Mastering Monteverdi’s Orfeo

You’ve conducted a lot of opera around the world, but you haven’t recorded many. The Taverner Consort and Players made two different recordings of Dido and Aeneas, and then there’s a famous recording of the Florentine intermedi performed at a Medici wedding in 1589. But this is the first time the group has tackled something commonly regarded as one of the great full-scale Baroque operas in the studio. Has the experience of conducting Handel, Mozart and Gluck operas at Opera Atelier, Drottningholm and Göttingen fed into your approach to Monteverdi’s Orfeo?

I’m happy to say no, not at all! I feel it’s completely separate, and can’t see any obvious connection between Orfeo and 18th-century operas such as Mozart’s Idomeneo and Handel’s Amadigi – both of which I recently loved conducting in the theatre. In many ways my broad approach to Orfeo is exactly the same – for better or for worse – as when I first did it in concert performances at Bruges and Barcelona, and I’ve also done stage versions in Oslo and Boston, and more recently the Taverners did a concert performance at the Herrenchiemsee Festival in the Bavarian Alps.

My feeling right from the beginning is that Orfeo is not an opera at all in the way we now think of opera. It’s not ‘the first great opera’ – forget that. Monteverdi was 39 when he wrote Orfeo, and up to that point his background was that of a Renaissance composer – he was born in what we call the Renaissance, and he was trained by a Renaissance composer. In 1607 Orfeo was something new for him, and he was blissfully ignorant of all the vices (and also good things) of later opera. In Orfeo there’s nothing clichéd, predictable, routine or degenerate; he didn’t fall into any of the traps that would later be fallen into of making operas vehicles for singing canaries – it wasn’t about vocal display in that sense. Also, in my view, it’s miles and miles away from the much later Venetian operas that he wrote – which don’t do those same things for me. I know they’re good, but they are of a different sort and come from a different world.

Monteverdi’s two editions of the score of Orfeo printed by Amadino in Venice (in 1609 and 1615) and a letter from one of the Gonzaga family all refer to it as a Favola in Musica, which is distinct from what the early 18th century called dramma per musica, or what we’d later complacently simply call ‘opera’. A ‘fable in music’ has an implication that it is not the same thing as a literal narrative drama that you’d find in later opera, whether it’s Giulio Cesare, La Traviata or Albert Herring.

Monteverdi wasn’t consciously making a distinction between those labels because he didn’t know them. Orfeo is poetic rather than natural, and it’s affecting rather than ‘dramatic’. I mean, there is drama in it, and it is appropriate to do it in the right sort of theatre under the right sort of circumstances, but those modern preoccupations about music theatre are not what it was originally about. If you approach Monteverdi as proto-opera – even as proto-late-17th century opera, let alone fully fledged high Baroque opera, you’re going to miss the point of it. It is poetry – a sophisticated and intellectual (though not at all heavy-handed) retelling of a story that everyone knew, and told in the most refined way you can possibly imagine. That doesn’t make it precious at all; it is wonderfully human despite all the baggage of a plot taken from ancient history, and the musical fable is done so magnificently well that during the project I couldn’t stop dreaming about it.

A letter written by one of the Gonzagas explains that the libretto was printed for the first performance so that the audience could read along – so therefore it’s not a translation for an audience who doesn’t know Italian (like Handel’s London opera wordbooks), but for an Italian audience who presumably had no trouble understanding the singers.

Yes, the audience was a small literary academy who convened to savour the poetry, and who were knowledgeable about the story. So they didn’t need an elaborate Technicolor production spelling out ‘this is Orpheus, this is Apollo’ because they were reading the text and already knew.

So you’re aiming for conversational drama within a modestly proportioned room in the Ducal palace in Mantua rather than the bigger spectacle we’re accustomed to in a modern opera house?

Apart from the fact that the 1609 edition of the score has a line mentioning a passage ‘sung behind the curtain’, there’s no strong evidence that I’ve seen that it was ‘produced’ in the sense that there were elaborate costumes, scenic changes and complicated stage action. Apollo sings as he descends, and then afterwards he and Orfeo sing as they ascend together, which implies some kind of machine, but actually none of the documents from the Gonzaga circle say ‘tomorrow there’s going to be a production with all the scenic effects but the costumes haven’t been finished yet’. They talk about the excitement of a fable in music taking place, and that talented young singers have been struggling to learn their parts in time, but there’s nothing in the archives relating to matters we might expect for the materials and preparations required for a theatrical-style production. I suspect by our modern terms it was a very simple production, acted simply, and with emphasis on the poetry rather than on physical action.

The Taverner Consort and Players’ performance of Orfeo seems much closer to the spirit of an extended poetic madrigal, or even the genre of intermedio imported from Florence.

Yes, absolutely. The way I see the score of Orfeo is that it has at least one foot in the intermedio tradition – that’s what the score, instrumentation and other indications tell me, though of course it took the genre to a new degree on the back of all the experiments and intellectual thought that went into the reinvestigation of ancient Greek principles of declamation, and so on. The miracle is not just that Monteverdi combined those two things, but that it was combined so brilliantly combined. He came up with something stunning!

Also, the connection to Florence is important. The court at Mantua looked in some ways to Ferrara as a cultural model, but it also looked in particular to Florence. That’s why we ignored Venetian or Milanese pitch, considering them probably irrelevant to what Monteverdi was doing in Mantua; A=440 is a demonstrable historical pitch for Florence, and probably close to what Monteverdi was using in Mantua. Also the slightly lower pitch helps our singers to be within speaking range of their voices – because the higher you go the more artificial and unnatural the exclamation becomes. 

We chose to use modern Florentine Italian, which is regarded as the equivalent of our Queen’s English, Oxford English, or BBC English – whatever one of those we’re supposed to use for English! I made the policy that we wouldn’t try to do historical pronunciation. It’s partly practical and economical – rehearsals would get much less far musically if we were worrying about spending lots of time on historical pronunciation. But actually, modern Florentine Italian is not so far from that anyway, and our specialist language coach Giovanna del Perugia is actually from Florence, and lives not far from Oxford, so it was terrific having her on hand throughout the project. In rehearsals she was a bit cautious at first, but then in the recording sessions she was terrific to work with because she got to know everybody and became more extrovert in the way she ticked people off saying ‘No, no, any Italian would think you are rubbish! It should go like this…’!

Another interesting feature of The Taverner Players’ approach is the actual instrument types used by the string players. In one short passage the violinists were using new instruments constructed especially for the project, and the cellists were actually playing bass violins (resting them on a chair), holding the instruments in their arms but with a strap for convenience.

There is one short passage where Monteverdi specifies two ‘violini piccolo alla francese’, so Oliver Webber and I spent months trying to figure out what the heck that meant, sending emails back and forth, and trying to work out whether they were little violins, and, if so, what was their tuning? Or were they members of the rebec family, more like the pochette – and there are two different types of pochette as well. In the end, we couldn’t prove things for sure either way, but we went for the pochette, and two were specially built following an instrument that has surfaced relatively recently in Italy, from slightly later in the 17th century – so it has some pedigree.

As for what might be called the cellists, there were no cellos at that time, but the bass violin, which was the bass member of the violin band at that time. Whether it makes a big difference or not, I believe that the part above that in the five-part writing, which is written in the tenor clef, is not for a viola, but for an intermediary size instrument between a viola and a cello, tuned a fourth or fifth between viola and cello; I suppose we might call it a tenor violin. The way you play it, whether da spalla (at the shoulder), between your legs, or on a chair is irrelevant. It’s the tuning and the function of it that’s important. So Monteverdi’s string band had a Renaissance constitution of two equal small instruments at the top, the next part down as one size down – the equivalent of the viola – and then this instrument tuned a fifth or a fourth between the modern viola and the modern cello, and then the bass violin at the bottom. The double bass is not part of that string band except occasionally.

In the five-part string ritornelli you got your keyboard and plucked continuo players to drop out most of the time.

You say ‘drop out’, but I’d say they’re not necessarily supposed to be there in the first place. They’re not needed. I’m not saying I’m absolutely right and this is what Monteverdi definitely intended, but I certainly think it’s a very plausible option and it has the advantage of pacing things better, or giving a forward motion as you hand over from one instrumental ensemble to another: we flow from the singers with their continuo accompaniment to the side-stage string band in the ritornelli.

I noticed that the cornettist Gawain Glenton was serving throughout the project as an ornamentation guru to all the other players and singers.

 That was very useful. The person who has made ornamentation of music from this period really come to life is [cornettist] Bruce Dickey, who is a scholar and good friend of mine. He has looked at the original treatises and has now taught new generations of cornett players, including Gawain and Jamie Savan, and they all know off-the-cuff how to embellish early 17th-century music really successfully. Gawain didn’t have that much to do as a cornett player in Orfeo, but his wife Kirsty was playing the harp, so rather than have him hanging around I thought I’d ask him to discuss ornamentation with the singers. One thing he brought to the project that helped enormously was that he mentioned if you look at Caccini’s ornamented versions the ornaments often don’t happen where we think they should happen. We assume we have to do something on the penultimate note of a cadence, but that often sounds very contrived and ends up with seeming like an ornament we’ve heard 1000 times before. But Gawain observed that often the embellishment has finished by that point, so the penultimate note is almost totally free and rhythmical, so that has implications on how the continuo plays at the cadence, and the music therefore avoids the problem of stopping and starting every line, and thus flows more naturally. Getting away from embellishment as a predictably cadential function, and instead using it as something special to emphasise particular words helped our approach to the whole piece enormously.

Francesco Rasi, the original Orfeo, is sometimes described in scholarly literature as somebody who could sing either bass or tenor parts, so one increasingly common modern solution is to cast a baritone in the part. But you’ve chosen Charles Daniels – a Monteverdian of good pedigree – who has a resolutely high tenor timbre.

I think our modern definition of what baritones are, as opposed to tenors, is spurious. Rasi does seem to have been basically a tenor in the early 17th-century use of the term, in that he had a good low range. The part itself is absolutely a standard tenor range; it’s not a low tenor part. I’m sure Rasi could’ve sung lower, and there is some Caccini he sang that does go lower than Orfeo, but a tenor part in the early-17th century is not the same as a tenor part now. Their definition of a tenor was ‘the ordinary voice’, in terms of range – which for us nowadays we might define as a baritone because we’ve decided to categorise things that way. We have bass-baritones – although, curiously, we don’t have ‘tenor-baritones’ – that’s just the way we think, like we don’t seem to have contraltos anymore because they’re all called mezzos! But human beings haven’t necessarily changed all that much – it’s just the way we use and categorise them.

The standard 16th-century alto clef was usually sung by the voice that we now call a tenor who goes up to Gs and As. Whereas the tenor part written in a tenor clef did not have the highest notes that are now sung by modern tenors, but nonetheless the part of Orfeo is composed in a standard tenor range for the 16th and early-17th centuries. John Eliot Gardiner transposed Orfeo up a tone to fit Anthony Rolfe Johnson, which is fair enough. I wouldn’t do that because it skews the instruments playing in the wrong keys, so I would not contemplate changing keys in order to accommodate the singer in a piece like that.

So that leaves us with the question of what kind of modern voice should sing Orpheus. Charles Daniels happens to have sung some very high tenor parts in his early youth – he sometimes sang alto parts in this repertory, with effortless high As and beyond. But in fact he’s always had very good low notes, down to a low A. The range is not a problem for him at all, and what that means in Orfeo is that he is singing in the speaking area of his voice, and his top notes are not unduly forced or difficult – which, in Monteverdi's time, would have been considered vulgar and undesirable. Any hint of strain would have been inappropriate, and the newly emerging tradition of Italianate bel canto was that the top should sound easy – whereas the French tenors went on pushing for another century and more. The French liked loud singing, but the Italians were fonder of soft singing. So our approach comes from the tradition of soft singing – not church singing, but chamber singing.

But for Apollo, who has a similar vocal range to Orfeo, you have cast a baritone!

Our young Israeli baritone Guy Pelc got better and better during the project. My dilemma was that Apollo is Orpheus’s father, but is also eternally youthful, and is at least as good a singer as Orpheus, if not better. We haven’t got costumes to differentiate them on a recording, so if you took Charles Daniels and a clone, or double-tracked him, the listener to a recording wouldn’t get the idea of the two contrasting characters. So I was really struggling to think of who the right person was, and then I heard Guy in Israel and got him to sing with us. He’d never been exposed to this sort of stuff before, and it was great!

He’s really musical, very intelligent, has got a good voice and fine technique already, and I think he’s got a really outstanding career ahead of him. He was open-minded, and there was one phrase for Apollo for which we tried a kind of ornament he’d never done before, and he picked it up amazingly quickly and just did it straight off. Even with the best, most experienced and open-minded kind of singers, it’s very difficult to get that feeling of improvisation and freshness. Ideally I want both maturity and freshness from my singers!

Recordings of Orfeo usually have a soprano singing La Musica in the prologue, but you’ve chosen the high countertenor David Hurley. 

David Hurley sang in the Taverner Choir when I did a Handel oratorio recording years ago, but I heard his recording of Barbarino devotional motets and was impressed by two things in particular. First, here is a man singing in soprano range that doesn’t sound strained, but comfortable and natural at the top of his voice. The second thing is that he can articulate vocal runs beautifully and very delicately. Ornaments can be sung like they’re in parentheses; some singers articulate a run very deliberately as if to make the audience notice every individual note of the flourish, but it’s not supposed to be like that – ornaments should be simple gestures. But the main reason I wanted David Hurley to sing La Musica was that I wanted to start the opera in a way that felt like an intermedio. We might see the name La Musica and think she’s a person, but it isn’t like that – it’s an allegorical figure. I’m not against a female soprano singing it now, but I needed to characterise the voices on the record very clearly. I wanted to achieve a clearly audible distinction between the soprano-range parts. Moreover, Monteverdi’s La Musica was sung by the adolescent castrato Magli, a teenager described as ‘a youth’. Although he had been castrated, his voice might have been rather closer to a boy’s than a fully-fledged operatic castrato such as those we imagine singing for Handel over a century later.

Christopher Purves is another delightful cameo from an ex-Taverner Choir alumna.

Chris came back from doing George Benjamin’s Written on Skin in Aix-en-Provence and sang Plutone for us. Chris read English at Cambridge, and is a very intelligent dramatic actor, and he understands character and drama. It’s not a thunderous rumbling bellowing Plutone, but a mature performance that is like he is genuinely speaking. It’s a very short passage, and it’s quite tricky to know how to play the God of the Underworld, who in the space of only a few lines admits he loves Proserpine, and responds to her seductive imploring that he should let Orpheus take his dead bride back to Earth. So Chris’s experience of flesh-and-blood characterisation is useful. In most operas the music tells you about the sentiments of the character in boxed numbers, whether it’s rage or seduction, but Monteverdi’s recitative needs someone who is really on top of it linguistically and dramatically.

Any performers tackling Orfeo have to deal with the issue of instrumentation because there are different schools of thought about what you do, when you do it, and how much. I noticed that facsimiles of both Monteverdi’s 1609 and 1615 editions and the original Mantuan libretto were all floating around between the musicians during the project. What decisions did you make?

The two editions are the same in terms of instrumentation, but the actual list of instruments Monteverdi gives at the front are slightly different. There are contradictions between the two sources too numerous to mention, but I think there’s a lot that you can deduce from it if you understand particularly the Renaissance conception of intermedio scoring. You can see how Monteverdi is thinking, and it’s relatively clear that the brass is symbolically separate from the strings in being associated with the underworld as opposed to the Thracian fields.

There are other things that I think are more frequently misunderstood, but the important question is really the accompaniment of recitative, which is almost all just continuo – not strings and continuo. I think it’s wrong to assume it’s a free-for-all when Monteverdi didn’t specify something. When he doesn’t specify, it doesn’t mean you can do anything you like. When he does specify something, it means there is a departure from the norm. So the question to ask is what the norm is. I don’t wish to knock other recordings or performances, but there is a general tendency nowadays to revel in large continuo sections of multiple instruments – not just one or two, but four or five – who wring the changes almost constantly. Well, I agree changes should be wrung, but infrequently in order to make very specific points tellingly.

I think the over-activity of multiple continuo instruments that change texture on almost every line is dangerous.The singer needs the timing, pacing and judicious amount of volume to support what they’re doing, but if the idea is duplicated or pre-empted by over-excitable continuo it undermines the singer, and anything instrumental that jumps up and down clamouring for attention in the wrong places runs the risk of cutting against the poetry. So I wanted to have a continuo style that offered lots of flexibility to support the focus on language rather than using all the players at once, which would detract from it. I chose to have just one or two continuo instruments accompanying the solo singers, moving backwards and forwards with that singer, rather than having to organise this trundling machine of lots of pluckers and rich colours all the time.

So many reviews of modern performances of Baroque opera legitimately describe the thrilling or exciting accompaniments of the continuo team, but for me the best review of a continuo team is that you don’t notice them: they are playing stylishly, beautifully, very sensitively, but the listener is concentrating on the voice. Although various instruments are at our disposal, I think using them sparingly creates a greater effect and draws the listener in. I don’t see it as my job to spray-gun the audience with ideas, but actually to draw them into what the singer, or actor, is doing. It’s a fundamental aspect, but the modern world of opera doesn’t often do that, and tends to do the opposite.

The Taverner Consort and Players' new recording of Monteverdi's L'Orfeo is reviewed in the June 2013 issue of Gramophone. Click here to buy the recording from Amazon

Tweet