urrhythm -- characterize Johnson-Laird rhythmic prototypes in a Humdrum passage


urrhythm inputfile [ > outputfile.urr]


The urrhythm command outputs a single Humdrum spine (**URrhythm) containing data representing rhythmic `prototypes' (Ur-rhythms) evident in a musical passage. The command implements a variation of Johnson-Laird's theory of rhythmic prototypes (see REFERENCES below). In order to identify the rhythmic prototypes, urrhythm requires information about note onsets and metric position. These may be provided via two input spines: **kern (or **recip) and **metpos.

Johnson-Laird's rhythmic-prototype theory can be applied only to musical passages conforming to some established metric context, such as 2/4, 3/2, or 12/8 meters. The urrhythm command handles all regular types of meters (simple and compound, duple, triple and quadruple). Specifically, any meter having a "numerator" of 2, 3, 4, 6, 9, or 12 can be processed. Urrhym adapts to changes of meter, but is unable to handle irregular meters. If an irregular meter is encountered in the input an error is generated and the command terminates.

urrhythm characterizes each beat in a passage as belonging to one of three beat types: Note (N), Syncopation (S), or Other (O). Only major beats are characterized in this way. Hence, in 3/4 or 9/8 meters, three beats will be characterized for each complete measure. Similarly, in 4/2 and 12/16, four beats will be characterized for each complete measure.

A "Note" (signified in the output by the letter `N') is defined as a beat that coincides with a note onset.

A "Syncopation" (signified by the letter `S') is defined as arising when no note-onset happens on a beat whose position in the metric hierarchy is greater than that of the most recent note onset. By way of example, imagine a measure in 4/4 meter containing a quarter-note, followed by a half-note, followed by a quarter-note. The third beat position does not coincide with a note onset. The most recent note onset prior to the third beat occurs on beat two. Since beat three is a more important metric position than beat two, beat three is deemed to be syncopated.

Syncopated beats can happen only after the first note onset; subsequent syncopated moments will require another note onset (i.e. two syncopated moments can't occur in a row without some note onset intervening).

An "Other" (signified by the letter `O') is any beat that is not a Note (N) or a syncopation (S).

It is recommended that output files produced using the urrhythm command should be given names with the distinguishing `.urr' extension.


The urrhythm command provides only a help option:
-hdisplays a help screen summarizing the command syntax
Options are specified in the command line.


The following example illustrates the operation of urrhythm. The first two spines (**kern and **metpos) constitute the input. The third spine (**URrhythm) is added by the urrhythm command. All three spines are given in the output.


The urrhythm command is currently unable to handle Humdrum spine-path changes -- such as join-path, exchange-path, or split-path. If spine-path changes are encountered an error is issued and the command terminates.


DOS 2.0 and up, with the MKS Toolkit. OS/2 with the MKS Toolkit. UNIX systems supporting the Korn shell or Bourne shell command interpreters, and revised awk (1985).


**kern (2), **metpos (2), metpos (4), **recip (2), timebase (4), synco (4), **URrhythm (2)


The urrhythm command differs from Johnson-Laird's theory in the definition of syncopation. Johnson-Laird's theory requires that a listener be able to identify a syncopation retrospectively. That is, a listener is able to determine whether the current beat is a syncopation, only by determining what happens at the beginning of the next beat. The algorithm used here avoids the theoretical assumption of backward listening. (See Simpson & Huron, 1993.)


Johnson-Laird, P. "Rhythm and meter: A theory at the computational level," Psychomusicology, Vol. 10 (1991) pp. 88-106.

Simpson, J. & Huron, D. "The perception of rhythmic similarity: A test of a modified version of Johnson-Laird's theory," Canadian Acoustics, Vol. 21, No. 3 (1993) pp. 89-90.