The modern conventions for meter are, for example, that in 4/4 time, the metrical emphases are S-w+-s-w-, 3/4 is S-w-w, and so on.

Did Bach understand meter in these more-or-less strictly regulated ways, or were metrical emphases more fluid for him?

The question is particularly relevant in fugues, because the subject does not always enter on the same beat, particularly in stretto passages like the one below from Bach's fugue in C minor (WTC II, BWV 871), mm. 16–17. The subject appears first in the alto voice, beginning on the "and" of 1, and then enters in the soprano on the "and" of 3.

BWV 871, mm. 16-17
Image source: IMSLP (Hermann Keller, ed.; Peters)

Would Bach have expected

  1. the subject maintains its "first-entry" stresses ("and" of 1) even when its metrical placement changes; that is, the soprano would put the main emphasis on the Eb in m. 17, b. 3,
  2. the subject's stresses follow its actual metrical placement; meaning soprano emphasis on the G in m. 17, b. 1, or
  3. Bach was agnostic about meter, so metric designations (time signatures) and bar lines were more notational convenience than performance instruction (so to speak)? The entire passage would be played with each beat being relatively equal.
  • 2
    Interesting question. There is quite a lot of material from the period about meter, and I suppose it's probably fairly safe to assume that Bach's view was in the mainstream (i.e., not 3), but I don't know whether he is on the record on this question. If he isn't directly, his son CPE probably had something to say about it in his famous "essay."
    – phoog
    Commented Jan 11 at 7:37
  • 1
    @phoog Here's a promising-looking paper, but I can't read it just now: Mid-Bar Downbeat in Bach’s Keyboard Music. Both the journal and the author are of high repute.
    – Aaron
    Commented Jan 11 at 8:39
  • 3
    I feel the question is splitting hair a bit. Difference between 1& and 3& is not so big. WTC were composed for instruments with no or limited dynamics, so differentiating between them would be even more difficult, especially at the rather not very fast tempo of the piece. Then, regarding point 3. this fugue, and other Bach pieces in general follow the meter consistently, there are no unexpected 2/4 or 6/4 measures between 4/4, which suggests that the whole composition is purposeful. Commented Jan 11 at 8:40
  • 2
    Great question. I don't have an answer (wondered about this a considerable amount myself), but what I would highlight is that many of Bach's pieces do give substantially more relevant instruction than just the meter: many of them are labelled by types of dance. A Sarabande will always have very different emphasis from a Menuett, though both are in 3/4. Commented Jan 11 at 10:42
  • 1
    @ToddWilcox and Aaron I doubt that Mozart's ideas about meter were all that different from Bach's -- at least I am fairly sure that they were more similar to Bach's than to Rachmaninoff's. Certainly, if you look at the counterpoint in Mozart's Requiem, the main subject of the introit starts on either beat 2 or 4, and the Kyrie subject starts on 1 or 3. I think the importance of multiple levels of metrical hierarchy is sometimes overstated: sometimes it's a useful concept but often it's not. This may have changed over time but it's not absolutely tied to period or style.
    – phoog
    Commented Jan 12 at 12:36

2 Answers 2


This is a very interesting question, and I wish I had the time I would need to incorporate more source material into this answer.

Some thoughts:

  • in a passage such as the example, in 4/4 meter with only eighths and 16ths in the contrapuntal texture, the distinction between 1 and 3 is largely theoretical. This will be true in any period. The concept of good taste is always present in treatises of the baroque period. If altering the articulation or phrasing of a fugue subject makes sense in some context, it should be considered, whether because the subject lies differently in the meter or for any other reason. (This also means that different choices may be reasonable -- there may be incorrect choices, but that doesn't mean that there's only one correct choice.)

  • treatises that do talk about metrical considerations tend not to go much past a two-level hierarchy of "good" notes and "bad" notes, usually in the context of articulation: bowing of string instruments, tonguing of wind instruments, keyboard fingering, etc. Similar to the previous point, beats 1 and 3 are certainly different if you are playing half notes, possibly different if you are playing quarter notes, and probably not so different if you are playing eighth notes or shorter.

  • while we could dispose of the example given in the question through the above points, there are probably more interesting examples to be found. A few weeks ago I was looking at the Credo from the B-minor mass. It is in 4/2 (notated as alla breve), and the theme enters mostly on beat one, a few times on beat 3, and once on beat 4. But this example is also of limited use because the theme is in whole notes (with an initial breve), so when you start on beat 4, all of the notes start on a weak beat. But what of the performance of those statements starting on 3 versus those starting on 1? Since we have text here, the most likely is that they should be the same; otherwise we would sometimes have CRE-do IN u-NUM de-UM instead of CRE-do in U-num DE-um.

  • back to keyboard works, we have the first fugue of the Well-Tempered Clavier, where in measure 15 the tenor enters with the theme a beat after the alto. Should the stress be different? Probably not in general, but here the motion is in eighth notes, and even with text a rhetorical variation in stress could be desirable. For example, suppose this were adapted into a choral fugue with the text wir danken dir: There's no reason to avoid sometimes singing "we THANK you" and sometimes "we thank YOU'"; variety is the spice of life. (In fact this theme is similar to the subject of BWV 29.2 on the text "wir danken dir, Gott"; that fugue appears twice in the B-minor mass, first as Amen and then as Dona nobis pacem.)

In summary, a strictly hierarchical approach to meter at multiple simultaneous levels is probably not indicated, nor is a strict approach to the application of meter to phrasing or articulation when a fugue subject or other melodic fragment lies differently in the meter compared to its initial statement. As always, good taste prevails.


Bach's understanding of meter is good enough that his vocal works match word accents to musical accents (which makes for "singability") and you figure out an explicit stylistic exception like a hemiola out without the need for a meter change.

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.