In my apprenticing to jembe masters in Mali, I was thrown into real music almost from the very first day. My masters were not experienced in formalized music teaching because in their milieu there is no such thing. The appropriation of musical skills in their tradition is fully embodied in performance practice. Even when I paid them for sitting down and showing things to me, which was part of my earliest and mostly frustrating approach to jembe playing in Mali, my masters weren't willing (nor perhaps able) to slow down, segment phrases, and play rhythms in a “simplified” way and then "add" the appropriate swing feel only in a later step. I thus needed to imitate full-fledged music, including its swing timing patterns, from the outset. In the learning-by-doing pedagogy of Malian drumming, one begins with the nuanced rhythms, without any sense of a simplified (quantized) basic pulse. Even small kids drumming along on toy-drums or chairs, play with swing. (See the example of a novice ensemble playing with full swing in the page on field-recordings of Bamana drumming.) Symbolic representation such as written notations, which may convey an abstract sense of isochrony relatively independent from performance practice, are not used in this pedagogy.

Having had no formal training in Western music, I grew up (in my "second socialization" in Mali) with swung subdivisions affording metric reference. In both musicological and music-psychological literature, swing-timing patterns were explained as the deviation from some underyling structural isochrony. This idea appeared to me as unrealistic, falling into the trap of equating an analytical framework (a "ruler") with a perceptually relevant structure (the "real thing"). It seemed implausible to conceive of the conspiciously stable and basic timing patterns in Malian drumming as "deviations" from an allegedly underlying isochrony. My qualitative research thus suggested the alternative hypothesis that swing timing patterns in this cultural context can constitute full-fledged metric reference structures.