"ma ko MA ko SA" … "ma MA ku SA"

« previous post | next post »

Yesterday, Ben Zimmer traced the nonsense-syllable chant at the end of Michael Jackson's Wanna Be Startin Somethin back to its roots in Manu Dibango's Soul Makossa, a 1973 Cameroonian hit that played a role in the origins of disco in New York City. The chants in these songs are nice examples of a phenomenon that I discussed a couple of years ago ("Rock syncopation: stress shifts or polyrhythms?", 11/26/2007), where linguistic accents and musical beats start off aligned at the beginning of a phrase, and then go out of sync, typically with one or more of the later textual accents shifted "to the left", i.e. ahead in time, relative to the apparent musical beat.

In Soul Makossa, the basic background rhythm has a period of 16 units of about 1/8 second each, with a drumbeat on 1, 5, and 13; a stronger accent on 9; and upbeats on 4 and 12. (Beats 2 and 6 are weaker and are omitted in some repetitions of the pattern.)

Audio clip: Adobe Flash Player (version 9 or above) is required to play this audio clip. Download the latest version here. You also need to have JavaScript enabled in your browser.

Slowed down by a factor of two, it sounds like this:

Audio clip: Adobe Flash Player (version 9 or above) is required to play this audio clip. Download the latest version here. You also need to have JavaScript enabled in your browser.

In a simplified schematic form, we could represent it as something like this:

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
X
.
.
x
X
.
.
.
x
.
.
x
X
.
.
.

So far, this can be interpreted as a fairly vanilla form of 4/4 time, with the 16 minimal time units notated as 16th notes, and the quarter-note pulse on 1, 5, 9, 13.

Dibango's language, Duala, probably doesn't have anything that corresponds to English stress, but (at least to the ears of a native English speaker) the refrain seems to be accented as "ma ma KO, ma ma SA, ma ko MA ko SA", and so according to the traditional norms of European tune-text alignment, where accented syllables line up with stronger musical beats, you'd expect to see the four most strongly accented syllables lining up with the quarter-note beat, i.e. on the 16th-notes 1, 5, 9, and 13.

And that's where the first three end up — but not the final SA, which ends up one 16th-note earlier.

Audio clip: Adobe Flash Player (version 9 or above) is required to play this audio clip. Download the latest version here. You also need to have JavaScript enabled in your browser.

Schematically,

15
16
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
.
.
X
.
.
x
X
.
.
.
x
.
.
x
X
.
.
.
ma
ma
KO
ma
ma
SA
ma
ko
MA
ko
SA

At half speed:

Audio clip: Adobe Flash Player (version 9 or above) is required to play this audio clip. Download the latest version here. You also need to have JavaScript enabled in your browser.

Michael Jackson's version of the chant has one fewer syllables, but still four perceptible accents: "ma ma SE ma ma SA ma MA ku SA". And the background percussion has four square beats per bar. But only the first two of the chant's accents align with the quarter-note beat — on 1 and 5 of the 16 implicit time-units. And the others line up in two (mostly alternating) different ways:

Audio clip: Adobe Flash Player (version 9 or above) is required to play this audio clip. Download the latest version here. You also need to have JavaScript enabled in your browser.

Schematically, the first alignment is:

15 16 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
X X X X
ma ma SE ma ma SA ma MA ku SA (ma ma

And the second one:

15 16 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
X X X X
ma ma SE ma ma SA ma MA ku SA (ma ma

The two versions at half speed:

Audio clip: Adobe Flash Player (version 9 or above) is required to play this audio clip. Download the latest version here. You also need to have JavaScript enabled in your browser.

It's worth noting that Michael Jackson's first alignment divides the measure into the series of time intervals

(2+1+1)+(2+1)+(2+1)+4+2 =

4+3+3+4+2

which is a cyclically permuted version of the son clave: 3+3+4+2+4.

And the second alignment has the interval sequence

(2+1+1)+(2+1)+(2+2)+3+2 =

4+3+4+3+2

which is a similarly-rotated version of the rumba clave 3+4+3+2+4.

I should emphasize my opinion, explained at greater length here and here, that the key to understanding text-setting in African-influenced popular music — which is most of the music of the past century or so — is to recognize that the underlying metrical pattern is often not just a regular beat created by hierarchical binary or ternary subdivisions, but rather something more like the habanera or the son. On this view, the "syncopations" are not just random interesting shifts of accents "off the beat" in performance, but are rather accents placed "on the beat" relative to a different rhythmic pattern (of which several may be going on in parallel).

I don't have any larger point to make about this, for now — but I think this stuff is interesting and deserves further study. Of course, it makes more sense to look at English lyrics where we know where the lexical stresses are — it's plausible to doubt my perceptions of accent patterns in these chants. But I enjoyed listening to the examples that Ben brought up yesterday, and I thought the metrical details were worth a post.



4 Comments

  1. peter said,

    June 28, 2009 @ 4:24 am

    I am reminded of that witty explanation of musical syncopation as:

    "emPHAsis on the wrong syLLABle"

  2. Randy Alexander said,

    June 28, 2009 @ 9:11 am

    For the relationship between MJ's chants and the clave rhythms, I'm not sure "permutation" would be the best way to describe them, because that word is used in musical set theory to mean inversion (of pitch) and retrograde (pitch or rhythm).

    http://en.wikipedia.org/wiki/Permutation_(music)

    For readers who didn't catch how these rhythms are permuted, the MJ versions start on what would be the last group of each of the clave patterns.

    For example, the son clave pattern is:
    3+3+4+2+4

    But MJ starts on the last group (4 beats), so he's taking that group and moving it around to the front:
    4+3+3+4+2

    Musical set theory does sometimes use "rotation" to describe this process of starting on an earlier or later element in the pattern, but it's not so common. If I were to personally choose a term for this, I would borrow "translation" from geometry, but this is often used to show an analogy with pitch transposition (translation along the y axis) as opposed to a different process (translation along the x (time) axis).

    This "rotation" or rhythmic "translation" is very common in African music from a western perspective. In the west, we understand music to have meter, with distinct beginnings and endings to each measure, but from what I understand of African music, that is an alien concept there. A "measure" in a lot of African music is more like a circle — there is no "beginning" per se.

    [(myl) Thanks! I meant to use cyclic permutation in the general mathematical sense, but you're right that there could be confusion with the musical meanings. Perhaps "rotation" would be better.

    There's some interesting work on the mathematics of such cyclic patterns as applied to musical scales and rhythms, e.g. in the work of Godfried Toussaint, cited here.

    But my understanding is that different "rotations" or whatever of the same sequence are actually treated as different rhythms, in African, Afro-Cuban and other musical traditions. Thus Toussaint, in the cited paper, derives ten named 12/8-time bell patterns as rotations of three basic sequences. ]

  3. Aaron Davies said,

    June 29, 2009 @ 1:32 am

    the bloodhound gang version usually follows MJ's form, but occasionally throws in an extra "ma" before the third group, giving "ma ma SE ma ma SA ma-ma ma ku SA".

  4. Dennis Brennan said,

    June 30, 2009 @ 2:55 pm

    On a realted note, check the link in my name above.

RSS feed for comments on this post