# Do accidentals override key signature and previous accidentals?

I am wondering how the accidental in the first chord (see what is circled) is played? Does any accidental simply move the note up or down a half-step from what the note is supposed to be based on the key signature or does it override the key signature all together? Specifically, is the accidental G# played as a natural G (since G is already a flat in the key signature - thus a half step up) or actually as a G#?

• Out of curiosity, what piece is this? I'm assuming 20th century, and there are some fairly strange enharmonic choices being made. Mar 10, 2014 at 21:47
• Notice that you also have natural accidentals in the first chord (A-natural, rather than A-flat). If the accidentals were cumulative (e.g., a sharp sign meant "half step up from what this would be otherwise", then the A-natural would have been denoted with a sharp sign on the A-flat. If you're playing that triad as A-C#-E, then the only consistent interpretation is that the sharp sign on G means G#, not "half a step above Gb, thus G-natural." Mar 10, 2014 at 23:47

The chord played is Amaj7, made up with A C# E and G#.The key sig. is Db/Bbm.I guess that the G# is shown instead of a possible Ab, which is technically correct for that chord.

All accidentals over-ride the key sig., for the rest of the bar they're marked in.Sometimes the author will be helpful and remind the player that a particular accidental is not needed in the next bar by using another accidental,which can get confusing !

Any accidental will show the actual note to be played, not just sharpening a flat, for example. In that key sig.,the Gb is changed to a G# with an accidental. It doesn't 'sharpen the flat' to make it a G natural.

• I find that almost every time I'm confused by a courtesy accidental, it's because I missed an earlier accidental (or, in some cases, a key change)--so even then, they're helpful. Mar 10, 2014 at 21:33
• @KyleStrand, agreed, but Tim's right, it's unnecessarily confusing. This is why I think composers should use parentheses around courtesy accidentals, much like in the OP's example. Mar 10, 2014 at 21:45
• Definitely paranthesise them! But definitely include them in such cases, too. Mar 11, 2014 at 1:00

Accidentals override key signatures and previous accidentals. The circled chord has two G# and and one C# note. Having "additive" accidentals would make it very hard to read music. In this excerpt, the next octave chords in the top staff would then be B-flat, then B-doubleflat, and then either G natural (if adding to the previous accidental) or or G double flat (if adding to the key signature)...

• @Tim - nonpop is pointing out how confusing it would be if accidentals were additive. Putting naturals on the B's on beat two of the RH would have to mean Bb as in the key signature, the flats on the Bs on beat 3 would mean double-flat since B is already flat in the key signature. The fact that you got lost by it proves the point! Mar 10, 2014 at 21:43

As has been mentioned in other answers, alterations are not relative, they're absolute. So a sharp makes the note sharp, regardless of the alteration it had due to the key you're playing in; so if the key states Ab and you suddenly see an accidental sharp on an A, you play A#. Same goes for the inverse case.

So I just wanted to add that there is a sign to remove an alteration: the natural. Your example includes the natural sign in the second chord on the treble clef; that one is used to "neutralize" a sharp or flat, so you play B natural there (which is what you were thinking the sharps did in the first chord).

• Because alterations are not additive, it is possible in some rare cases for a sharp sign to actually lower a note's pitch (if the note was double-sharped earlier in the measure) or for a flat sign to raise it (if the note was double-flatted). Oct 16, 2014 at 19:45

Accidentals take precedence to any alteration from the signature. Moreover, they apply until the end of the bar for every pitch. If there is a G# in a bar, then, until a new accidental or the end of the bar, all the following Gs on the same octave should be played sharp.

• But since that last rule is not always followed in practice, I recommend anyone generating sheet music to explicitly mark the accidental (or lack thereof) in each octave within the measure. Mar 10, 2014 at 17:25
• @CarlWitthoft: I would suggest that the biggest reason for adding "cautionary accidentals" in other octaves would be to make it clear that a lack of an accidental was deliberate; they should not be included in cases where the lack of an accidental would be "obvious" [e.g. in D major, if an E7 chord is followed by an A7 whose G is in a different octave from the preceding G#, I would see no need for a natural sign on the G.] Oct 16, 2014 at 19:50

The key signature is always secondary to any accidentals next to the note. In the chord you circled you would play a G#, C#, and another G# instead of a Gb, C, and another Gb. They never cancel each other out just play what is written.

The key signature is not overridden. This is just a chord borrowed from a different key and the accidentals continue until the end of the measure unless otherwise noted.

As noted in the other replies, accidentals are "absolute", with the meaning unfazed by preceding material or key signature. There is one exception: when "weakening" a previous double accidental (key signature or previous in the bar), the resulting single accidental is often printed combined with an immediately preceding natural sign.

While the existing answers are correct for modern music, there was a time before the invention of the natural sign in the middle of the seventeenth century, when music was less chromatic and harmonies rarely strayed very far from the tonic. In those days, the sharp sign was used to cancel the flat sign.

An example may be found at the end of the first Kyrie of Byrd's Mass for 4 voices, the last note of the second staff in the image.

• – Richard
Dec 7, 2021 at 13:16
• @Richard nice. I didn't realize this practice persisted so late. I found naturals in a movable-type print (similar to this but much later) from 1660, as well as in a print using a different style of movable type that I didn't find a date for, but it was music of Dieterich Buxtehude. Dec 7, 2021 at 13:24