In reading from 100 Orchestration Tips by Thomas Goss, I've come across a tip which I'm perfectly happy accepting, but (perhaps as I'm not a horn player nor a trumpet player) I do not understand.

34. Horn Key Signatures

"... concert horn and trumpet players prefer not to have key signatures."

"... if you ask them what they really want, you'll get anything from a polite suggestion to a lengthy dissertation on why key signatures are irrelevant to horn. ..."

Tip 34: Horn Key Signatures (excerpt)

My question (primarily to horn players or trumpet players) is:

  • Why is it that key-signatures are 'irrelevant' to horn and trumpet?

I don't need a lengthy dissertation but if you could at least explain the crux or main reasons that key-signatures are not preferred, I would find that helpful.

  • Surely that means the dots are clogged up with even more accidentals than normal?
    – Tim
    Commented Aug 28, 2021 at 9:09
  • 4
    I'd actually suspect the statement is false for trumpet players - all the concert band music I've seen has key signatures for trumpet, and so does the early 1900s-era Boosey & Hawkes-published sheet music for Elgar's Pomp and Circumstance Marches - but that very same Boosey & Hawkes sheet music has no key signatures for French horn throughout.
    – Dekkadeci
    Commented Aug 28, 2021 at 11:45
  • @Dekkadeci you're probably right. Natural trumpets didn't have the same tradition of having 10 or more transpositions available as did natural horns.
    – phoog
    Commented Aug 28, 2021 at 12:05

4 Answers 4


I play trumpet and horn, and have played both in orchestras for many years.

Both instruments used to not be fully chromatic (mainly playing their overtones) and so they'd be crooked into the key of the piece (usually). Thus, the part is transposed to read in C and the parts always end up with no key signature (or rather, the key signature of C major).

Now, I don't have any proof of this, but I strongly suspect what happened is that somebody studied a bunch of music from the 19th century and earlier, saw that parts never have a key signature, and concluded that "trumpets and horns don't use key signatures", which misses the point entirely. But they published that in some orchestration book, and it gets passed down and not questioned.

What's true is that, for better or worse, there is a tradition in classical orchestral music to eschew key signatures in trumpet and horn parts. What is NOT true is that the players themselves "don't read key signatures" or however the author in question wants to put it. Literally all other music for horn and trumpet uses key signatures. Brass quintets, wind ensembles, brass bands, jazz bands, musical theatre pits, etc. all use key signatures. And many modern classical composers are bucking the tradition and including key signatures in orchestral pieces too. So clearly, a player who is even remotely rounded is thoroughly used to reading key signatures and should have no problem with them whatsoever.

And obviously I don't know every trumpeter and hornist in the world, but I've never personally encountered anyone who is either so siloed into orchestral playing that they never encounter key signatures, or is such a zealot for tradition that they get angry about a modern composer using them. So I don't know who these players are that Thomas Goss interacts with, but they sound like clowns to me. Or maybe they're just doodling because they're orchestral brass players and they don't play much and they're bored?

In short, I fully support ditching any tradition based on old crooked natural trumpets and horns. These instruments are now fully chromatic and should be treated the same as the rest of the orchestra. And I may just be one guy, but Thomas Goss is also just one guy, and I at least play the instruments. As a direct retort to:

"... if you ask them what they really want, you'll get anything from a polite suggestion to a lengthy dissertation on why key signatures are irrelevant to horn. ..."

I really want consistency, and saying that key signatures are somehow "irrelevant" to horn makes no sense whatsoever.

  • So, from your experience Tip 34. is basically wrong; I should just ignore it. Orchestral horn and trumpet players do indeed prefer key signatures. Right? Commented Aug 30, 2021 at 14:31
  • @ElementsinSpace Again, I can't speak to what every person prefers, and apparently Goss has found a pocket of players that prefer no key signatures. But there really should be no problem using key signatures, and you have my blessing (FWIW) to roll your eyes at anyone who complains, should you ever even encounter one.
    – MattPutnam
    Commented Aug 30, 2021 at 17:05

Like Dekkadeci, I'll note that the main body of the passage in question does not mention trumpets.

Why is it that key-signatures are 'irrelevant' to horn?

The tradition arose because natural horns always played as transposing instruments in the key of the piece; beginning in the late 18th or early 19th century, the transposition could change in the middle of a movement: if the piece moved to a new tonal area the horn would change to match, always being notated in C (for minor keys, the horn would be assigned a transposition to a closely related major key). The practice of not using key signatures follows from this fairly, ahem, naturally. Trumpets didn't change transposition as readily, and their parts are more likely to have key signatures, but if you search the internet for images you'll also find several of trumpets with sets of crooks.

If you look at older scores, you'll see notations like "muta in Sol" which means "change to G." Originally, this meant inserting an extra length of tubing, or removing one, to change the length of the horn. After valves became standard, the F transposition became standard, too, but the practice of not using key signatures remained.

  • So, early horns players would have a set of different crooks for the different keys? But wouldn't there have to be something (like in older scores) to tell them which key everyone else is in, so they could change to the apppropriate crook?
    – Tim
    Commented Aug 28, 2021 at 13:29
  • 2
    This is definitely the answer for horn. One of the reasons horn is notoriously difficult among brass is horn players have to read all kinds of literature including natural horn and often transpose on the fly, unless they have time to make or get a transcription, which generally they learn not to need instead. Commented Aug 28, 2021 at 16:17
  • I have seen “natural” trumpet parts written, which I found confusing. And I can’t remember where. But I don’t think that carried through to modern notation in the same way that horn notation has. Commented Aug 28, 2021 at 16:21
  • 1
    @Tim to change to the appropriate crook, they don't need to know what key anyone else is in; they just need to know what key they are supposed to be in. If everyone else is in D minor, one pair of horns might be in D and the other in F or B flat. They'd alternate notes depending on whose overtone series the note is in. The part might say Corno 1° in Re at the beginning and then later Muta in Fa or whatever, which means "take out the D crook and put in the F crook." Later in the 19th century you could get some very eclectic combinations to support the increasing chromaticism.
    – phoog
    Commented Aug 28, 2021 at 16:23
  • 2
    @ToddWilcox it was certainly the practice into the second half of the 18th century, probably a good deal into the 19th. But keyed trumpets already existed in Haydn's day, so trumpeters were liberated, if you will, from the overtone series rather earlier. Also, people seem to have been more conservative about horns: even a few decades after valved horns were well established, some composers continued to prefer natural horns. I don't remember whom I've read this about, but it was definitely in the latter part of the 19th century. Brahms, perhaps.
    – phoog
    Commented Aug 28, 2021 at 16:28

It's quite possible that those concert horn players are following an earlier music tradition where French horn parts were consistently not notated with key signatures, regardless of the home key of the piece.

Implying such a music tradition exists, the Boosey & Hawkes-published scores of Edward Elgar's Pomp and Circumstance Marches No. 1-4, which were all published in the early 1900s, all have no key signature for the horn parts throughout, regardless of the marches' home keys and key changes. (You can check these scores on IMSLP. Interestingly, the horn parts of Pomp and Circumstance March No. 5, which was first published in 1930 and should enter the public domain in the USA in 2026, do have key signatures.)

The closest I can find to concert trumpet players also not preferring key signatures are those Boosey & Hawkes scores similarly not having key signatures for parts for F trumpets (I'll note that I've never been in a concert band with a F trumpet, despite being in 3 school concert bands) and a claim I have heard that film music is often similarly not notated with key signatures, regardless of home key or number of key changes. I'll note that those Boosey & Hawkes scores all had key signatures for the trumpet parts for trumpets other than F trumpets, and all the concert band and marching band sheet music I've seen have key signatures for trumpet parts.

  • 2
    The "no key signatures for trumpets" convention only holds for orchestral music. Wind band & brass band trumpet parts generally have key signatures (as you note). See Gould's Behind Bars, pp. 263, 539, and 546 for orchestra, wind band, & brass band respectively. Commented Aug 28, 2021 at 13:56
  • As an example, see the trumpet part for Tchaikovsky's Fourth Symphony. The first movement starts in concert F minor and modulates to B major & F major before returning to F minor, but there's not a key signature to be seen in the trumpet part — just a lot of accidentals. Commented Aug 28, 2021 at 14:04
  • 2
    The trumpet part from the Boosey & Hawkes edition of "Pomp & Circumstance" also has no key signature, using accidentals instead. Commented Aug 28, 2021 at 14:17
  • @Michael Seifert - Unusual - the trumpets behave just like horns for that score. My claim that the Pomp and Circumstance trumpet parts have key signatures is mainly based on Marches No. 3 and 4.
    – Dekkadeci
    Commented Aug 28, 2021 at 19:06

Yes, historically valveless horns and trumpets were built (or adapted with interchangeable slides) in different keys, but always written 'in C'.

A modern orchestral horn player will still get given parts for 'Horn in D', 'Horn in Eb' etc etc. and will transpose them on the fly for his modern 'Horn in F'. Which will probably be a 'Double horn in F/Bb'. Unless they follow the increasing modern practice of using replicas of the original instruments... Have I made the point that a modern hornist's life involves a LOT of transposition? Trumpet players also use instruments in Bb, C, D, Eb (and more) to play parts in various keys.

Do these facts excuse a seemingly perverse insistence on no key signatures? Perhaps. Or maybe they just enjoy being special! Anyway, it's only a symphony thing. Horns and trumpets cope with key signatures perfectly well in band, commercial and theatre music.

There's a similar thing with harp players. In traditional technique, harp is played with four fingers of each hand. No thumbs. {Whoops! Correction - no 5th finger. But the point stands.) But while today's players of every other instrument delight in 'alternative' techniques, ask a harpist to pluck a string with their thumb and you'll hit a brick wall.

  • 3
    I'm sure I've seen harpists use their thumbs.
    – phoog
    Commented Aug 28, 2021 at 18:51
  • See for example youtu.be/A0cyHGMsJBM?t=485
    – phoog
    Commented Aug 28, 2021 at 18:59
  • 5
    it's the pinkie finger that harp players don't use Commented Aug 28, 2021 at 19:01
  • @ElementsinSpace it must be opposite day.
    – phoog
    Commented Aug 28, 2021 at 19:05
  • "The fifth finger is not used." - Walter Piston, Orchestration
    – Jos
    Commented Aug 29, 2021 at 13:25

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.