Annotation for Johnson-Laird, Philip N.
Rhythm and Meter: A Theory at the Computational Level
Annotation (by Bill Tilghman):
- This article presents a theory of rhythm and meter that attempts
to explain why rhythms, unlike harmonic sequences, require only
weak computational power and can be perceived and created using
only a finite working memory. According to the theory, melodies
are divided into discreet phrases by means of various temporal
and tonal cues, and the rhythm of each phrase
consists of a sequence of event onsets based on a metrical
framework. As an aid to the cognition of rhythm, listeners can
detect structural resemblences among similar rhythmic sequences.
These resemblences are reflected in the theory by means of
rhythmic families; each member
of a family is understood as a variant of some prototypical
underlying rhythm, or UR-rhythm. An UR-rhythm is defined
by what is the most significant class of event that occurs on
each beat of the phrase. There are three classes of event, listed
here from greatest to least significance:
Thus both of the following 2-bar rhythms are members of the same
family, which is represented by the UR-rhythm notated at the
- Sync: The beat contains a syncopated note (i.e.,
an onset on a metrical unit of less importance than one that
intervenes prior to the next onset), which may or may not be
preceded by other notes, on or off the beat.
- Note: The beat begins with an onset, which may or may
not be followed by other onsets. If there are other onsets, they
do not initiate syncopated notes.
- Other: All other possibilities, such as no onsets at
all, or one or more unsyncopated notes occuring after the start
of the beat.