Who put the C in Allegri?

Everybody who's listened to a "Great Choral Classics" recording knows the piece. You know, this one. The one with that impressive high note every other verse. It's a fixture at Ash Wednesday services, too, providing a suitably solemn background to the imposition of ashes. It's been a classic for a while; there's a famous recording of it, in English (as was then fashionable; I wonder when performing in translation became gauche?) from 1963 or thereabouts, and the liner notes tell how the soloist was only tapped just before the service, still grimy from a game of rugger that afternoon. (That soloist, Roy Goodman, went on to be a conductor and was for a time director of a chamber orchestra in Manitoba. Fun fact.) In high school, when I needed to unwind from a day of adolescent complexities, I would listen to Allegri's Miserere, and let the high C's float away like my troubles. The thing is, at some level the piece is a lie. For one, the chant is all wrong; the choral verses are based on a Tonus Peregrinus chant, not the mode 2 chant it is almost always sung with. But for two, the *C* is a lie. This first became obvious to me when I sang the piece as transcribed by Liszt, who made all the verses the same: high Gs, but no high Cs. Looking at it, it seemed the C was just...the previous verse up a fourth, for some reason. But let me back up. Liszt? What's he got to do with it? Well, the piece was famously the property of the Sistine Chapel choir, who didn't sing what was on the page. Allegri had written a fairly simple set of chords around the base chant, but the choir ornamented it at leisure, perhaps doing some ornaments long established by their predecessors, perhaps adding their own, until at some point through the process of mutation and evolution one arrives at today's product. Is it still Allegri, then? A good seventeenth-century composer, Allegri probably *expected* that his composition would change, and would have been most disappointed in a choir that merely sang what he *wrote*. (In other words, he expected there to be something unexpected.) Being able to ornament decorously was important for any Baroque singer; being able to improvise in a group, meanwhile, had been a practice going back to the organum of the middle ages. Allegri might also have been disappointed, by the same token, that nobody changes the notes anymore--if it *doesn't* have that high C, it isn't "the" Allegri Miserere, for a lot of people. One is reminded of the ship of Theseus, each of its boards gradually replaced by another, until it both is and isn't Theseus's ship.
Alright, but how did this piece, in this version, become a fixture of the English choral tradition, in particular? I happened to read yesterday, while looking up Ivor Algernon Atkins (author of a most excellent Anglican chant, friend of Elgar, and translator of the St. Matthew Passion, and arranger of Cornelius's Three Kings for chorus and soloist) that he is credited with doing the job.
Novello published an edition back in 1905, as you can see in this list here. (As an aside, I have tried to find out more about "Ada Marguerite Burnside", lady composer of the Edwardian era, but she seems to have died, at age 45, not long after the publication of her pieces here, and--since I presume she didn't hold an important organist post--her composition and legacies have disappeared without a trace.) This Novello edition, priced at 4 pence, may still be found for one's perusal, for example here. One might note the soprano part in this edition stays below an E-flat--within the compass of an adult male choir.
But the astute musician of the early 1900's might also have read about the piece in a lengthy entry in Grove's Dictionary of Music and Musicians (then merely in it's third edition!). Grove was somewhat more colourful in those days, and the entry describes the whole service at which the piece is sung, and relates a famous (semi-apocryphal) anecdote of a teenage Mozart hearing the piece and then transcribing it precisely at a single listen. This edition supposedly made it into the hands of the music historian/chronicler/personality Charles Burney, who published it in the eighteenth century. (The secrecy surrounding the piece added a bit of extra hype). But Burney's edition--though the Grove claims Mozart's transcription was praised by the singers of the Chapel of that time, and that it has been handed down unchanged to the present day--has no high C's either. (It seems likely it merely shows the "simple" version, and that Mozart's version never circulated--if indeed it resembled the version we know today.) All this came from the hand of William Rockstro, né Rackstraw, a decidedly Victorian character fond of Romanticizing early music. Rockstro includes a description of the piece by Mendelssohn (uncited, so I can't find it) who, like Liszt and Mozart, had gone to hear the famous piece for himself. Mendelssohn *does* mention a high C--at least, he does according to Rockstro. I see elsewhere this may be because the choir at that time sang the entire piece up in pitch, for effect (probably about a major third, which Mendelssohn put up another half step to account for pitch differences in Rome--but he seems to have heard it at several pitches, and in any case the choir had a tendency to go flat by the end.)This was not how Rockstro set the piece in his illustration, however; in his version, the first half of the verse is at the written pitch, and the second half transposed up to make the high C's Mendelssohn described.
When Novello issued another edition of the piece in 1951 (there was an intervening edition, also with the high notes, ca. 1931), Ivor Atkins dutifully noted that his abbellimenti were based on none other than Rockstro, thus further propagating the (error? embellishment? creative license?) of nineteenth-century Grove. He also cites Pietro Alfieri, a nineteenth-century fellow who attempted to bring back classics of the Catholic musical tradition from Vatican manuscripts. Alfieri's edition mentions that G's are sung as B's, that is to say, a third up, and conceives of the piece as one using a fixed set of embellishments provided in a table at the back. The same embellishments can be used for the other (less renowned) Miserere sung by the papal choir, by Tomasso Bai; it seems the practice was to alternate versions of the one and the other, and they were in any case basically variants on a theme to start with. The embellishments might be old; they might have been a recent practice in Alfieri's day (ca.1840). But once again--no high C. That, it would seem, is all Mr. Rockstro.
Incidentally, the Tallis Scholars recorded a version with Tonus Peregrinus and with some semi-improvised embellishments. Even then, though, they couldn't face leaving out the high C--it's simply, at this point, too much a part of the piece. There is also an oddly campy recording that reconstructs the Allegri along with the later Miserere by Tommaso Bai with which the chapel choir evidently alternated verses. But is this Bai's composition, or Bai's version of Allegri, or Bai as he might have heard it, or is it how the nineteenth-century papal choir might have sung a two-hundred year old piece?
Who knows. At this point, "Allegri" is only about 40% Allegri, 30% papal choir embellishments as handed down over centuries of oral tradition and periodic transcriptions, 20% Tommaso Bai's verses, and at least 10% an egregious but much-propagated error in Grove. Which goes to show, I suppose, that encyclopedias can have a lasting impact.


post script: there's also a very interesting recording by the Sixteen detailing some of this evolution, based on the work of Ben Byram-Wigfield, whose essay I linked to in part above, and who considers all of this in considerably more detail.