Further evidence for embodied cognition: The link between spontaneous respiratory oscillations and number processing

Abstract

Embodied accounts of cognition propose that bodily states shape mental representations. We test this proposal in numerical cognition by asking whether spontaneous respiration biases random number generation (RNG). Building on prior work by Belli and collaborators (2021), we adopt a self-paced handwriting RNG paradigm that preserves the continuous, cyclic nature of breathing. Healthy adults will write digits 1–9 on a tablet while abdominal respiration is recorded via a piezoelectric belt. For each trial, we will extract the breathing phase (inhalation/exhalation) and the abdominal expansion at that moment. We preregister three hypotheses: (Hp1) inhalation will bias production toward larger numbers relative to exhalation; (Hp2) greater abdominal expansion will predict larger numbers independent of phase; and (Hp3) inhalation and larger expansion will be associated with increased written-digit size. The predefined sample size (N = 30) will provide 99.9% power to detect effects on the order reported by prior work. By coupling fine-grained respiratory dynamics with self-paced behavior, the study offers a stringent test of whether interoceptive and somatosensory signals couple with abstract magnitude processing. Confirmatory evidence would provide convergent, method-robust support for embodied influences on numerical cognition and will reveal a parallel impact on action, strengthening brain–body accounts of magnitude representation.