Looking at the following piece, the information says that the key is G minor. However. the score in the "original key" has a key signature of D minor. Moreover, the score transposed to A minor has a key signature of E minor.

I am aware that:

  1. The music is not strictly in a minor key, but rather in a mode, although I don't see why this would have any effect.
  2. There are accidentals next to the notes that make the music effectively be in the listed key (so for G minor, the E's are notated with a flat to make it sound G minor, although with a signature of D minor).

I'm not sure if this is bad editing and if when I arrange the piece I can/should use the listed key as the key signature. Or maybe it's some convention that I've yet to see in this type of music.

page 1

page 2

  • With Eb and F# featured, and the first two notes making most of the root harmony, yes, it looks like it is in G minor. Possibly, someone thought it was G Dorian, which could be represented by 1 flat (for F), but then it was too late...
    – Tim
    Dec 8, 2016 at 17:07
  • 2
    Realise that publishers sometimes do exactly what they want regardless of how much head scratching it leads to
    – Neil Meyer
    Dec 8, 2016 at 17:20
  • 3
    @NeilMeyer "...publishers sometimes do exactly what they want regardless of how much head scratching it leads to..." Quite the opposite: in all the examples in this thread, the publishers printed what the composer actually wrote. If you want to replace the entire history of music notation with what is currently fashionable among the "musical notation style police", you are welcome to your opinion - but don't try to force it on the rest of the world!
    – user19146
    Dec 8, 2016 at 21:34

3 Answers 3


You see this occasionally in Bach as well! Check out this piece in D minor:

enter image description here

The short answer is that, as musical practice was settling towards tonality, some composers were still using key signatures that, by today's standards, are missing a flat. These key signatures typically arose because the composer was using a transposition of a church mode.

In effect it looks like the pieces are being written in Dorian if you go by the key signature and ignore the later accidentals. (In fact, the Bach piece above is sometimes given the nickname "Dorian"!) But since the accidentals are there, it's just a piece in minor.

  • Also that F sharp resolves like the leading tone of a minor key should. If we were under some sort of mode it would probably not have that resolution.
    – Neil Meyer
    Dec 8, 2016 at 17:31
  • Yes, but notice that that F# is at a cadence point on "Gottes"; but in m. 3 it's F-natural!
    – Richard
    Dec 8, 2016 at 17:32
  • 5
    This practice with "minor" key signatures survived at least up to some of Haydn's early works. In fact, in some of Haydn's first uses of a "modern" minor key signature, he also wrote "unnecessary" accidentals in the score as if an old-style key signature was still being used! Being a practical working performer and conductor as well as a composer, he was probably well aware of what his contemporaries were used to reading, and the sort of mistakes they would make when playing from the unfamiliar "modern" minor key notation.
    – user19146
    Dec 8, 2016 at 21:42
  • Thanks. What would be the longs answer? I would assume that if the piece isn't in Dorian then it wouldn't be missing flat, but other "discrepancies" would occur according to the mode. How are the written key signatures chosen for a piece in this case? Dec 9, 2016 at 14:28
  • Sorry, but the Dorian by Bach that's not really in D minor. I don't see any point in your reasoning that a presence of accidentals means the piece is not dorian. Maybe it's not strictly in dorian mode, but it's certainly not in D minor. (What are pieces like this in precisely and technically can be a topic of a lot of debate, as they break a lot of modal principles, but they still carry their modal nature.)
    – yo'
    Oct 15, 2022 at 18:20

The key signature takes the form it does because the Dorian and minor modes were considered very nearly synonymous by many, possibly most, Baroque theorists (including Rameau). Dorian mode was seriously considered to be the precursor to the minor mode. There were a couple of reasons for this:

  • the standard mutable tones for the Dorian mode were 6 and ♭7 (e.g., B and C in D Dorian), the former frequently being flatted (particularly, in imitative counterpoint, in tonal responses that turn toward the subdominant region, as happens in the Schütz example), and the latter being raised (particularly in cadences).
  • The Dorian mode, like the major mode, requires only that the 4th degree be raised to accomplish a perfect cadence to the dominant. Contrast that with the Aeolian mode, where both 4 and 6 must be raised to cadence to the dominant.

Baroque (and some Renaissance) key signatures are often written with one fewer flat than is commonly used now. (Sharp keys seem to sometimes use one less sharp but I think I have seen one more sharp too.) There are some studies on the matter but most are not freely available on the net. Wikipedia has a short paragraph about this practice under "Key Signatures."

It's been suggested that one may need fewer written accidentals with the Baroque version. (There are also suggestions in various publications and on various net sites that minor keys should have "harmonic minor" signatures.)

I did a bit of counting by hand (mostly Bach, and a few other short pieces) to check if the number of written accidentals (those not in the key signature) would be smaller. I didn't find much difference (but I didn't keep the results as the sample was small and perhaps not representative.) The current practice of using the key signature of the relative major has the advantage of being regular and not needing extra variations for various modes.

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.