Learning to Attenuate Myself: A Predictive Processing Account of Body-Scan Meditation and the Dissolution of Bodily Boundaries

Meditation practices often involve sustaining attention on the body. Typically, attention is understood to enhance both neural resource allocation and the subjective salience of the attended target. However, in deep meditative states, practitioners sometimes report a dissolution of bodily boundaries, a phenomenon known in Pali as bhaṅga. This presents a paradox: why does focused attention, which typically heightens sensory perception, instead lead to its dissolution? This paper addresses this apparent contradiction by integrating computational, phenomenological, and empirical perspectives on attention, interoception and meditation. We focus on the body-scan technique, as practiced in Theravada Buddhist traditions, and its powerful capacity to produce experiences of the dissolution of bodily boundaries. Working within the Predictive Processing framework, we propose that this "dissolution" of bodily boundaries results from the body-scan's impact on attentional processes. We argue that by enhancing the precision of low-level predictions related to somatosensory signals, the body-scan effectively attenuates these signals, thereby diminishing perception of the body’s boundaries. In support of this claim, we first describe the body-scan technique and its phenomenological outcomes. We then introduce key concepts from the predictive processing framework and provide a detailed analysis of attentional processes during the body-scan. We conclude that the attenuation of somatosensory signals during the body-scan not only contributes to the experience of bhaṅga but also suggests a broader potential of this practice for enhancing well-being. With appropriate therapeutic integration, this attentional modulation offers promising applications in addressing conditions characterized by disrupted self-regulation, such as addiction and emotional dysregulation.