This perhaps just betrays my ignorance, but I don't understand why in the attached piece, in 3/4 time, certain bars, such as 10 and 12, have 6 instead of 3 quarter notes, with no change in the time signature.

What hints are there for the performer to know such a change occurs, and why is the notation as such? Is this device characteristic only of 17/18th century music?

Most importantly, how not to get confused when singing this?enter image description here

  • 5
    I'll let others elaborate, but this is about hemiola Jan 29 at 19:03
  • 2
    Do you have the edition information? other than pointing out the score has an implied switch to 6/4, and it does that to notate three groups of 2 beats, which is the hemiola, without the need to use ties over the bar line in 3/4, I wonder how the historic 18th century scores looked. My modern scores of Baroque music sometimes explicitly notate meter changes. I wonder if your edition is preserving how the meter change was original implied, not written out, of it was a modern editor's choice. Jan 29 at 21:33
  • 1
    @MichaelCurtis I looked it up earlier; the 1884 Chrysander score, at least, has just like this but with vertical strike marks through the middle half note Jan 29 at 22:43
  • 1
    Sadly I don't have any addition information about this particular score. But I suspect it was produced in Musescore by our local conductor, so it could well be that it is simply a mistake on his part. Else, from what I read about the definition of the hemiola, it doesn't quite seem to match, let alone be of any help to the person deciphering the music
    – z8080
    Jan 30 at 10:53
  • 2
    I wouldn't think it's a mistake. Whoever notated it, took some extra steps to produce these double-measures. That's intentional. Jan 30 at 12:25

5 Answers 5


What hints are there for the performer to know such a change occurs

Count the beats in the measure!

and why is the notation as such?

The answer to this question probably lies in the original sources for the opera. Unfortunately, none of these seem to be available online. In addition to manuscript scores, one can also look at manuscript performing parts. A comprehensive list of available sources may be found at RISM.

Another possible source of this information is the editorial notes to any modern scholarly edition. You should be able to find these by visiting any research music library. You may also be able to find a copy in a sheet music store, though sheet music stores are increasingly difficult to find these days.

Is this device characteristic only of 17/18th century music?

Not only, but primarily. You'll find it in the Renaissance, though it was largely less formulaic then, and you can certainly find it in the nineteenth century, though it is typically found in music that evokes the older classical and baroque styles.

While I could not find manuscript sources online, nor (unsurprisingly) the editorial notes accompanying the modern editions, I did find a scan of the contemporaneous engraved score. This has an odd quirk that I haven't seen before. There aren't any bars that are twice as long as the others; equivalently, there are no missing bar lines. Neither are the measures in question notated with ties as shown in other answers. Instead, the half note is centered on the bar line:

detail of the Walsh edition of Xerxes showing an excerpt from the movement in question

This is very odd by modern standards, but it is not so odd if you consider that ties across the barline were also avoided for dotted rhythms by writing the dot for a note that starts on the last beat of one measure on the downbeat of the following measure:

detail showing augmentation dot on downbeat

The scan can be found on Google Books (the movement in question is on page 97).

Most importantly, how not to get confused when singing this?

At first, the easiest way, if the tempo is moderate to slow as it is here, is to count as it is barred here: instead of 1-2-3, 1-2-3 count 1-and-2-and-3-and -- but be sure that the quarter note beat remains constant. Once you're more familiar with the style, where this rhythmic shift happens at almost every cadence in triple meter, you will start doing it automatically.

  • No wonder the urtext publishers revolted and used ties instead.
    – Dekkadeci
    Feb 1 at 4:35
  • @Dekkadeci indeed. We do have to remember to take "urtext" with a grain of salt. Urtext editions of Bach's keyboard works, for example, wouldn't be particularly saleable if they were published using the original soprano clef for the upper staff.
    – phoog
    Feb 1 at 7:46

If it makes you happier, you could edit those bars to insert the missing bar lines, then convert the second minim/half note to two tied crotchets/quarter notes, e.g. enter image description here

As Andy Bonner notes in a comment, it's hemiola. Every now and then the music changes to have the same beat pattern, but at half the speed (double the duration of each note). For the least subtle example of hemiola ever, listen to America from West Side Story.

  • The OP doesn't seem to be dismayed by the notation; they just need an explanation of hemiola Jan 29 at 22:44
  • 2
    That is of course a trivial edit to make but what would make me happier is to understand to what purpose this change of notation is used in the first place, assuming that it isn't a trivial mistake (see previous comment)
    – z8080
    Jan 30 at 10:54
  • @AndyBonner: The wikipedia page you linked to above describes hemiola as "three beats of equal value in the time normally occupied by two beats" - eg "♩♩♩" in the time of "♩. ♩.", but this seems to be "one beat in the time of two" ("♩" in the time of "♪♪"). Can the term refer to both, then?
    – psmears
    Jan 30 at 11:11
  • @psmears This is the same as the Wikipedia example with the note values doubled (three half-notes in the space of two dotted half-notes). It's actually more typical and would be a better example for Wikipedia.
    – PiedPiper
    Jan 30 at 11:51
  • @PiedPiper: I guess my confusion is that there aren't any dotted-half-note rhythms around the bars in question to contrast with; is the idea that the contrast is with pairs of bars from the surrounding music? (Otherwise you could equally claim that the first line of the piece is all hemiolas because it has "♩♩♩" played in the time of "♩. ♩."...)
    – psmears
    Jan 30 at 12:23

There is an urtext version available here: https://imslp.org/wiki/Serse,_HWV_40_(Handel,_George_Frideric)

The part in question is on page 161 of the score: enter image description here

...and it uses 3/4 meter with ties over the bar line.

The rhythmic effect of that passage is called hemiola which is a rhythm based on the ratio of 3:2. In this particular case it is a switch from metrical grouping by 3, which is the 3/4 meter, to grouping by 2, which is the three two-beat durations. You can think of it like a type of syncopation, because of the way the sense of strong beats shifts around. In 3/4 the third beat of measure is normally weak, but with the hemiola rhythm that third beat tied to the next one beat gets a kind of emphasis that contrast to the prevailing 3/4 meter.

Switching meter from 3/4 to an implied 6/4 is just a way to avoid notating ties over the bar lines. For good or bad that choice was made by the editor of the specific edition you have.

  • The editor's choice is actually closer to the Walsh edition than the supposed urtext edition, whose editor has chosen to add ties where it seems that none exist in the source.
    – phoog
    Jan 31 at 22:11
  • @phoog is that Walsh edition available at IMSLP? Feb 1 at 20:08
  • I don't think so. I found it on Google Books.
    – phoog
    Feb 2 at 16:22

Yes, it's a hemiola. A brief drop into 'half time', normally in a triple meter. They occur all through the history of notated music, and are sometimes written with apparent disregard for the rules of notation, as in your example.

How not to get confused? Just believe the notation. Now you know it ISN'T a mistake, it won't be confusing!


Putting my neck on the line here: it may not be a hemiola, but a poor writing, or printing, where bar lines have been missed, and ties not applied. Any piece in 3/4 time should have three crotchets in each bar (excluding anacruces!), and it's patently obvious that this piece has ended u with 6 instead in some bars.

The fact that two minims won't fit into a bar in 3/4 time is fundamental, so to keep the timing of the tune, there would need to be a bar line after 3 beats, and the second minim in that bar tied across in the form of one crotchet tied to the first crotchet in the next bar. It will all sound the same, count the same, read and play the same - and be technically correct.

I've yet to find a hemiola which isn't contained within a single bar, which abides by the 'rule' of correct number of beats in each bar. And while knowing what a hemiola is, there's no need to understand it for this example.

First port of call must be your conductor. Second, probably, re-write it so there's a bar line and tied crotchets.

And the usual unexplained dvs will follow...

  • 2
    They are hemiolas. Some editor has tried to make them even more obvious than they would be already.
    – PiedPiper
    Jan 30 at 11:47
  • See Schumann example from the wikipedia page for a two-bar hemiola example en.wikipedia.org/wiki/Hemiola Jan 30 at 12:14
  • @user1079505 - I agree totally that there is at least one hemiola there, but there's also the requisite number of bar lines - which is my point. To make it really obvious, there should be a time signature change!
    – Tim
    Jan 30 at 12:36
  • @Tim While adding time signatures (or barlines and tied notes) would indeed be strictly correct, your conductor may have thought that the intention was already very clear, and that adding time signatures at the start of 4 consecutive bars would have made it harder, not easier to read.
    – gidds
    Jan 30 at 14:29
  • It's a common enough way of making these hemiolas visible, so it's not a violation of the rules, but a special exception. Think of it as a two-bar meter change without bothering to spell out the time signature. (Except it's more like "both meters are true simultaneously," thus the fun cognitive dissonance.) Jan 30 at 15:10

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.