Genealogies persist in the writing of histories, if not to suggest lineages then to delight or dismay us with the effects of accident and circumstance. The first generation of Italian operas were all court, that is private, entertainments, whether produced in Florence, Mantua, Rome or Turin. When this new form of entertainment first emigrated, it spoke Italian in foreign courts (except in Madrid), perhaps because it was so identified in its novelty with still novel Italian musical styles, with the dramaturgy of Italian spectacle and above all with the Italian way of singing. It travelled to the Habsburg imperial court early, in Prague, 1627. Like this performance, later occasional stagings in Vienna and Innsbruck remained in Italian and at first by Italians, though eventual progeny included scores by Gluck and Mozart. The court in Madrid, however, heard in 1627 a version of opera with a libretto by Lope de Vega, La selva sin amor; its score was by a theatrically inexperienced Tuscan lutenist, Filippo Piccinni. Madrid did not essay opera again until 1660. The court of Sigismund III in Warsaw heard Italian opera beginning in 1628, possibly with a Mantuan score. Opera made its way to Paris only in 1645, where, after acquiring a French text and French dancing and singing, it became one of the glories of the royal court. The cradle of opera, in all these cases, was the court, with all its resources in terms of money, an obligated audience and more than willing professionals, whether imported or local.

One of Così fan tutte's more amusing episodes occurs in the first-act finale when Despina marches on stage disguised as a doctor and, in a spoof on Mesmerism, produces a magnet to heal the stricken Albanians. She introduces this little scenario with a ridiculous salutation: ‘Salvete, amabiles, bones puelles’. The source of this mangled Latin is interesting: it is present in Mozart's autograph, but not in Da Ponte's libretto, where the Latin appears in its proper form, ‘bonae puellae’. Only two words, to be sure, but they have attracted attention because of the rare glimpse they offer of Mozart's direct intervention in a Da Ponte text, of a parting of ways between composer and librettist. Mozart's change has traditionally been interpreted not as an error at all, but as a kind of ‘correction’ of Despina's character: since the servant Despina lacks the status and education to know proper Latin, Mozart portrays her in a way that is true to her station. As one commentator says, she ‘understands only half of what she says; she imitates what she has picked up from frequent doctor's visits to her masters’.

In a recent number of 19th-Century Music, Allan Atlas offered a complex analytical reading of Puccini's Madama Butterfly, hinging on a pair of keys forming what he called ‘crossed tonal areas’. The tonality G flat major in the second-act trio ‘Io so che alle sue pene’ plays a crucial role in Atlas's reading, and he had to account for the fact that the trio was originally conceived a halfstep higher, in G major, and was so performed at the première in Milan. He asked:

can we assign a significant role to the Gb of the trio when it represents not Puccini's original – and presumably well-reasoned – intention, but what might be little more than a compositional accident necessitated by the practicalities of performance, and a specific one – Brescia, May 1904 – at that?

And in his footnote 13 to the foregoing, Atlas listed six studies discussing halfstep transpositions, three in Verdi and three more in Puccini.

Finally, love, – love retranslated again into nature! Not the love of a ‘cultured maiden’! No Senta-sentimentality! But love as fate, as fatality, cynical, innocent, cruel, – and thus true to nature! Love, which in its expedients is the war of the sexes, and in its basis their mortal hatred.