The Work Without a Number

Track nine ended with a piece of music that never decided what it was called.

Track ten has the opposite problem. Nobody has ever been confused about what this music is called. It is called Brahms. It has been called Brahms since 1880, it is called Brahms on the sleeve of this album, and it will be called Brahms for as long as anyone plays it.

Brahms spent the rest of his life saying it shouldn't be.

First, the physical facts, because they are unusual. Track ten runs four minutes and one second. That makes it the longest thing on the record so far — the first track on this album to pass four minutes — and it arrives directly out of the shortest, the Massenet, which was gone in one minute fifty-five. The album drops to its smallest and then opens all the way out. And the piece itself has two halves: it begins Andantino, slow and low and in no hurry at all, and then, about halfway through, it turns Vivace and never looks back. The title on the album names only the first half. And one more piece of geography: on the double vinyl this is the first track of side C — the opening side of the second record. New disc, needle down, and this is what you hear.

Now look at the credit, because on this album that is always where the story is. Deutsche Grammophon prints it 21 Hungarian Dances, WoO 1: No. 17 in F-Sharp Minor. Andantino. Garrett's own site prints it Hungarian Dance No. 17 in F-Sharp Minor. Andantino, from 21 Hungarian Dances, WoO 1. Different lengths, same three small characters sitting in the middle of both — and almost nobody reads them.

WoO. Werk ohne Opuszahl. Work without opus number.

Brahms numbered his works. Op. 1 through Op. 122, the symphonies and the concertos and the sonatas, a catalogue kept in order by a man who cared a great deal about order. WoO 1 is the first entry in the other list — the list of everything he published outside that catalogue. And the Hungarian Dances are at the top of it not because a later scholar filed them there. They are there because Brahms insisted.

Henle's edition puts it about as plainly as it can be put. "'No opus number' is what Brahms gave his astonished publisher to understand, when he offered him the Hungarian Dances for publication in 1869. He placed a great deal of importance on the fact that he had merely 'set' popular Hungarian melodies and not composed new works as such."

He was right to insist on it, and he was not careful enough about it.

Of the twenty-one dances, only three — Nos. 11, 14 and 16 — are reckoned to be his own tunes; the judgement comes down to us from Joachim, who was in a position to know. The rest he had picked up. Some from the Hungarian violinist Ede Reményi, who took the twenty-year-old Brahms on tour in 1853 and taught him this music by playing it at him night after night. Some from café bands in Hamburg and Vienna. Some from printed sheets.

One of them turned out to belong to somebody who was still alive. Hungarian Dance No. 5 — the one everybody can hum — is built on a csárdás called Bártfai emlék by the bandmaster Béla Kéler, and Kéler said so, publicly. Brahms's answer was that he had taken it for a folk song, and that no greater compliment could be paid to a composer than to find his melodies being sung by peasants in remote villages. Which is a graceful reply and also, from the man holding the royalties, an easy one.

He was accused. He was never, so far as any record shows, sued. But he clearly knew what he was doing was delicate, because of what he wrote to his publisher about the dances. It is the most revealing sentence he ever produced about them, and it is one line long.

"They are, incidentally, genuine puszta and gypsy children. Not begotten by me, only raised on milk and bread…"

That is his word, not ours, and we have left it in his mouth rather than tidy it away, because the tidying is exactly what the last hundred and fifty years have already done to this music. So let us say the other half plainly, in our own voice: the people he was describing were Roma musicians. They were working players in the coffee houses and dance halls of Vienna and Budapest and the Hungarian countryside, they were the authors of a great deal of what is on this record, and not one of them has a name in this story.

That is not a modern accusation dressed up as history. It is measurable. The musicologist Jon Banks went through the 78rpm discs of "traditional" Hungarian music made in the years just after Brahms died, and found that nearly half of the Brahms melodies turn up in the working repertory of the recording bands — many of whose players had been active in Vienna in Brahms's own lifetime. The Hungarian Dances, he concluded, are an invented amalgamation of several distinct genres: café music, military band music, popular song, salon pieces. Brahms's idea of Hungarian music simply was this music. He heard it from the people who played it, and the people who played it were not asked.

So here is where the first half of the story leaves us. He put his name on the cover, and then spent the rest of his life refusing the one word — composed — that would have made the name mean what everyone assumed it meant. He refused the opus number. He wrote the disclaimer down in a letter. He did roughly everything a scrupulous man of 1869 could think to do.

And of everything he wrote — the four symphonies, the German Requiem, the concerto that every violinist on earth is still climbing — this is the music the world actually took home.

• • •

Follow this one dance, No. 17, and count the hands that have been on it.

First, a tune, in circulation, unattributed, played by people whose names nobody wrote down.

Then Brahms, who sets it for piano four hands and publishes it with Simrock in 1880, in the last of the four books. No. 17 opens Book IV — the first dance of the final set, which, a hundred and forty-six years later, is exactly where it has ended up again.

Then Joseph Joachim, who arranges all twenty-one for violin and piano. This is the step that matters most to us, because it is the reason a violinist ever meets this music at all: the original is a piano duet, and without Joachim there is no violin in the story. His versions are not polite reductions, either — they are harder than the original, written by a virtuoso for virtuosos.

And Joachim is in this story twice, at both ends. It was Reményi — the man who taught Brahms these tunes on the road, and who would later accuse him of stealing them — who introduced Brahms to Joachim in the first place, in that same year, 1853. One Hungarian violinist put the music into Brahms's hands. Another put it onto the violin. They were the same introduction.

Then Antonín Dvořák, who orchestrates Nos. 17 to 21. And the reason Dvořák is anywhere near this is Brahms: it was Brahms, sitting on the prize committee that kept awarding Dvořák a stipend, who sent him to his own publisher. Simrock then asked the new man for something dance-shaped of his own, and got the Slavonic Dances of 1878 — modelled on the Hungarian Dances, as everyone including Dvořák said at the time. But note the difference, because it is the whole argument in miniature: Dvořák wrote his own melodies, and gave the set an opus number. Brahms used other people's, and refused one. And it is Dvořák, the man who signed his, who ends up orchestrating the ones Brahms wouldn't.

Then Fritz Kreisler, who makes his own version of No. 17 for violin and piano — and transposes it down a semitone, into F minor, which is how a good many violinists have met this dance ever since.

Stop there for a second, because you have met that name on this album already. Kreisler is track six: La chasse, published in 1911 as the rediscovered caprice of a dead eighteenth-century Frenchman who had not written a note of it, and confessed as a forgery only in 1935. That is the man who arranged this Brahms.

And he is the mirror image of it. Kreisler took his own music and put somebody else's name on it. Brahms took somebody else's music and put his own name on it — and then spent thirty years insisting the name meant nothing more than the man who wrote it down. Two composers, two sets of quotation marks, pointing in opposite directions. If you want the full version of Kreisler's side of it, it is here: The Violinist Who Invented His Own Masters.

Then, in 2026, Garrett and van der Heijden, for violin, piano, guitar and orchestra.

And both of those printed credits keep F-sharp minor — Brahms's key, the key of the piano duet, and not the F minor the violin tradition settled into after Kreisler. On a sleeve where every title has been checked twice, that is worth noticing. It reaches back past the violinists to the page as Brahms set it.

• • •

Which brings us to the choice, and to something this series has now watched happen three times.

Say "Brahms" and "Hungarian Dance" to anybody at all and you get one answer: No. 5. It is the one in the films, the one in the adverts, the one every violinist has in the encore folder.

David Garrett already has it. It is track six of Encore, from 2008, and his own site lists it on Unlimited and on the live DVDs besides. He has been playing it for the better part of twenty years.

So he reached past it and took No. 17.

That is now three tracks in a row. On track eight, Paganini arrived without a caprice, because the caprices were already on a record of his own. On track nine, Massenet arrived without the Méditation from Thaïs, because the Méditation was already on Classic Romance. On track ten, Brahms arrives without No. 5, because No. 5 is already on Encore.

Three famous names. Three famous pieces withheld. The same reason every time — not shyness, not contrarianism, but a man with twenty-five slots to fill who has decided he is not going to spend any of them repeating himself. Track ten is where that stops looking like a coincidence and starts looking like the method.

And there is a sting in the tail of it. The Brahms that David Garrett actually carries into concert halls is the opus-numbered one: the Concerto in D, Op. 77, the Double Concerto, the three sonatas, the Piano Quintet. None of it is on this album. The one piece of Brahms he brought is the one piece Brahms declined to sign.

This series keeps asking a single question: what actually carries music past the people who made it?

Track six's answer was that immortality doesn't check the name on the door. Track eight's was that it takes no instructions from the person most concerned. Track nine's was that it doesn't care what you call it.

Track ten sharpens track six into something harder. The name on the door goes up whether you want it there or not.

Brahms did everything within his power to decline the credit — no opus number, the word set instead of composed, a letter disowning paternity in as many words. It made no difference whatsoever. A hundred and fifty years on, the tune is his, the sleeve of a 2026 record says Brahms, and the musicians who actually played it first are a category rather than a list of names. Immortality assigned the credit. It did not take instructions on how.

So when the line begins on track ten — low, unhurried, four minutes, the longest thing on the record — that is a Hungarian dance nobody can put an author to, played on an instrument it was not written for, in an arrangement that is the fifth of its kind, under the name of the one man who kept saying it was never his.

Immortal — David Garrett for Deutsche Grammophon — arrives 9 October 2026: 20 tracks in the standard edition, 25 in the deluxe, on digital, CD and vinyl, with pre-orders open now on Deutsche Grammophon. One track at a time, we keep counting the album in.

Sources & notes: the album credit as Deutsche Grammophon and david-garrett.com print it; the opus-number refusal from G. Henle Verlag’s edition of the Hungarian Dances Nos. 1–10; Brahms’s letter to Fritz Simrock — „Es sind übrigens echte Pußta- und Zigeunerkinder. Also nicht von mir gezeugt, sondern nur mit Milch und Brot aufgezogen …" — translated in the text above, with his own term reproduced as he wrote it and answered in our own voice in the paragraph that follows; Joachim’s judgement that only Nos. 11, 14 and 16 are Brahms’s own, and Dvořák’s orchestration of Nos. 17–21, from the Hyperion programme note; Kreisler’s F minor arrangement from the Boosey & Hawkes catalogue; the sources of the melodies from Jon Banks, ‘Brahms’s Hungarian Dances and the Early Csárdás Recordings’, Music and Letters 102/4 (2021), and RomArchive’s Brahms and the Roma; Garrett’s recordings of Hungarian Dance No. 5 from david-garrett.com. We have not asserted that Brahms was ever sued — the accusations by Béla Kéler and Ede Reményi are documented, a court case is not — nor identified the source melody behind No. 17, which we could not establish. The key given is the one the printed credits carry.