The Oldest Voice

Note: This week’s post is longer than usual, because its subject gets more interesting the farther you look into it - history buffs, enjoy!

Peter Schram - portrait by unknown artist

When we speak of the most ancient recorded voices, we usually mean Tamagno and Maurel, who sang in the premiere of Otello, or maybe Marianne Brandt, the first Kundry, or Adelina Patti, the reigning queen of (old-fashioned) opera for decades following her debut in 1859. But the earliest opera singer we can hear today was already performing lead roles before any of those was even born. He himself was born two full centuries ago, was trained by an Italian tenor older than Rossini, sang with Jenny Lind and with Wagner’s muse Wilhelmine Schröder-Devrient. He collaborated with Hans Christian Andersen. Søren Kierkegaard went to hear him at the Royal Theater in Copenhagen.

Record-history specialists will already know the name under discussion: Peter Schram, or more completely Peter Ludvig Nicolai Schram, 1819-1895, whose voice was captured in a primitive cylinder recording on 5 September 1889, his seventieth birthday and the occasion of his operatic farewell as Leporello in Don Giovanni, a role he had been singing at the Kongelige Theater since 1844.

The Cylinder

Schram as Leporello

Schram’s recording was made not during the performance but at a party afterwards, hosted by the diplomat and merchant Gottfried Rubens, who had recently become the Scandinavian agent for Thomas Edison’s “perfected phonograph.” (The Edison marketing tour of 1888-1889 is the source of all the earliest cylinder recordings we can hear today.) Rubens was, as far as we can tell, selling machines and not recordings at this time, but he recorded many actors and musicians in Copenhagen for the purpose of demonstrating the new mechanical marvel at his shop, and his collection of cylinders is the earliest substantial group that has been preserved through the years. Peter Schram’s contribution consists of Leporello’s opening solo “Notte e giorno faticar” and the first pages of his aria “Madamina, il catalogo è questo,” sung in Danish and without accompaniment.

As in archeology or paleontology, so in music: the more ancient the artifact, the greater the fascination with even fragmentary traces. So everybody was interested when the prolific music writer Henry Pleasants got word of the improbable relic about forty years ago (until then it was unknown outside a small circle of Danish collectors). Pleasants enlisted his friend Harold C. Schonberg, chief critic of the New York Times, to spread the news. People were curious, but the actual listening proved disconcerting for many.

Here is what Herr Schram sang that night 130 years ago, with the Danish text and English translation (big thanks to my polymath brother J.D. Crutchfield for help with this and the other translations from Danish).


Sjelden Penge, Prygl des fleer!
Staae om Natten udenfor
Medens herren inde leer,
Det var hidtil mine Kaar!
Hvorfor selv ej Herre være
Fanden være Tjener meer!
Det for galt er, paa min Ære,
Mens han Elskovslønnen henter
Jeg paa Gaden staaer og venter.
Herre kan jeg gerne være,
Fanden være Tjener meer!
Hvad er det? – Der kommer nogen.
Det er bedst, jeg søger Krogen,
Ganske stille staaer jeg der!
Hvis min Donna, de behager at høre
Denne Liste, som jeg har at føre
Paa de Skjønne, hans Kunst monne røre,
Smukke Ting da, de skal faae at see!
Først i Italien et hundred og tyve!
En Snees Tydske, for ikke at lyve.
Frankrigs Pigebørn rundt om ham flyve;
Men ved Spanien, staaer tusind og tre!

Little pay, frequent beatings,
Standing watch outside at night,
While my master laughs within,
This was hitherto my lot!
Why not be a lord myself?
To hell with any more servitude!
It’s too unjust, on my honor!
While he collects the wages of love,
I stand in the street and wait.
I’d gladly be a master.
The hell with any more servitude!
What was that? - Someone’s coming.
I’d best look for a corner
And stand there quite silently.

Listen, my lady, if you please,
To this list I have to keep
Of the beauties his talents have touched,
Lovely things are there to see!
First in Italy a hundred and twenty,
A score of Germans, to tell the truth.
French girls fly all around him,
But in Spain are a thousand and three!

The easy part for Pleasants and Schonberg to describe was that Schram still used turns, appoggiaturas and cadenzas just like those written down for students by Mozart and his contemporaries. That was interesting. But the singing...

Well, first, anybody old enough to have worked in a traditional European repertory house has encountered this type: a basso getting on in years, no longer presentable in roles that depend on fine singing, but sufficiently sturdy of voice to get through character parts in a kind of Sprechstimme harangue, and sufficiently confident and reliable on the stage to be a genuinely useful member of the ensemble. That is what we seem to be hearing in the staccato syllables that emerge from the dim cylinder. The few sustained notes don’t exactly make you sorry there was no time for the slow part of the Catalogue Song. Presumably the “beautiful, rich” voice that the Allgemeine Musikalische Zeitung had praised in 1847, and that that Lind herself tried to recruit for Stockholm earlier in that decade, was a thing of the past by 1889. Fair enough.

But the rhythm! Was he feeble-minded with age? Was he a musical illiterate who depended on the conductor’s gestures to know when it was time to sing? Was he so arrhythmic that he could not keep the beat without hearing an orchestra? Or was he maybe, you was a post-performance birthday party...perhaps Rubens served something other than water and coffee…?

We can never know the answer to that last one, but we can rule out the others. Illiterate, no: he was a composer who published seven volumes of songs (here’s one). Arrhythmic? Then he could hardly have been “much in demand for quartet evenings” as a cellist, as one encyclopedia entry tells us. Feeble-minded with age? Here we come to the most fascinating thing. Schram’s 1889 farewell to opera was not his farewell to the Kongelige Theater; he had long been employed in a double capacity, alternating operatic and straight dramatic performances, and in the latter he continued right up to his death in 1895.

Every book and encyclopedia article on Danish theater counts Schram as a significant figure in the new dramas of Ibsen, Heiberg, and Drachmann, the classics of Holberg, the translated masterpieces of Schiller, Shakespeare and Hugo. Edvard Brandes, in Danske Skuespilkunst (“Danish Acting,” Copenhagen, 1880) devotes a whole chapter to him - one of 15 actors so honored, and the only opera singer among them. The farther we look into his life, the more important he seems to have been.

Life and Career

Giuseppe Siboni

Schram was born into a middle-class family with a tradition of musical accomplishment. One uncle was a well-known violinist, and the youngster was trained on piano and all the instruments of the string family from childhood. When he was twelve his parents put him up for admission to the Royal Academy of Music, recently founded by King Christian VIII under the directorship of the Italian tenor Giuseppe Siboni. Siboni’s entrance-examination notes survive in the archives: “Not a good voice, but a perfect ear, and all other dispositions perfect; very good.” Schram was accepted, and within a year he had a walk-on part on the stage of the Royal Theater. Two years later he had his first speaking role, and when his voice changed it turned out to be good after all. He studied with Siboni up to the latter’s death in 1839, and in 1841, aged 22, he made his opera debut as Bertram in Robert le diable.

Siboni is a fascinating figure himself - a star tenor born in 1780, the same year as the castrato G.B. Velluti, with whom he eventually sang at La Scala. Siboni’s operatic debut was in 1797 in Florence and his career quickly became international; he was at Prague by 1800 and the King’s Theatre (London) by 1806. He was successful in Vienna, where he knew and performed with Beethoven, and it was apparently he who introduced Schubert to the latter’s great interpreter and champion Michael Vogl. In 1819, after several previous attempts, King Christian persuaded Siboni to accept a court appointment to sing in and supervise Italian operas at the Royal Theater, to train native singers in Italian style, and (from 1825) to direct the Academy, still in existence today.  The idea that we might be able to hear the pupil of a teacher whose life overlapped Mozart’s is a bit mind-boggling.

Schram as Papageno

Meanwhile Schram soon came in contact with another legendary pedagogue. Jenny Lind had been in the Robert cast as a guest star, and immediately recommended Schram to her home theater in Stockholm and to her teacher, Manuel García jr., who after a successful course of study in the off-season proposed the young Dane to the Opéra in Paris. (Schram recalled in an 1893 autobiographical sketch that he had at first irritated García by sight-reading an aria so well that the teacher thought he was pretending not to have seen it before. At home that night, García composed a fresh aria to put the young show-off in his place, and when Schram read the new one just as well the next day, he became a firm favorite.)

In both these cases the Royal Theater forestalled the foreign engagements by raising Schram’s salary, which may have been his goal all along; in the end, with only brief guest engagements elsewhere, he stayed at the Royal Theater for the rest of his life, a career of sixty-three years on the same stage. For decades he sang a wide and varied repertory - Cardinal Brogni, Bertram, Marcel, Kaspar in Der Freischütz, but also Papageno, Sergeant Sulpice, Leporello, and the Bartolos of Mozart and Rossini. In 1864 he introduced Gounod’s Mephistopheles to Copenhagen, and in 1872 Wagner’s Hans Sachs.

But having started as an actor, he retained a second contract for purely dramatic parts, and from about 1870 he focused on these more and more. On the opera side he retained mostly his buffo repertory, while on drama nights he played a wide variety of character roles, serious and comic alike, meanwhile sharing the stage with Ristori and Bernhardt in guest engagements as he had shared it with Lind and Schröder-Devrient in his singing prime. From this phase of his career there are dozens of photographs of Schram in costume, and he photographed better than most: for one thing, he had a clear gift for delineation of character through simple body stance.


By the time Danske Skuespilkunst appeared, Schram had played over 170 roles. Brandes gives eyewitness comments on his acting in well over a dozen of them - by no means uncritical, but in sum deeply appreciative. He singles out an unmatched ability “to play uncertainty in all its nuances.” At random: Cola in Henrik Hertz’s The Youngest, a lazy gadabout whose role was “not much developed by the author’s hand,” was by Schram

...raised almost to a modern Caliban [...] His defiance was exactly that of a stubborn donkey afraid of the prod. One sensed how malignantly he wanted to give all the gentlefolk a kick, and that only his cowardice impeded a rebellion like Caliban’s. Schram’s mark on the role was this way of combining anxiety with sloth.

Brandes had an eye for things that are hard for actors to do, like playing ambivalence, or unusual for the roles in question, like the electrifying effect of genuine terror in Schram’s Mephistopheles when confronted with the raised crosses. He selects for detailed commentary two of Schram’s interpretations that he considered definitive:  Leporello and Aslaksen, the printer whose malleable reaction to prevailing winds of opinion is central to the crisis in Ibsen’s En Folkefiende, “An Enemy of the People.”

The description of Leporello starts with mixed intentions:  

Take him when he unrolls the list [...] of his master’s conquests. He admires Don Juan: "A hell of a fellow! If I were a knight and good looking and rich, by God, I’d lead the same life!"  In fantasy, he has possessed all these women; as confidant, he has in any case had some part in them. And his eyes shine with lust. There is an echo of Don Juan's jubilation in the andante at the end of the Catalogue Song—and yet he takes pity on the innocent lamb that is sacrificed upon the altar of his lord.  What nuances does Schram not express here with tone and glance!

A commercial illustration of the younger Schram playing the role in the 1850s captures this rather strikingly: has anyone else depicted Leporello looking as though he wants to comfort Elvira?

There is another description of the actor’s capacity for bringing cowardice and abasement to life: his Leporello “shrinks like a worm” when the Commendatore’s statue speaks in the graveyard, and gives “a wonderful wail” when confessing after the scene in which he dresses in his master’s clothes: “the self-abasement he puts in his plea for forgiveness is indescribable.” And there is appreciation of a subtle character choice in a difficult episode just before that:

...Leporello has tricked Elvira, on Don Juan's instructions, and she lovingly embraces the guilty rascal. It's a pretty shabby scene. Schram saves it by the grotesque way in which he utters his line "the joke is not so bad." The audience laughs, and is not offended.

Schram as Leporello

I wish I knew exactly what Brandes meant by that. Possibly that Schram, by allowing his own character to appear in an unsympathetic light, spared the public from feeling complicit in the cruel joke by identifying with Leporello’s p.o.v.? He makes a contrast with Pierre Gailhard at the Paris Opera, whose reading of the line he calls “saftig” (juicy, which can also mean sensuous or racy). He also says that the Leporellos he has seen in all the other capitals of Europe are not fit to tie Schram’s shoelaces.

Aslaksen in En Folkefiende is a man on the spot, since he owns the printing press that either will or will not publish Dr. Stockmann’s exposé of pollution in the town’s spa waters. In the first two acts, he sees both virtue and glory in publishing. But as corrupt officials make clear their reasons for wanting the news suppressed (or branded fake), Aslaksen vacillates, and in the end he finds himself unhappily chairing the meeting that declares Stockmann “an enemy of the people.” Brandes on Schram’s reading:

With a brilliant energy, he has here gathered all the features of spinelessness into the sad picture [...] This is how the gifted man comes to look who does not believe in himself, and succumbs in the soul-destroying struggle for existence. Schram elevates the wretched, drunken printer into an almost tragic figure. Never has a human being looked more pitiful than this fellow with the furrowed brow whose every wrinkle tells of a year’s misfortunes like the rings of a tree trunk. It is as though the hand of fate had pressed his head down between his shoulders, as if half-working, half-carousing nights had drawn all the blood from this blue-pale face. What a superb makeup job! And the black suit that was once so fine and sociable is now shabby and greasy, precisely the mark of a true pauvre honteux. The dragging voice, whose plaintive tone comes close to weeping, makes every line a masterpiece.

Schram as Aslaksen

I feel I can practically hear that voice coming through the blurry photograph (and doesn’t it look like a man in the very moment of halting a step he has started to take?).  Brandes concludes his account with a weirdly prophetic question: “Why, indeed, may we not bring it to the phonograph, for the instruction and enjoyment of future generations?” This is startling, given that at the time no commercial model had yet emerged from Edison’s workshop. But it brings us back to the cylinder from which we are still hoping to derive some instruction and maybe even enjoyment.

The singing

Before returning to the bizarre aspect of the interpretation, we should note that we cannot be exactly sure of the cylinder’s correct playback speed, and thus cannot be sure just what the tone of the voice was in 1889. What we can say is that at least one of the two excerpts is sung in something other than its original key: Schram finishes “Notte e giorno” on a low note and then launches into “Madamina” on the same note an octave higher. Natural enough if you’re singing a capella at a party, but in the score they don’t stand in that relationship; the first piece is supposed to end on an F and the second should start on F-sharp. So if the entrance aria is in key, the Catalogue Song is a half-tone down, or if the latter is in key, the former is a half-tone up. We have to guess; there is no reference point and no “standard speed” for these home-produced cylinders.

After listening both ways many times, my guess is “Notte e giorno” in F and “Madamina”  in D-flat - but here is a fragment “guessed” the other way, with “Notte e giorno” in F-sharp, for the interest of comparison.  (In this snippet I’ve also applied even more aggressive noise-reduction software and cut out the worst bangs caused by cracks in the physical cylinder. You’re not really supposed to do that - there is always a risk that something substantive will vanish along with the unwanted sounds. But this is an extreme case.)


Back to the rhythm, though. We’re clear on the fact that people sang with more freedom back then. But this much freedom, in this kind of music? With a bit of searching, it’s possible to hear a 1904 recording of the same solo by one of Schram’s early successors as Leporello at the Kongelige Theater: Helge Nissen (1871-1926), a prolific bass-baritone who sang everything from Osmin to Wotan to Count Di Luna, and sometimes conducted as well.  Here is how he did it:


OK, not as free as Schram, but far from literal. Both recordings basically underline the lesson of Georg Henschel’s (see “Die beiden Grenadiere,” Record of the Week for October 2, 2018) - namely, that togetherness and steadiness did not rank so high in the 19th century’s list of musical priorities as they do in ours. This is the only early Danish version I could find, but several German ones are comparably loose. (By the way, the squawk at the beginning is a pitchpipe playing a C, so that listeners could “tune” their gramophones - adjustable in those days with a simple screw mechanism - to the speed at which the record had been made. If more companies had adopted this policy a lot of problems could have been avoided!)

So where does this leave us with the cylinder, which is rhythmically looser still? Even to take an educated guess, we have to form some kind of opinion on the narrower question: is the way Schram sang without accompaniment representative of the way he sang with it? There’s no way to answer that for sure, but it obviously makes a difference whether we think “no way” or “maybe so.” Couldn’t we at least test the hypothesis by sitting down to “accompany” him in 2019?

Here’s what I got when I tried that:


Schram as Per Degn

Yes, it’s a bit of a scramble, and I’m sure with a few more tries it might be done more convincingly. But the only places I had to tweak the original recording are the two spots where Leporello is silent for almost two full bars. Schram didn’t wait that long, so I estimated their length based on the tempo he was singing before, and added a corresponding amount of “silence” (i.e. cylinder noise) to the original sound-file before trying to accompany it. When he is actually singing, though, there’s no real problem - barely even a difficulty. I had to hurry or wait more often than we are used to doing in Mozart, but it was nothing an orchestra couldn’t do just as easily (Nissen too speeds and slows in ways that would keep the band alert). Apart from those moments, though, the players could basically maintain a steady beat and count on the singer to tally up with them where he needed to, however freely he might declaim in the meantime. The words in the first phrase follow their own rhythm, but the next downbeat arrives on time. The seemingly too-short A at “Fanden være Tjener meer” (“e non voglio più servir”) turns out to be a coherent syncopation.

Conclusion: yes, it’s at least possible that what Schram sang at the party was pretty much what he had sung on stage earlier in the evening. And that gives us a lot to think about.

Schram as Don Ranudo

Meanwhile: when we imagine what else might have been recorded in 1889 if the musical world had been alert to the miraculous new possibility, scraps like this may seem small compensation. We have a pretty clear idea of music-making from 1900 forward and nothing at all before 1888, so the dozen years in between are a tantalizing zone of unfulfilled possibility, punctuated by a few relics from amateurs and enthusiasts - those whose cylinders managed not to be discarded by their heirs, left in a demolished building, cracked by temperature changes, overrun with mold, bombed in two world wars, or simply mislaid. Julius Block in Moscow and St. Petersburg; Ruperto Regordosa in Barcelona; our Gottfried Rubens in Copenhagen - all of them recording artists with whom they happened to have social contact - had their collections preserved by lucky chance. Who knows what treasures were created and lost by similar hobbyists in New York, Berlin, Vienna, Milan, Paris, London?

But “chance” did not do so badly in settling on Peter Schram - a singer trained by an eighteenth-century Italian, active on stage for fully two-thirds of the nineteenth, and a serious participant in the theatrical currents that would drive opera’s transformations in the twentieth. And Schram did not do so badly either: with a 170-second shot at immortality, he managed to leave a time-capsule that can still teach us a thing or two in the twenty-first.


Schram portrait photo

Believe it or not, Schram’s turns out not to be the very earliest recorded opera voice after all. The same Rubens collection holds four cylinders made by Ludvig Phister (1807-1896) in his mid-eighties. Phister was mostly a straight actor - only one of the cylinders contains any singing, and that is a very thin-voiced bit of folksong ending a spoken scene from an unidentified play. He sometimes took small comic roles in operettas and singspiels, most of them probably involving little or no music. But scouring the archives, one finds that in his first few seasons with the company he also participated in “real” operas - notably, starting on March 29 1830, the bass role of Orbazzano in the production of Rossini’s Tancredi supervised by Siboni. So, strictly speaking, he counts as an opera singer twelve years older than Peter Schram. Who knows what else will emerge? Rossini’s own protégée Marietta Alboni was still singing Tancredi’s aria at private gatherings when Edison’s agents first spread out over Europe…

Teatro Nuovo puts great emphasis on learning from the singers who had never heard, or heard of, microphone singing - primitive recordings from more than a century ago, forming a link to the traditions of opera’s heyday and the infinite potential of the natural, unassisted human voice. Check this space regularly for samples, and click here for some pointers on how to listen.