I recently began preparing a new edition of the B-flat Cello Concerto usually attributed to Michael Haydn. I now have an almost publishable score with critical notes, prepared directly from the surviving manuscript parts.

After studying the source closely, I do not believe the concerto is by Michael Haydn. I am even less persuaded by Mark Starr’s suggestion that it may instead be by Joseph Haydn.

The manuscript is anonymous. It does not name a composer. Its presence in the Michael Haydn collection in Budapest may explain how the attribution began, but provenance is not proof of authorship.

The music does not make a convincing case for Michael Haydn either. Its scale, orchestration, solo-tutti organization, and treatment of thematic material do not fit comfortably with his authenticated instrumental music.

Starr’s case for Joseph Haydn is even weaker.

After corresponding with him directly, I learned that there is no larger body of unpublished evidence behind the attribution. His case appears to rest on a resemblance between the 6/8 finale and the finale of Joseph Haydn’s D-major Cello Concerto, his feeling that the concerto is “too good” and “too big” for Michael Haydn, and the manuscript’s presence in the Michael Haydn collection.

None of that establishes Joseph Haydn as the composer.

The location of the manuscript may help explain the old Michael Haydn attribution, but it does not point toward Joseph. Calling the concerto too good for Michael Haydn is a personal judgment, not evidence. And the resemblance between the finales is not especially remarkable. Hunting-style rondos in 6/8 were everywhere in the late eighteenth century.

Composers shared motives, sequences, rhythms, cadential formulas, orchestral gestures, and whole sections of musical vocabulary. Joseph Haydn’s style is not a collection of little fingerprints. It is found in the pacing, proportions, harmony, rhythmic organization, orchestral thinking, and the way expectations are created and redirected across an entire movement.

This concerto does not think or move like Joseph Haydn.

The first movement opens with a long orchestral tutti. Later orchestral passages are interrupted by short solo appearances. The development is built largely from new material, and the second theme from the exposition does not return normally in the recapitulation. The work’s Romance–Rondo pairing and its general handling of soloist and orchestra also point toward a different musical world.

These are not tiny details that can be outweighed by one cheerful 6/8 theme.

When I first played the solo part, it reminded me broadly of the world of Carl Stamitz. It is not especially Boccherinian, and it does not yet feel like the later cello writing of Romberg, Joseph Reicha, or Kraft.

More recently, after comparing the concerto with authenticated works by Antonio Rosetti, I have become increasingly interested in Rosetti as a possible attribution. The scoring, melodic style, phrase construction, horn writing, orchestral gestures, and solo-tutti exchanges seem much more at home in his musical world than in that of either Joseph or Michael Haydn.

At this point, Rosetti is the strongest candidate I have found.

That is still only a hypothesis.

No cello concerto appears in the existing thematic catalog of Rosetti’s music, and I have found no documentary evidence connecting him directly with the Budapest manuscript. A serious attribution would require broader comparison with authenticated works, along with more research into the paper, copyist, provenance, and history of the source.

For now, the honest description is something like: Cello Concerto in B-flat major, anonymous, formerly attributed to Michael Haydn; possibly by Antonio Rosetti.

I have spent years studying and playing hundreds of cello concertos and concert pieces from this broad period, from Costanzi and Gretsch through Piatti, Swert, Popper, and many composers in between. I certainly have not examined every surviving work, but I know the repertoire and its different schools of cello writing unusually well.

A performer who has worked through enough of this music begins to recognize the grammar of instrumental writing. How do the sequences fall under the hand? How are the registers connected? Where do the shifts lie? How is the passagework built? How do the technical patterns relate to the form?

None of that proves authorship. But it can help place a work within a broader instrumental tradition.

Two Haydnesque ideas cannot outweigh the language of the whole concerto.

This raises a larger question: how easily does a musical hypothesis become an accepted fact?

Someone discovers or reconstructs a source. A famous performer or editor becomes associated with it. The claim gets repeated in program notes, recordings, competitions, and new editions. Eventually, almost no one remembers what was actually demonstrated.

Viktor Kubatsky offers a useful example.

In 1956, he published a reconstruction of Tchaikovsky’s original version of the Rococo Variations. The later story became wonderfully simple: Tchaikovsky wrote the authentic version, Fitzenhagen mutilated it, and Kubatsky restored the truth.

The source history is more complicated. Tchaikovsky and Fitzenhagen worked together, and several manuscript states survive. Kubatsky did real source work, but the mythology around his edition eventually turned one reconstruction into the single pure original and made Fitzenhagen the villain.

The string world loves this kind of story:

Composer good. Cellist bad. Modern editor restores truth.

Real collaboration is usually messier.

Kubatsky’s edition of Shostakovich’s Cello Sonata presents a related problem. Kubatsky was the dedicatee and first performer, which gives his edition real historical importance. But he was not the cellist who created the work’s lasting performance tradition. Shafran and, especially, Rostropovich performed and recorded the sonata with Shostakovich.

In 1968, about thirty-four years after the premiere, Kubatsky prepared a new edition. It was approved by Shostakovich and published after Kubatsky’s death.

But what does “approved by Shostakovich” actually mean?

Did Shostakovich originate every new bowing, articulation, fingering, dynamic, and textual change? Did he examine every detail? Was Kubatsky preserving decisions from 1934, or presenting his own later interpretation?

We do not know.

The problem begins when Kubatsky’s authority as dedicatee and first performer quietly becomes the authority of the composer. A performance edition may be valuable without being treated like an autograph. “Approved by the composer” is not the same as “written by the composer.”

This is how authority gets laundered.

The theory that Anna Magdalena Bach composed the Cello Suites is a more extreme example. Martin Jarvis based the claim partly on handwriting analysis and supposed corrections in her manuscript copy. The theory received enormous attention because it offered everything a modern audience could want: a famous masterpiece, a neglected woman, a historical injustice, and apparently forensic evidence capable of overturning the canon.

But the manuscript identifies Johann Sebastian Bach as the composer and Anna Magdalena as the copyist.

The theory required pushing aside the clearest evidence on the page.

Anna Magdalena was an important musician, singer, copyist, and participant in Bach’s working life. Recognizing that does not require inventing works for her. That is not historical justice. It is simply bad scholarship.

Then there are cases involving actual deception: Kreisler’s supposed rediscoveries, the Casadesus concertos, the “Albinoni” Adagio, and Winfried Michel’s false Haydn sonatas.

Michel’s case is especially relevant. He took genuine four-measure openings by Haydn and composed complete sonatas around them. Several respected musicians and scholars initially accepted the works.

The openings were Haydn.

The sonatas were not.

A few genuine motives did not make them genuine Haydn.

That should serve as a warning.

These cases are not identical. Kubatsky was not a forger, and proposing a new attribution is not dishonest. But speculation must remain speculation.

In the case of the B-flat concerto, there is no larger unpublished argument waiting to be evaluated. The Joseph Haydn attribution rests on a thematic resemblance, a judgment about the quality and scale of the work, and the manuscript’s connection with the Michael Haydn collection.

That is not enough.

The source is anonymous. The provenance does not point to Joseph Haydn. “Too good for Michael” is not a scholarly argument. And one familiar 6/8 theme cannot outweigh the language of an entire concerto.

Performers also belong in this discussion. Playing a piece does not prove authorship, but instrumental writing has its own language. Cellists can feel whether the passagework grows naturally from the instrument, what technical tradition it belongs to, and whether it consistently behaves like a particular composer’s work.

Source study can then test or correct those impressions.

After playing the concerto, reconstructing the score, and studying the surviving parts closely, I am not claiming to know who wrote it.

I am saying that the Michael Haydn attribution rests mainly on provenance, while the Joseph Haydn attribution rests on a thematic resemblance and a subjective judgment about the quality of the music. Neither case is persuasive.

Rosetti currently offers a more promising direction, but Rosetti also remains a hypothesis. Until there is better evidence, the anonymous source should be allowed to remain anonymous.

A few motives do not make a Haydn. A famous dedicatee does not turn a performance edition into the composer’s autograph. A reconstruction does not automatically uncover one pure original. And a weak attribution does not become stronger because it is repeated.