Nascent Hypermeter in
Bach: The Development of Style and PerceptionHearing hypermeter in
real time involves hearing in terms of a set of well-known possibilities, or
schemas. But how did these schemas
develop? Did composers imagine
hypermeter and then write regular music so that others might hear it too? Or could hypermeter be an emergent property
of statistical regularities that exist for other reasons?
This paper
investigates these questions, looking at Bach’s music as a test case, placed as
it is between the norms of the 17th century (compound meter, irregular phrase
lengths, relatively low salience of bar lines in compound meter, few
indications of hypermeter) and those of the 19th (salient notated downbeats,
robust hypermeter, standard hypermetrical manipulations). Two hypotheses are
tested, using a corpus of ritornello movements.
First, the notated measure was being stabilized. In a variety of meters, some simple and some
compound, the metrical placement of ritornello returns and of cadences suggests
an intermediate stage in which half-bar displacements are more prevalent than
in later music, but in which they are not arbitrary, rather reflecting
hierarchies of key and of form. Second,
four-bar construction of ritornellos has a prevalence that far exceeds chance. There are furthermore signs that Bach intended
hypermetrical hearing, as ritornellos with a multiple of four bars are often
followed by episodes starting on the subsequent measure, while ritornellos that
contain one additional bar beyond a multiple of four often end with a phrase
overlap, the cadence of the ritornello coinciding with the beginning of the
episode.
Statistical evidence
for hypermeter in Bach must be balanced, however, by consideration of examples
in which a regular four-bar frame receives an internal organization that would
seem to render the frame unhearable.
Such examples suggest: that hypermetrical hearing was for Bach only one
of the possible results of four-bar construction; and that listeners should
expect the unexpected.
|
|
