I didn't think this topic would become an issue in my composing -- after all, what could be more basic than notating accidentals? -- but recently, while working with a couple of pianists to record my pieces, it came to my attention that my notation isn't always that clear, or even if it's unambiguous, it may still be tricky to read, thereby leading to mistakes in performance.
I currently use Lilypond for my scores, and it offers quite a few alternatives for notating accidentals: https://lilypond.org/doc/v2.23/Documentation/notation/displaying-pitches#automatic-accidentals
I had been using the "default" setting, not having paid attention to this issue until recently, but it turns out that it leads to notation that isn't always the easiest for pianists to read.
So my question is: what's the best style to use? What would you recommend? I see "piano" and "piano-cautionary" as the most likely candidates, as least for piano pieces. But what about other types of music? What do y'all think?
And in general, what conventions do you use in notating accidentals?
P.S. this topic is not talking about spelling notes as sharps or flats. That's a separate issue, and in Lilypond generally isn't a problem because it already requires the composer to input sharps or flats explicitly. The issue is with the printing of accidentals after another one has already printed elsewhere: e.g., if there's a C# in a bar and another C# in the same bar -- should the 2nd # be printed or not? What if they're an octave apart? What if there's a C natural elsewhere in the same bar? Or in the previous bar? Or in another staff of the same instrument (e.g., RH vs LH in a piano score)? Etc..
Replies
Hi, i'm using also Lilypond time to time and use generally "Neo-modern" option in Lilypond for cautionnary accidentals.
S.
Do you always use neo-modern? Or does it depend on the type of piece you're writing?
I'm always writing Neo-modern music 😂
Fair enough. 😅
Hi HS. Although Im a pianist, I only have general guidelines in this regard. For more traditional harmonic structures which clearly have tonal centers, I would use accidentals that show the nature of harmonic relation, which include double flats and double sharps, of course. It is a way of leaving breadcrumbs for those who want to see the harmonic path. Although, in this context I also think about a players' easability with reading the score, and balance that out as well. For example - if a Gb7 sets up a modulation to Cb Ma. ---- wouldn't write the new section in Cb Ma, but B Ma.
With my extended harmonic pieces - I disregard trying to spell a connected tonal center, as that is always moving, and just see each step as a cluster of notes that reference a type of chord - which doesn't necessarily describe it's relation to the previous cluster.
When writing 12 tone tunes, for easy readabilty I'll (mostly) choose either: mostly sharps and naturals, or mostly flats with naturals, so that it is much easier within each measure to see what has been altered.
This, of course, is a personal approach, and wouldn't expect it to line up nicely with composition 'pedagogy'
Also, Im not inclined to let the software do its thing - unedited.
Ps. Oh, yes, the cautionary accidentals - when allowed, the software overuses this - to ridiculous extent, sometimes.
Yeah one reason I didn't use "piano" or "piano-cautionary" in my original score was that it introduced a lot of printed accidentals -- some of which I felt were unnecessary. But as it turns out, the pianist found the notation rather confusing, and therefore misread quite a number of implied accidentals. I have since switched to "piano-cautionary" to be on the safe side -- better to print a few more accidentals than to risk wrong notes in the recording, IMO.
I suppose if it came down to it, I could do it manually -- Lilypond does allow you to force the printing of individual accidentals even if it wouldn't have been printed by default. But obviously it's rather tedious to micromanage the notation this way; so if there was a default option that did the job 95% of the time, I'd rather go with that and just tweak a few individual notes as needed.
The notation should comply with the harmony unless you deliberately want to notate a couple of notes enharmonic. As long as the music is tonal, the type of music is of no importance.
I think there's some misunderstanding as to what exactly the issue is that I'm referring to here. The issue is not so much whether a note is spelled as a sharp or a flat -- Lilypond input syntax already requires explicit spelling of sharps or flats, and generally does not have a problem in this department.
The issue is when, and whether, an accidental is printed for a note after the same accidental has been printed elsewhere in the score. For example, let's say in measure 10 beat 1 there's a C#, and in beat 4 there's another C#. Should the # be printed for the 2nd note as well? What if the first C# is an octave higher/lower? What if it occurs on another staff (e.g., in the piano's LH part, where the 2nd C# is in the RH)? What if there's a C natural in measure 11 -- should a natural sign be printed or should it be assumed that the new measure has "reset" all previous accidentals? What if the previous accidental was in a different voice in the same staff (e.g., in a fugue or a combined instrumental score where two instruments are sharing the same staff)?
These are some of the areas where sometimes things can become confusing, if the composer assumes that a new measure has "reset" the accidentals to the current key signature, for example, but the pianist is still thinking in terms of the previous accidentals due to a temporary modulation in the music. Or if there's a C# in the LH but a C natural in the RH -- should a natural sign (optionally parenthesized) be printed so that the pianist does not mistake the C natural for an implied C#?
"
"For example, let's say in measure 10 beat 1 there's a C#, and in beat 4 there's another C#. Should the # be printed for the 2nd note as well? "
I would say no - unless too many notes are written in-between and you want to be certain the player remembers - use sharp in parentheses.
"What if the first C# is an octave higher/lower? "
To my thinking, the sharp only applies to the exact pitch it was originally written.
"What if it occurs on another staff (e.g., in the piano's LH part, where the 2nd C# is in the RH)? "
The 2nd C# must be sharped - but an argument can be made that the sharp is not needed if it is at the same pitch, but if it is layed out in fugue fashion - different voices, I would trouble with adding a sharp - at least in parentheses.
"What if there's a C natural in measure 11 -- should a natural sign be printed or should it be assumed that the new measure has "reset" all previous accidentals?"
I think it should be assumed the new measure resets.
"What if the previous accidental was in a different voice in the same staff (e.g., in a fugue or a combined instrumental score where two instruments are sharing the same staff)?"
Two voices - same pitch - 2nd accidental probably not needed - (could use paretheses). If different pitch, then Yes, a new accidental is required.
"These are some of the areas where sometimes things can become confusing, if the composer assumes that a new measure has "reset" the accidentals to the current key signature, for example, but the pianist is still thinking in terms of the previous accidentals due to a temporary modulation in the music."
I wouldn't concern that the pianst is still thinking in terms of previous bar's accidentals.
"Or if there's a C# in the LH but a C natural in the RH -- should a natural sign (optionally parenthesized) be printed so that the pianist does not mistake the C natural for an implied C#?"
Yes, I would, although, again, the sharp (should) only apply to an exact pitch. But if it is the exact pitch but changing clefs, I would definitely sharp it again.