Allegri’s Miserere: Conclusions

"Some have stood in an opinion, which to me seems very improbable, that men accustomed to descanting can sing together on a plainsong without singing either false chords or forbidden descant, one to another, which, till I see it, I will ever think impossible. For though they should all be most excellent men, and every one of their lessons by itself so well framed to the ground, yet it is impossible for them to be true, one to another, unless one man should cause all the rest to sing the same as he sang before them."

Thomas Morley, A Plain & Easy Introduction to Practical Music, 1597


This research has largely been superceded by the much more detailed work in Graham O'Reilly's book Allegri's Miserere in the Sistine Chapel (Boydell Press 2020). I recommend it to anyone interested in the subject.

Performance Practice

It is clear that throughout the eighteenth century Allegri's work was performed as written in MSS 185 & 206, ornamentation aside. It is this that the fourteen year old Mozart would have heard in 1770, and it is from this that the piece's reputation spread. The similarity between the ornaments in MS 31525 and those detailed by Alfieri, Amann and Mendelssohn would suggest a certain authority.

abellimenti 1

The biggest legend lies in the practice of the performance itself. These ornaments are not Baroque additions to the work, but pre-extant Renaissance ones! They are originally based on the Papal Chapel choir's ability to improvise counterpoint around a cantus firmus, and with this came ornamentation of certain contrapuntal sequences. Such masters as Ockeghem and Josquin have left records of how they expected specific polyphonic movement to be ornamented, particularly movement up and down the scale. Today, imitative decoration is still camped up by Lay Clerks singing verse anthems in cathedrals all over Britain, who can recognise the appropriate gem for a certain harmony. So, it may be that Allegri composed the work to show off these existing ornaments in a new light. It is also significant that Bai's Miserere was designed to use the same ornaments as Allegri's version. The similarity between the recorded ornaments over an hundred years confirms that this was not improvisation, but a regulated system.

abellimenti 2

However, by the eighteenth century, this ability was almost extinct outside the Vatican, so did they still perform in the same way as their forebears, or did they learn it by rote? It is not clear whether the Papal Choir were still tutored in such traditions of Palestrina by the eighteenth century. But could the ornaments presented here be sung ad libitum without any collusion? Over the history of the work, it is hard to imagine that this was so rarely consigned to paper. Indeed, the music for Tenebrae was rehearsed, a rare treat in those days, because it would be performed in near darkness. The source of the ornaments is beyond doubt from the renaissance teachings of such masters as Palestrina. How they were performed in the 18th century is still questionable.

Those ornaments in the first choir, shown in MS 31525 and elsewhere, are clearly not based in polyphonic theory, but merely a baroque attempt to complement its renaissance partner. Of course, it may be that MS 185 is the ornamentation for the choir a 5. Burney's model for this choir is then another refinement, possibly written by the young Mozart himself. Up to Top


The legend of secrecy stemmed, no doubt, from the failure to appreciate the source of the ornaments. The only rite of initiation was the standard musical training of the day. There never was a decree of excommunication issued by Urban VIII. But such a myth increases the currency of copies (the forbidden always comes at a premium), and also tells you what people were doing. (In the same way that a "No Smoking" sign tells you people smoke.)

There were supposedly three authorised copies outside the Vatican, held by Emperor Leopold I, the King of Portugal, and Padre Martini. Leopold had requested the Pope that the imperial choir in Vienna might perform it, and he was duly sent a copy, but it was clearly unornamented, as he complained to the Pope that he had been sent some inferior work. Padre Martini (1706 - 1784) was sent one out of the high regard in which he was held throughout Europe as a music scholar. Burney is said to have made his copy from Giuseppe Santarelli (1710-1790), the chief conductor of the Papal Chapel, and then compared it to that of Padre Martini in 1770. It is curious, therefore, that his choir a 5 differs significantly from the Vatican sources and most other 18th-century sources, and his choir a 4 is less ornate to the point of boredom. Mozart also visited Padre Martini in Bologna prior to his attending the Holy Week performance in 1770. It was at this performance that he, at the age of 14, is said to have copied out the forbidden music.

Leopold, in a letter to his wife of 14th April, says:
You have often heard of the famous Miserere in Rome, which is so greatly prized that the performers are forbidden on pain of excommunication to take away a single part of it, copy it or give it to anyone. But we have it already. Wolfgang has written it down and we would have sent it to Salzburg in this letter, if it were not necessary for us to be there to perform it. But the manner of performance contributes more to its effect than the composition itself. Moreover, as it is one of the secrets of Rome, we do not wish to let it fall into other hands.

However, doubt has been cast on much of this story. The legend of Wolfgang returning on Good Friday to correct his score, with his manuscript under his hat, is pure 19th-century invention. In any case, the young Mozart did not 'reveal' the secret to the world: the work had already been performed twice in London, and copies were on sale in the Via del Corso to gentlemen on the Grand Tour.

Up to Top

Remaining Questions

One of the biggest mysteries is how the first choir became altered, as written in ms 185, and by whom? Tommaso Bai's setting of 1713 was produced to echo that of Allegri, and the psalm was later sung as a mixture of these two pieces. The re-worked Allegri is not taken from Bai's own setting, but could Bai have subsequently re-styled Allegri's original verses? He died in 1718, which would seem to rule him out; but Allegri's own music doesn't make it into the Vatican library until some years after his death, either.

Another point concerns the plethora of unadorned manuscripts. Why did no-one gripe, as Leopold I had done, after being shown an unadorned copy if they had heard the piece performed? Did none of these musicologists on the Grand Tour attempt to fill in the gaps after paying through the nose for a manuscript which wasn't all they had hoped? The ability of this renaissance embellishment was not known outside the Vatican - how then did foreign singers hope to perform it without the notes?

Finally, could Burney's different version be a variation created by Mozart? Burney's published version is different from all other sources, being more elaborate and polished. But if Mozart had shown Burney his copy, why did Burney not publish the abbellimenti? There was no real danger of excommunication, though perhaps popular opinion might have condemned the act. Leopold Mozart's claim that he did "not wish to let it fall into other hands" is strange.

The question of pitch is also intriguing. Alfieri's comment that the Papal Choir 'converted G into B diapason', coupled with Mendelssohn's copy up a fourth gives strong ground that the work was transposed. Mendelssohn's top G up a fourth is of course where the top C enters the story. So the top C could still flourish by performing the work in C minor! Up to Top

Summary and Chronology

1514 Constanzo Festa composes the Miserere as a falsobordone for two choirs, one of 5 voices, the other of 4, for use in Holy Week liturgy.

1638 Gregorio Allegri composes his setting of Miserere in a similar format, allowing the singers to use a number of ornamentation techniques.

1652 Allegri dies.

1661 Allegri's setting is written into two books, which contain twelve different settings of the Miserere, each a falsobordone for two choirs of 5 and 4 voices.

1713 Tommaso Bai's setting of the Miserere, borrowing heavily from Allegri, is written into the Vatican library.

1731 Allegri's first choir appears re-written in a heavily altered form.

1735 The work is first performed in London.

1770 Mozart hears the Sistine Chapel choir and obtains a copy.

1770 Charles Burney meets Mozart.

1771 Burney publishes the Miserere, without abbellimenti, ending the papal monopoly. Reprints soon appear in Germany, France and Italy.

1831 Felix Mendelssohn transcribes the Miserere, seemingly performed up a fourth from written pitch.

1840 Pietro Alfieri publishes his detailed account of the abbellimenti.

1880 W.S. Rockstro incorrectly reproduces the second choir to illustrate an article in the first edition of Grove's Dictionary of Music and Musicians. Mendelssohn's transcription of the top G ornament, up a fourth, is inserted into the second half of the verse.

1932 Robert Haas publishes an edition with abbellimenti, copying the error directly from Grove's, producing the 'top C' version.

1935 Julius Amann publishes Allegris Miserere und die Auffuhrungspraxis in der Sixtina, detailing the correct abbellimenti and performance practice in the Sistine Chapel.

1951 Ivor Atkins copies Haas' error into his English text edition. Subsequent editions lend authority and credence to the error by repeating it. Up to Top

A Performing Edition

Having reviewed the evidence from manuscript sources, the next task is obviously to produce an authentic edition which can be performed by singers trained in contemporary methods and with contemporary notation. The next section of the essay deals with defining and producing such an edition, and the requirements upon choirs that perform it.

The performing edition is based on Vatican MS 185 for the first choir a 5, and British Library MS 31525 for the second choir a 4, with the ornaments rationalised and underlying corrections made from Vatican MS 206. Note values have been halved to make the archaic notation slightly more 'contemporary' with the eighteenth century, all bar-lines are editorial, written pitch has been maintained, although the key signature has been given an extra flat to replace an abundance of source accidentals. There are no editorial accidentals or dynamics. MS 185 gives the time signature as C.

In the second choir, MS 31525 shows the speed of the ornaments to be counteracted by a slentando. In Alfieri's notation, this slentando is effected by elongating the notated harmony in the other voices. This is where the performance technique is critical. The lead treble may elaborate on the arpeggio, in the style of a cadenza. The other three voices must hang on their notes until (s)he has finished.

The nine part choir of the last verse is taken from Allegri's original, MSS 205 & 206, with the first tenor part superposed in keeping with the rest of the work. All deviations from Allegri's original are based on this construction, although some are based more loosely than others. However, the pause on the final note is added from MS 185, and Burney's ultimate direction is included.

The plainsong is taken from the liber usualis, and is the full version of the Tonus Peregrinus, on which the falsobordone is based. The speech rhythm has been notated from MS 31395, but only in the first verse of each choir, as in Burney. The text has been checked against the Vulgate, with spelling and punctuation updated.

The crotchet pulse should be felt at approximately MM = 70, perhaps faster or slower according to taste and acoustic. The weight of the two choirs should be even, ideally one to a part in each choir. A lone singer or separate group should (exclusively) sing the plainsong verses. The speech rhythm sections should be sung according to the nature of the words. The work should, of course, be performed unaccompanied, as there has never been an organ in the Papal Chapel (this is, of course, the origin of the expression a cappella).
More ambitious performers might like to try the work up a minor 3rd, or even up a fourth, for maximum authenticity! Substitution of the second choir's bass for a tenor might be advised, but the first choir would present little problem for today's singers. Up to Top

The Performance Secret

"But the manner of performance contributes more to its effect than the composition itself."

By comparing Allegri's original manuscripts with the later versions containing the ornaments, it can clearly be seen what has happened to the piece and how the choir is to sing it. Allegri's original notes are doubled in length, whilst the first treble has elaborated on a simple downward scale. The lower three voices all hold onto their notes, not moving until the first treble gives the very obvious hint of a C followed a C# on two equal note values, thus re-establishing the tempo.

Therefore the first treble may elaborate on the existing ornaments in free time, in the style of a cadenza, while the remaining three parts hold onto their notes, only moving on to the next note only when the treble is ready! It is also worth noting that the top three parts each have ornaments at different times, so allowing each of them to show off their skills while the remaining parts hold onto the chord. Obviously, all the parts cannot extemporise at the same time, as this would lead to chaos.

It is this performance practice which makes the Miserere so special, and which also requires a high degree of musicality from its performers.

Download an expanded version of this essay, complete with musical examples. Up to Top


Ancient Groove Music is pleased to offer three different editions of Allegri's Miserere, reflecting the varied history of this work. We also offer the setting by Tommaso Bai, which features the same abbellimenti used by the Sistine Chapel Choir as those in Allegri's work. Other works by Allegri are also available.

1. A new performing edition created for The Sixteen by Harry Christophers & Ben Byram-Wigfield, showing the evolution of the work from simple fauxbourdon to the Top C version. The initial verses portray Allegri's original unadorned fauxbordon, then later verses detail the Sistine Chapel embellishments, before the final verses conclude with the ‘Top C’ version, best known today, which was never performed in Rome, but it merely a serendipitous scribal error.
| Buy copies |

A recording of The Sixteen performing this edition is available from The Sixteen website.

2. A scholarly edition as described here, providing a authentic rendering of the work as it was performed in the Sistine Chapel in the 17th and 18th century. | Buy copies |

3. Allegri's original setting, WITHOUT abbellimenti is offered as a free download.

4. The setting by Tommaso Bai (c.1650 - 1714), also composed to be sung in the Sistine Chapel, which uses the same abbellimenti. | Info | Buy copies |

<< Manuscript Sources | Biography>>