February 3, 2019


Frescobaldi and Early Italian Organs

February 3, 2019


Frescobaldi and Early Italian Organs

Frescobaldi's Bergamasca from Fiori Musicali, first edition.

Frescobaldi's Bergamasca from Fiori Musicali, first edition.


Scholar Christopher Stembridge discusses his new editions of Frescobaldi's music and extensive experience with early Italian organs with Vox Humana Associate Editor Guy Whatley.


Christopher, many thanks for agreeing to this interview. To begin, could you talk about your background? As one of the leading scholars on Italian keyboard music, how did you develop such an interest this repertoire?

To answer your last question first: when I was growing up, the music one heard and played tended to be rather Northern-European-centered, Bach, Mozart, Beethoven, Brahms with maybe a bit of Vivaldi thrown in, so perhaps my later interest in Italian music was to some extent a reaction to this environment. I suppose I grew up with these preconceptions. At the age of eight I went to boarding school and sang in the choir at Lichfield Cathedral, and amazingly we didn’t sing any Palestrina — in fact, I sang there for five years and we probably didn’t sing a note of Italian music (although we did of course sing some wonderful music such as Byrd and Gibbons). I remember the great Italian organist Fernando Germani coming to give a recital in Lichfield, and I'm sure even he played mainly Bach and César Franck! The music scene in Munich when I was studying there was very German-centered — and certainly for the Germans, Verdi was never considered as important as Wagner. And French music didn’t get much of a look-in at all. I actually gave the first complete Munich performance of Messiaen's Nativité some 33 years after it had been written!

The c. 1580 anonymous organo di legno at the Silberne Kapelle in Innsbruck, Austria.

But on the whole, when playing the organ, I don't think I was particularly interested in huge instruments with lots of stops, and I remember even in my university days thinking that the Italians had basically got it right — a single keyboard, a few stops, and they just played the music. This "small is beautiful" idea was very much in vogue in the 1960s, so I related to that. Although I studied languages (not music) in Cambridge, I was organ scholar at Downing College, which had a two manual Bevington organ with only three stops on the Great — and they were all eight foot! And in fact I’m rather grateful for that experience, because if that’s what you have, you try very hard to make it as interesting as possible.

My first meeting with the early Italian organ music was in 1972, the very first year they had the summer early music course in Innsbruck. I went to Tagliavini’s class based around Frescobaldi's Fiori Musicali, which was held on the only surviving organo di legno from the sixteenth century in the Silberne Kapelle. After some summers at the Haarlem Zomerakademie — where I started studying harpsichord with Kenneth Gilbert — I later went to Tagliavini's course in Pistoia, which opened the door to a wider range of early Italian keyboard music including that of Ascanio Mayone, the only copy of whose Libro Primo is in the British Library. I transcribed that, and with Kenneth Gilbert's encouragement made an edition of it. I also did some research work in Oxford on Italian vocal music of the period, reconstructing the missing voice in Frescobaldi's madrigals and motets.

The c. 1580 anonymous organo di legno at the Silberne Kapelle in Innsbruck, Austria.

Do you recall the first time that you played on a historic organ? How did that impact your interpretation and understanding of touch?

The 1519 Giovanni Piffero organ at Santissima Annunziata in Siena, Italy.

It was actually not an organ that first opened that world to me. When I was a student in Cambridge I saw an advertisement in the Musical Times — someone was selling an inexpensive Alec Hodson clavichord, so I bought it. It was simply wonderful to have an instrument in the room, and I took it to Munich with me when I studied there. My room had a bed, the clavichord, there were all nine volumes of the Grove Dictionary (1953 edition), and a table and a stove — and that was my world. I spent hours and hours playing through things like the Fitzwilliam Virginal Book on the clavichord without thinking anything about it at all. Now I think, without a doubt, that the clavichord was my most important teacher as far as touch and articulation is concerned. I have seen the same thing happen to my students. When I was teaching a summer course for the Chigiana in Siena on the beautiful 1519 Piffero organ there, most of the students were Italian, so they should have been conversant with such historic instruments. But the exception was someone from Estonia who had not had any organ at her disposal, but had prepared everything she played on her small clavichord, and her touch and articulation put the all the others in the shade.

The 1519 Giovanni Piffero organ at Santissima Annunziata in Siena, Italy.

In North America we have many historically-informed instruments, but the vast majority are modeled after organs in Northern Europe. How can we play Italian music on such instruments with their very different voicing, actions, and overall ethos?

I think you just have to take every organ on its own merits and see what is going to work. One thing I often find myself doing, particularly in something like an elevation toccata, is to play it on a four-foot principal an octave lower; for the missing bottom octave, I then find an eight foot stop in the pedal that matches. Of course you need to be aware what kind of sound Frescobaldi had in mind, but I think you can avoid being pedantic.

The generally-accepted concept of the early Italian organ is almost always based on the Antegnati instruments, which didn't have stopped pipes or reeds. But this is not true of other organs of the period in Italy. So once we realize that in Frescobaldi’s Rome there were many interesting instruments around, including several with two manuals, we should feel free to do what makes musical sense. The idea that Italian organs are tiny and lacking in color goes back to Mattheson, who thought that the stoplist of San Marco was ridiculous, and how could you make music on that! He didn’t appreciate what a 24-foot organ with five octaves could do!

You published an edition of Frescobaldi's Fiori Musicali in open score, just like the first printing, but with modern clefs. How does notation shape our performances and musical understanding? Do you think we lose anything playing from modern notation?

The idea of producing an edition of Fiori Musicali in open score was really to share something with my students. When teaching, I was always getting students to write out ricercars in open score, in either old or modern clefs, whichever they preferred. I like to encourage students to play from open score, partly just because Frescobaldi tells us this is what we should be doing.

I was criticized because I put this edition into modern clefs, but I did this simply because I wanted to encourage people to play from open score. Frescobaldi says in the 1624 preface to the Capricci that the practice of playing from open score was dying out. Clearly he wanted to keep it alive, mentioning it again in the Fiori. But is interesting to note that in Fiori Musicali, unusually for a set of polyphonic pieces in this period, he always uses the same four clefs throughout, and consequently avoids certain modes. I think one reason he did this was to make it simpler to read the whole book in open score. There are some Kyrie versets where one voice just sits on the same note the entire time — presumably to make them even easier to read.

Recently, Italian music publishers have fallen into the habit of putting contrapuntal keyboard music into pseudo-intavolatura [two staves] where you end up simply not knowing what the parts are doing [if voices are crossing]. A lot of music has been recently published that wasn’t available except for the manuscripts and facsimiles, but as far as I’m concerned, it is still not really available if no one can see what the counterpoint is doing. The new Bärenreiter edition of Frescobaldi will be on two staves, but I would hope that Frescobaldi’s ghost could take a look at The Well-Tempered Clavier and realize that you can actually write four parts on two staves clearly (which is what we are trying to do). Sometimes people complain about things like too many ledger lines, but that’s what you need to do to make the part writing clear, which seems much more important to me than ease of just playing the right note at the right moment.

Speaking of your new Frescobaldi edition for Bärenreiter, what can you tell us about it?

When the Bärenreiter project started off I was going to do it together with Kenneth Gilbert. We had wonderful discussions and arguments at the beginning — I’ve written about this in the Routledge book Perspectives on Early Keyboard Music and Revival in the Twentieth Century (a Festschrift for Kenneth Gilbert). For the last two years I have been working on the last main volume which contains both the Fiori Musicali and the Aggiunta (Frescobaldi's appendix to the final reprint of his First Book of Toccatas). One of the new things about this edition (which was my imperative) has been to put the pieces into chronological order, as there is so much confusion about what works first appeared when. At first, I was slightly embarrassed about the juxtaposition of the Fiori and the Aggiunta, but now they are together I’m absolutely thrilled because there are resonances. After all, the very last piece of the Aggiunta is for organ, and I certainly think that harpsichordists should be playing things from the Fiori Musicali.

While Fiori Musicali has tended to be regarded as inextricably linked to the Catholic Mass, I suggest that people look closely at Frescobaldi’s preface. He says twice that, although he has set these verses to the Kyrie, you can play them on other occasions. Oddly, he also says that his main idea for this book is to provide the organist with music to play at Mass and Vespers, but there’s nothing in there for Vespers (although there is in the Second Book of Toccatas) and frankly I think that sentence is somewhat at variance with the rest of the preface. This whole book is didactic — you learn to play from open score, and Frescobaldi's text at the beginning of the tricky Bergamasca states that whoever plays this piece will "not have learned little."

The title Fiori Musicali had been used by many composers previously mainly for anthologies of secular vocal music. The word "fiori" was traditionally used to mean an anthology — in fact it relates to the notion of "anthology" (which is based on the Greek word for flower, "anthos"). If you look at the title page, there is no mention that the contents are for liturgical use; there is not even any indication that they are for organ versus any other instrument. It is also very clear that the pieces were not written at the same period or even for the same instrument: for example, some pieces are composed with a short octave in mind while others use the chromatic notes. I believe that this book is simply a collection of the best pieces which Frescobaldi had not yet published and, as he approached the end of his life, he needed to get them out. It could be that the dedicatee, Cardinal Barberini, requested something useful for organists to play at the Mass, and so maybe Frescobaldi merely adopted the pseudo-liturgical presentation either for that reason or simply to ensure good sales. Interestingly, Frescobaldi has based the Kyrie verses on many different variants of the relevant chants, and so in this edition I have put these in front of each verse — not for liturgical use, but to encourage the player's understanding of Frescobaldi's artistry in his many ways of treating a chant. The edition also includes some lovely facsimiles from some of the different graduals of the period.

The fact that you are living in Italy and producing an edition of Italian music for a German publisher reminds me that you mentioned earlier that you studied languages at Cambridge. That seems to have been a useful choice.

Today it might seem unusual, but in those days (at least in England) it was quite normal to think that the job of University was to educate you, not necessarily to set you up as a specialist. Thurston Dart studied math, David Wulstan biochemistry, Peter Williams English. My father had pushed me in the direction of learning German as an option at school "if I was going to be a musician" and, yes, I am eternally grateful for that. My other language at University was French, and in those days you could not enter Cambridge without Latin, so it was knowledge of Latin and French that enabled me much later to read things in Italian.

I think it is also really a great advantage not only to be able to read things but also actually to spend time in countries whose music is important to you. I suspect that hearing Italian spoken around you, with the lively flow of the language affects the way you play a toccata. You can take note of the intonation and enunciation of the language and the way it is expressed and articulated. And then, for instance, if you consider the different voicings of a German organ versus an Italian organ, the German one will have higher wind pressures and rather harder voicing. The initial 'chiff' is somewhat analogous to the use of hard consonants like “K” and “Z” in the language, whereas Italian, of course, is much more directed towards the vowels; this seems to be reflected in the more vocal gentler Italian sound.

The other side of the coin is that music is certainly not a language but something that is not bound by space or time: I think travel helps one to appreciate that.

But you can't travel in time.

Well, it is amazing what you can do. For instance, that little organ in Siena I was talking about earlier is still standing where it was put in 1519, surrounded by much earlier frescoes. And of course the space hasn't changed at all, so it is still in the acoustic for which it was built. The pipes had not been touched much over the centuries and so, since it was therefore possible to work out the original wind-pressure, we have not only a sound which is probably very close to what it was in 1519, but we have the action. It is very moving to sit down and play an instrument like that — it becomes a real time machine. That said, it can't tell us how people played on it nearly five centuries ago. And while it is necessary to study the historical context of music in order to understand it, and of course look for all the clues we might find in the written notes and in what someone like Frescobaldi writes, I think there is a real danger of putting "early music" into a separate box from music-making generally. For instance, when I am asked about my free approach to the Toccatas, I like to think that that comes partly from my own early contact with very different music.

I'm a Worcestershire lad and I grew up absolutely in love with the music of Elgar, and I would listen to records of Elgar conducting his own music (for example, the fantastic recording of Menuhin at the age of 16 playing the Violin Concerto). I would go cycling around Worcestershire (very much as Elgar did), and I would race down a hill and trudge up the next one very slowly. I feel that this, and listening to those performances, particular the Beatrice Harrison recording of the Cello Concerto with Elgar conducting, really influenced me. There is an extraordinary range of color and tempi, and that always grabbed me. At one stage I was too free with music that needed more regularity. I remember when I was studying in Munich my organ teacher thought that I should be using a metronome, and I simply don’t think that is the answer. I think you need to find your level in freedom, and that there will be a regular beat somewhere, but it will not be at the level of the sixteenth note. But you need time to settle down to that and find it yourself.

What are some other performance practice issues in this music?

I have become especially interested in the amazing number of ways that Frescobaldi uses triple time. Duple time just comes in one form, which of course does not mean it is all played at the same speed, but triple time comes in multiple variants, and it seems Frescobaldi was at pains to notate subtle differences in the intended tempo. The crowning achievement of this is the Cento Partite, and reading literature on the origins of the passacaglia and the ciacona one realizes that they are totally different animals. Now when I play the Cento Partite I enjoy the contrast between the passacagli, which started life as improvised ritornelli, and the lively dance which was the Spanish chacona. Of course there is always confusion about the difference between a chaconne and a passacaglia, but nearly all the discussions that people read are centered on post-Frescobaldi usage and nobody seems to be aware of the history of the rudimentary pieces that found their way from Spain to Italy in the sixteenth century when a plethora of popular books for Spanish guitar were published. It is useful to look at Thomas Walker's article "Ciaccona and Passacaglia: Remarks on Their Origin and Early History" in the Journal of the American Musicological Society 21 (1968), 300–320, and Richard Hudson's Passacaglio and Ciaccona: From Guitar Music to Italian Keyboard Variations in the Seventeenth Century (Ann Arbor: University of Michigan Press 1981).

Although I have some very distinguished colleagues who would disagree with me, I have become really convinced that when you have these strange juxtapositions of note-values in Toccata Nona (Book II) for example, with perhaps twelve notes in one hand and eight notes in the other hand, you are not supposed to play equal notes in each hand. That creates confusion and to my mind is rather unmusical, but you modify the rhythm of one of the parts. After all, when you look at Frescobaldi’s toccatas, you soon realize that enormous care has been lavished on the precise notation, and that subtle differences occur constantly throughout the text. But having gone to all of this trouble with the notation, the very first thing Frescobaldi says in his preface is that this music is not subject to a regular beat. Of course, this is the case with recitative anywhere, but I feel there is something rather Italian about this approach, analogous to the bureaucracy that reigns in Italy: everything has to be absolutely correct on paper, but in performance things are allowed to become more free, in the same way that the Italians know how to enjoy life in spite of the bureaucracy. Frescobaldi goes on to say that when it comes to a trill in one hand you don’t play what is written, but you play more notes, and even though each sixteenth note of the phrase in the other hand corresponds on paper to one note of the trill, the phrase is played slowly and with affect. So surely once you have accepted that you don’t take this notation literally it hardly makes musical or logical sense to play the conflicting rhythms mentioned above exactly as they are notated. You have to see which notes fit together, and there is almost never only one solution. I always find consulting Michael Collins's dissertation The Performance of Coloration, Hemiola, and Sesquialtera [1450–1750] (PhD dissertation, Stanford University, 1966) helpful when revisiting apparent problems of three against two.

What are your thoughts on historic tunings?

As with modern performance, there is the danger of a rather categorical approach. I really don’t think that meantone should always be understood in the very specific sense in which the thirds must be pure while all the fifths should sound identical, being narrowed by a quarter of a comma.

There are two very good sources from around 1600 that everybody quotes, but it would be good if they actually read them instead of just quoting them. One is Antegnati and one is Praetorius, and so we have the north and the south represented. If you read Antegnati, he indeed says that you make the thirds as pure as possible, but before this he has already talked about the fifths, stating that you should make them slightly narrow so that you can hardly notice it, and that does not really seem to be a description of a quarter-comma fifth! Then he says that the thirds should be as pure as possible, but he is not asking for them to be absolutely pure. If you follow Antegnati’s tuning instructions you start from F, first tuning A as a perfect third and then proceeding up by fifths; when you arrive at A, Antegnati says it should be in tune because you have tuned it as a third. But he doesn’t say don’t touch it. Similarly, you tune B a third above G, and when you come to B again by way of the fifths, he says the same thing. But when you play G again (to tune E flat), he says that you should not touch it. I think that is very significant because there would be absolutely no reason to think of changing it unless you had modified the tuning of the B since you first tuned it, spoiling the pure third. We can thus surmise that the thirds were not necessarily perfectly in tune. The last note that he tunes is B flat, and he tunes it as a fifth against E flat, and he doesn’t check it against D and he doesn’t check it against F, which suggests that these intervals might not be so well in tune!

Again, if you read Praetorius's Syntagma musicum carefully, he describes a pure meantone tuning, but he goes on to say that he has given you here a simplified version for the simple-minded organists and organ makers. He then says that he hopes that God will give him enough time to write another volume, so he obviously intended to go into much more detail. Again, I would see this as evidence for not sticking to a totally regularized tuning. Some years ago a symposium was held in Dresden because one day they hope to reconstruct the organ at the Schlosskapelle, which is where Praetorius was. The consensus was that this organ should be pure quarter-comma meantone, but I had to disagree. My biggest argument there was a toccata by Hans Leo Hassler, who of course was the organist of the Schlosskapelle when this organ was being designed and built, which starts off with a long held open fifth, G and D, (i.e. with no third) and I cannot imagine any musician being satisfied with that opening being played on a fifth narrowed by a quarter of a comma.

There are also two organs of the period that I know of — one is the Badia in Florence, one is the Carmine in Brescia — where we can probably accept that the pipes in the façade are all of the original lengths. In both those cases they actually give us the idea that G to D was a better fifth. The other thing is that the Italians were very insistent that all the modes have a different character, and so if all the modes have a different character, why should they all have identical tuning? A lot of music in the second mode finishes without a third in the final chord, just G and D. This interval is also found melodically in the seventh and eighth mode, which is supposed to be serene — no one would want to sing it if it was tuned like that! There is also a handwritten addition in a copy of Diruta’s Transilvano (kept at the Biblioteca Casanatense in Rome) suggesting some thirds are pure, while others are not pure, but some of which should be almost pure.

All this suggests that there were many variants of slightly unequal, if basically meantone, temperaments at the period. On a harpsichord we can adjust this daily if we want to, but of course we can't be knocking organ pipes around. Each instrument will need a tuning that suits it. It seems a pity that most restorations of historic organs of the period have insisted on regular quarter-comma meantone without question.

On my organ at home, I have the F major triad tuned with a very good third, but a not perfect fifth, and the G major triad is tuned with an almost perfect fifth but with a poorer third. And so they are both good, but different, and I just love playing pieces like William Byrd’s The Woods so Wild on that organ. I have just commissioned a new instrument by Bernhard Fleig, who built the arciorgano in Basel [see this Vox Humana article for more]. It is a little four-foot organ with the pipes based on those at Innsbruck.

What are you working on now and what new projects do you have planned?

I am continually fascinated by the different characters of the modes. It is absolutely critical for Italian music, but it is also completely relevant for the music of Bach. I have been trying to get to know all of the cantatas better and am trying to make a sort of database of all of the movements organized by year, key, instrumentation, time-signature, and tempo. This enables me to find, for instance, all the pieces in G minor that are in 3/8 time, and I keep noticing extraordinary things. Again and again one finds the same details, even the same musical ideas cropping up in pieces in the same key. Obviously the cantatas have texts, and so it is much easier to identify the effect that relates to the musical texture. For example, C minor often tends to be about resignation, about positively accepting one's fate. I think anybody who is going to play Bach's keyboard or organ music, indeed any music by Bach, needs to be aware of such characteristics. But all this is another story for another day, and there is way too much material there even for a book. This will fascinate me until long after I am dead!


Christopher Stembridge studied languages at Cambridge University and musicology at Oxford University. He was awarded the Turpin Prize on obtaining Fellowship of the Royal College of Organists. After 20 years as a lecturer in music at University College Cork (National University of Ireland), he moved to Northern Italy where he now lives. He gives regular master-classes on Renaissance organs and travels widely giving lectures, recitals and seminars in European, Russian and North American universities and conservatoires. For ten years he was Professor of Organ and Harpsichord at the Scuola di Musica Santa Cecilia, Brescia, and Professor for Organ at the Accademia Chigiana, Siena 1994–1996; he taught at the Montreal Summer Academy for Organ in 2003. In 2006 (Summer Semester) he held a guest professorship in harpsichord at the Gnesin Academy, Moscow. He has recently directed concerts with the early music group Insula magica in Novosibirsk, and the Baroque Orchestra of Armenia in Yerevan.

His special field of interest is Italian keyboard music of the renaissance and early baroque. He has published various papers on early music by de Macque, Mayone and Frescobaldi. He is currently preparing the new Bärenreiter edition of Frescobaldi’s organ and keyboard music for which he was recently awarded the Noah Greenberg Prize by the American Musicological Society. He wrote the chapter on Italy for the Cambridge Companion to the Organ (Cambridge University Press 1999).

He also plays the clavichord and the cembalo cromatico, a harpsichord with 19 notes per octave, a reconstruction of an instrument that was widely used in late sixteenth-century Italy — and together with Willard Martin of Pennsylvania has designed the reconstruction of a 19-note just intonation keyboard similar to that proposed by Descartes in 1643.

The views and opinions expressed in this article are those of the author, and do not necessarily reflect the position of Vox Humana.