On the Modality of Spinning
Limitations and Transformations of Siteswap when Applied to Flexible Props
This article examines the limits of siteswap—a rhythmic juggling notation—when applied to flexible and torque-based props. It shows that the absence of dwell time and spatial specificity makes the notation precise for rigid objects but insufficient for poi and mixed ensembles.
§ 1
Siteswap as an Abstract Model
Siteswap is an abstract model of juggling notation that records the rhythmic order in which objects appear at a point of interaction. It specifies how many beats after the current moment a given object must be used again.
However, siteswap does not describe the trajectory of movement, spatial configuration, or physical characteristics of performance. It therefore functions as a rhythmic scheme stripped of spatial specificity and kinetic detail.
This assumption proves effective when describing traditional forms of juggling with rigid props, where the primary movement reduces to a throw followed by a catch.
§ 2
Dwell Time and Continuous Dynamics
When this system is transferred to props with continuous dynamics, such as poi, fundamental discrepancies arise. The principal one is that object dwell time—which varies substantially depending on technique—is not represented in the notation, despite its critical role in shaping rhythm and movement structure.
In classical juggling, dwell time is treated as an inertial element that does not affect basic rhythm: an object may remain in the hand for one-third or half a beat without invalidating the pattern.
With poi, the situation differs: holding and transferring the prop are not discrete events but a continuous bodily action involving impulse transfer, sustaining rotation, and plane control. The emphasis shifts from the moment of release to the phase of sustaining movement, which traditional notation omits entirely.
§ 3
Loss of Descriptive Precision
This causes siteswap to lose descriptive precision, remaining merely a scheme of events detached from physical context.
Pattern 3, which in juggling denotes an even alternation of throws between hands, may correspond in poi spinning to many distinct forms: from a wall-plane spiral to an asymmetric weave or a pendulum with active correction.
All these realisations preserve the same rhythmic structure yet differ in hold character, torque distribution, spatial dynamics, and bodily interaction.
This becomes even clearer with complex patterns such as 5223. In siteswap this is an alternation of high, holding, and medium throws. In poi the same sequence may be interpreted as movement where «2» is an active rotation phase (not passive waiting before a plane change) and «5» is an intensified acceleration.
§ 4
Terminology and the Cascade
The notation continues to describe rhythmic structure but provides no information about performance form. It records when the next interaction should occur, but not what happens between interactions or by what means.
Terminology borrowed from classical juggling is also problematic. The notion of a «cascade» is applied by analogy yet has no direct equivalent in poi.
The absence of throw as a discrete event, the continuity of trajectory, and the influence of torque make exact reproduction of cascade structure impossible. The claim that «there is no cascade in poi» reflects not a denial of a rhythmic analogue but the impossibility of transferring form without distortion.
§ 5
Group Synchronisation
This problem intensifies in group work or when synchronising performers who use different types of props.
One participant juggling balls performs 3 as a standard cascade. Another, spinning poi, uses the same rhythmic pattern but at the level of bodily kinematics performs a fundamentally different action.
Without specifying performance parameters—hold method, rotation direction, plane, character of acceleration—a semantic divergence arises despite formal notational agreement.
§ 6
Conclusion
All of the above allows us to conclude that siteswap, while an effective abstraction, is not a universal system for describing movement. Its applicability to flexible and torque-based systems requires rethinking: either toward extended models that account for additional parameters, or toward a conscious limitation of its domain of use.
In its current form, siteswap continues to perform an important communicative and rhythmic function, but cannot serve as the sole descriptor in bodily practice that relies on continuous, phase-dependent movement.
Siteswap is thus less a language of description than a tool of ordering. It tells the performer when the next act of interaction with an object should occur, but leaves open the question of how that act is realised. This makes it valuable as a rhythmic grid, yet limited as a model of movement—especially in disciplines where the boundary between throw and spin is blurred.