Bucephalus was Alexander’s horse. Gassendi brings the famous animal into his objections to make a small but pointed comparison.

What the objection says

Descartes says the mind is not extended. Gassendi objects that this says what it is not, rather than what it is. Does knowing that something is not another thing give us a clear and distinct idea of it?

Suppose someone knows only that Bucephalus is not a fly. Is that a clear and distinct idea of Bucephalus?

The distance between the creatures makes the example work. You may exclude a fly without coming anywhere near an account of the horse.

What the reply says

In Latin, a fly is musca; music is musica. One letter separates them. In Descartes’ reply, the fly has become music.

He makes two moves. First, he says that calling the mind unextended was not intended as a complete account of its nature. It corrected those who took it to be extended. Similarly, if someone asserted that Bucephalus was music, denying that assertion would not be pointless.

Restore the fly in that first comparison and the reply still works. Denying that the horse is a fly can correct a mistaken claim without giving a complete account of the horse.

The second move is different. Gassendi argues in the same passage that a mind using an extended body must itself be extended. Descartes replies that this is no better than saying Bucephalus neighs—or bellows—and thus makes sounds that might be classed as music, therefore Bucephalus is music.

That analogy now depends on music. It is no longer simply answering the opponent’s example; it is an example of Descartes’ own.

The two appearances of music therefore have different jobs. The first occupies the place of Gassendi’s fly. The second belongs to a new comparison.

Someone noticed

Three years later, Gassendi’s renewed objections appeared in his large volume. He pointed out the discrepancy. The wording discussed here reaches us through a hostile commentator quoted by later editors: Gassendi reportedly says, with measured politeness, that Descartes has compared Bucephalus with music, whether intentionally or through a misreading. Gassendi himself, following both common sense and the printed text, had compared the horse with a fly.

Both possibilities remain in that wording: deliberate alteration, or reading error.

Descartes subsequently wrote to the translator of the Objections and Replies that he had neglected to answer Gassendi’s large book. The translator had arranged for friends to extract its strongest arguments and send him a digest; Descartes replied to that digest.

The editors report that the digest has not been found. The material used here does not establish whether Descartes saw this particular correction.

Into French

Another three years pass, and the French edition appears. In a pamphlet of 1650, Jacobus Revius discusses what happened to the passage. He quotes the French wording and records its pages. He is also the intermediary through whom the comment attributed to Gassendi above is read here.

Revius was an opponent of Descartes. The account is evidence from an adversarial report, not an unmediated window into the translator’s decisions.

According to his report, the French keeps the fly on Gassendi’s side and music on Descartes’ side. In modern spelling, mouche and musique are far more than one letter apart. Translation did not reconcile the examples.

But Revius records another alteration.

The sound that disappeared

In the Latin reply, the horse neighs or bellows. The French, in the passage Revius reports, keeps only the neighing.

His explanation is that the translator noticed that bellowing did not belong to a horse and removed it. The rest, he says, was left as it stood.

The deleted sound belonged to Descartes’ own second comparison. The music occupying the fly’s place in the first comparison remained. That is the distinction worth keeping in view.

What is not on the page

Did the translator notice the fly/music discrepancy? The surviving wording discussed here does not tell us.

If one wanted to restore the opponent’s example, the immediate place to intervene is clear: put the fly back where the first music stands. Descartes’ later analogy could remain, with an explanation if needed. Repairing that mismatch does not require deleting an entire passage.

According to the report, that repair was not made. Why it was not made is another question. The essay does not supply a motive that the evidence cannot disclose.

Editing somebody else

An author miscopies a word while reporting an opponent. Then the author builds an ingenious new analogy around that word.

An editor has two different things to handle. One sentence purports to report another person’s argument; another belongs to the author’s own reasoning. Restoring the first does not, by itself, authorize replacing the second.

Of course the editor might never notice the mismatch. The returned manuscript would look exactly the same. A visible result does not uniquely identify the attention—or intention—that produced it.

At the end

Revius finishes his note with a couplet. In the old story, a goddess substitutes a deer for a girl at the altar. Here music has been substituted for a fly.

The mismatch enters the book in 1641; Gassendi points it out in 1644; it passes into French in 1647; an opponent mocks it in 1650; the Adam–Tannery editors reproduce that mocking note in their Latin volume, published in 1904.

One pair of hands after another. The fly and the music remain on their respective sides. According to the report, only the bellowing disappeared on the way into French.