Francesco Guardi, “The Ladies’ Concert”, 1782. Alte Pinakothek, Munich, via Wikimedia Commons (public domain).
In 1723 a book appeared in Bologna with a combative title right on the cover: Opinioni de’ cantori antichi e moderni, o sieno osservazioni sopra il canto figurato. The author is Pier Francesco Tosi, a castrato born around 1653, who had sung across half of Europe and taught in London for decades. The book is the first great treatise on the art of singing ever printed, and also the first diatribe of an elderly professional against the young who no longer know the trade. The two things, for three hundred years, have tended to coincide.
Inside is the complete catalogue of what a singer had to possess: the messa di voce, that is the art of swelling and diminishing a held sound without the stitching showing; the distinction between chest voice and head voice and the way of welding them together; the appoggiatura, the trill in eight different kinds, the rule that cadenzas may be ornamented but not stretched until the piece is forgotten. And there is a term that appears here in print for the first time in history: rubamento di tempo, the rubato, allowed to the singer as long as the bass underneath does not stop. When a pianist slows down a Chopin nocturne, he is using an idea that a Bolognese castrato put in writing a century earlier.
But the page that makes your head spin today is another. Tosi taught that within an octave there are not twelve sounds: there are fifty-five, and he demanded that a serious singer tell them apart, because the sharp on the way up and the flat on the way down do not fall on the same spot. It was no eccentricity: it was the ordinary fineness of ear before equal temperament reduced the palette to twelve identical keys. Alessandro Baricco, in his Breve storia eretica della musica classica (Feltrinelli, 2025), uses Tosi precisely to measure how much we threw away in that passage: from fifty-five sounds to twelve, he writes, there is a leap that must be taken literally. Baricco’s book sits on the history and criticism shelf of the Melos library, next to the reproductions of the treatises it discusses.
Tosi had the publishing career that handbooks dream of: translated into English in 1743 by Galliard, into German in 1757 by Agricola, who added a commentary as long as the original. It is in the public domain and still reads with pleasure, because it is written by someone who does not theorise: he settles scores. With singers who shout, with those who imitate instruments, with teachers who pass pupils out of friendship. It sits in the Melos library next to Quantz, who measured tempos by the pulse and Gasparini, who taught composing while you play: three manuals that say the same thing, namely that in the eighteenth century written music was only half the work. The other half lived in the ears, and the ears of that time heard more notes than ours.