Fidgeting is Not Random: Rhythmic Leg Motion, Speech, and Gesture

Posted: 29 Oct 2008

See all articles by Nathaniel J. Smith

Nathaniel J. Smith

University of Edinburgh

Shweta Narayan

University of California, Berkeley

Date Written: October, 28 2008


Humans in interaction produce diverse movements. As analysts, we focus on 'speech' and 'gesture'; we subdivide 'gesture' further into such categories as spatial 'iconics' and prosodic 'beats'. We do not know how these categories relate to one another; and even the speech/gesture distinction can be unclear in signed languages.

We can gain perspective by considering these in relation to other, relatively unstudied, movements: self-adjustments, weight shifts, etc. We present initial work on one such class: rhythmic motions produced by (usually) the legs and feet. These motions are common; we analyzed videos of five such "jigglers", covering three different levels of discourse context (alone, one-sided conversation, interactive conversation) and two languages/cultures (American English and Tzoltzil).

The form of a jiggle is determined by four parameters: the moving joint, the time of onset, the periodicity, and the time of offset. The joint seems primarily determined by postural affordances - crossed legs lead to different motion than flat feet, etc. The other variables, however, are surprisingly sensitive to linguistic and gestural context; we have observed two fundamental patterns of systematicity.

The first is more familiar: 'pulses' of rhythmic motion occur aligned with other rhythmic activity in the environment. This activity might include the jiggler's speech prosody, the strokes of their gestures, the prosody and/or strokes of their interlocuter's speech and gesture, etc. This category can be further subdivided into true rhythmic entrainment, consisting of a multi-beat rhythm in time with some isochronous pulse, and single pulse alignment, in which only a single motion is made.

The second pattern is more interesting: leg motion occurs in alternation with some other activity in the environment. A wide range of activities may fill this role. For instance, one speaker talks, pauses, and then resumes talking, and - reminiscent of a verbal filled pause - during the pause kicks their leg several times, with the onset and offset of the motion aligned with the offset and onset of their speech. (Later, they produce the same phrase without a pause, but a kick occurs at the same location in a kind of "fidget catchment".) Another speaker's foot moves freely while they are speaking, but freezes while they produce gestural strokes. Later, this same individual is listening to their interlocuter speak, and their foot moves - except that it freezes whenever their interlocutor produces a gestural stroke. A final individual began jiggling when they felt their interlocutor was failing to provide new information, and stopped when a topic shift was signaled. This may explain the folk idea that fidgeting indicates boredom. While interlocuters may find it a useful signal, it is clear from our analysis that this behavior is an instance of a deeper and more systematic pattern.

Thus, while so far we can only speculate about the causal and functional properties of jiggles, they are clearly substantially rule-governed, sensitive to both formal and semantic aspects of ongoing discourse, widespread - both patterns were observed in all discourse contexts, and both languages - and deserving of further attention.

Suggested Citation

Smith, Nathaniel J. and Narayan, Shweta, Fidgeting is Not Random: Rhythmic Leg Motion, Speech, and Gesture (October, 28 2008). 9th Conference on Conceptual Structure, Discourse, & Language (CSDL9), Available at SSRN:

Nathaniel J. Smith (Contact Author)

University of Edinburgh ( email )

Old College
South Bridge
Edinburgh, Scotland EH8 9JY
United Kingdom

Shweta Narayan

University of California, Berkeley ( email )

310 Barrows Hall
Berkeley, CA 94720
United States

Do you have a job opening that you would like to promote on SSRN?

Paper statistics

Abstract Views
PlumX Metrics