidentals to draw attention to an important moment, or at times seemingly as pure whimsy, can be quite shocking to an ear attuned to the ordinary practice of the time. His modulations, which theory insisted were hexachord mutations, are at times almost reckless, and more often than one would like they raise questions as to where musica ficta is to be applied.

Sometimes Willaert begins and ends in different modes. Tinctoris was already aware of this deplorable practice towards the end of the previous century, and in his struggle to preserve modal order he ordains that the final bass note determines what mode a piece is in. Glareanus’ Dodecachordon (1547) brings theory somewhat up to date. He says most composers use his new modes 11 and 12 anyway – what we would call major keys. Willaert certainly does, often “transposed” (as theory insisted) down a fifth by adding a flat to the key signature.

The word “tempo” cropped up a little ways back. Nothing wrecks the impact of 16th-century music like the headlong tempi usually taken at present. There is plenty of contemporary information on what the general speed should be, with some adjustment allowed according to content and individual preference – but that doesn’t stop people who are supposed to be “historically informed” from throwing powerful dissonances, intricate figures, cross-rhythms, contrapuntal complexities and the oft-reiterated need for gravitas to the gale force winds of modernity. The hand marking the upbeat and downbeat measured by minims (half-notes) moves calmly, as we are told again and again. The music unfolds in the serene rhythm of cloister arches by Brunelleschi. Within that space its ramifications can be followed and appreciated. A beat of around 80 minims per minute is usually called for, the same as Quantz’s basic pulse. In many passages the harmonies change on every beat, and sometimes even at twice that speed; they simply can’t sink in if the tempo is much faster. (The ratio between this harmonic rhythm and the opening E-flat of Das Rheingold is approximately 400 or 800 to 1.) A well-made ricercar is not a race to the finish line, no more than a bottle of Château d’Yquem is to be guzzled down with French fries.


If any attention is now paid to16th-century ricercars, it is mainly to their role as precursors to Bach’s fugues. There is little or no appreciation of their intrinsic value on their own terms, the reason being that they are so much more demanding in their asceticism. The last time I saw Gustav Leonhardt at Huis Bartolotti, he had just played some Bach in the Nieuwe Kerk. He emphatically agreed with my comment at the downstairs dinner table that the Thomascantor is shockingly modern. His fugues are so much more accessible than the works of the previous century; with their welter of figuration and lush harmonies they are actually ornate, quasi-Mannerist versions of ricercars. To return to the Sistine Chapel: I can appreciate Michelangelo’s grand conceptions on the ceiling and behind the altar, but I’d much rather look at the quattrocento frescoes on the walls. The natural grace of their figures appeals to me far more than the Florentine’s stereotype-musclebound, titanic humanoids. When the titanic element in Bach becomes wearisome, I take refuge in ricercars.

Their usual road map is a constant crescendo of intensity, as a means of holding interest. Willaert sometimes turns this around, and ends more calmly than he begins. For example, Ricercar X of Musica Nova begins in the warlike Phrygian, but passes through phases of quieter modes to end peacefully. Is it too fanciful to see here an allegory of Venice’s wars?

Willaert was intimately connected with the Serenissima’s artistic and literary intelligentsia. Aretino, in his play Il Marescalco, called him a “forza di natura”. He certainly appears like one in a largely-unrecognized a portrait of him held by the Chicago Art Institute. They attribute it to Palma Giovane, but that painter was only 18 when Willaert died, and the portrait is of a man in the prime of life. The sitter looks like a cross between Zeus and the Cyclops. I am convinced it is by the composer’s friend Titian, who included a snippet from a chanson by Willaert in his Bacchanal of the Andrians. The large canvas is now in the Prado but was painted for the ducal camerino in Ferrara. The two had met there in 1516. Titian must have been amused by the similarity of the title of his mythological subject to his friend’s name.

Edward Lowinsky was the first to note the resemblance to the only authenticated portrait of Willaert. The composer is shown as an old man on the frontispiece of another Musica Nova. This later one (1568) was a posthumous collection of brilliant motets and madrigals composed for the Duke of Ferrara.

Zarlino refers to Willaert as the "eccelentissimo Adriano" some 20 times in the Istitutioni Harmoniche. Cipriano da Rore, the finest composer of the mid-century, called him “mio maestro”. His influence is incalculable; the later Venetian School is unthinkable without him. The 16 ricercars (a piece so titled in Musique de Joye is clearly a transcribed chanson) are a relatively insignificant fragment of his output, but they are some of the most precious of the forgotten jewels of the keyboard repertoire.

December 23, 2021

- back -