Question about transposing instrument key signatures in modern scores.

I should hang my head in shame not knowing the answer - but it does seem to be a bit subject to opinion so I'll ask here.

As per the title, should transposing instruments have key signatures in 'no key' scores?

I've got away with no key signatures on handwritten scores - at least our local orchestras haven't questioned it for clarinets/co anglaises  (and never for brass anyway) but doing more posh scores with notation software should I apply key signatures, like a clarinet in A should have the Eb key signature (as if the work is in key C or A minor)? Or should I just put in accidentals as I would any other instrument? 

Your opinions would be most appreciated.

Thank you.

No, if you decide to not use key signatures, no instrument should have them. Check scores such as the Rite of Spring for examples. Sibelius has the "no key" option already, that hides any signature on every instrument.

Thank you.

As I thought and - it makes sense. It's how I worked in the past but an opposite opinion has been in the air.

By the way, I got caught out by the Rite. You're right so far as the keyless sections go. At rehearsal 13 in my score, he changes to 3 flats (as in Eflat major) and the transposing instruments get their correct key signatures. He's in a key there. I don't know why - accidentals are as dense as anywhere else in that score. It posed a question. If he can suddenly go into a key, could the "keyless" sections be seen as in C major, for which key signatures for transposing instruments would be right.

8 hours ago, Quinn said:

Thank you.

As I thought and - it makes sense. It's how I worked in the past but an opposite opinion has been in the air.

By the way, I got caught out by the Rite. You're right so far as the keyless sections go. At rehearsal 13 in my score, he changes to 3 flats (as in Eflat major) and the transposing instruments get their correct key signatures. He's in a key there. I don't know why - accidentals are as dense as anywhere else in that score. It posed a question. If he can suddenly go into a key, could the "keyless" sections be seen as in C major, for which key signatures for transposing instruments would be right.

 

The parts with key signatures are relatively more diatonic, if you check the later sections in 4 and 5 flats you'll see that he sticks mostly to a single diatonic scale with maybe only 1 altered degree. The section with 3 flats is also relatively "grounded" (but ultra dissonant and non-functional), notice how there's an Eb7 "home" chord somewhere almost throughout, or at least the Eb-Bb pair. This specific case a bit of a close call, though, personally I'd have notated it as keyless. There's no section notated as C major IIRC (e.g. strings without key but the clarinets in Bb with 2 sharps).

Rehearsal 37 has the typical Rite-ish language, first a typical C7-ish octatonic chord with the characteristic major/minor 3rds (plus a b9), and then the woodwinds play above in another scale that's diatonic. Since none of this would fit any key signature, he uses none.

It's also possible to notate playing in 2/3/4 keys at once if each part sticks to a key very consistently (check Bartóks 1st bagatelle, Ives' 2nd sonata, reh. 91 of the Rite, Szymanowski's 1st quartet), but I don't recommend it because it's confusing.

There's a lot of personal taste involved, however, check Prokofiev's piano sonatas 2-9: all of them use mostly the same extended tonality and each movement ends in a clear tonic, but some have signatures and others don't. I personally don't use key signatures at all, unless I'm writing in common-practise harmony (or almost).