一项新的国际研究主张，冥想不是大脑的「静息」状态，而是一种动态、能改变心智的状态；研究在靠近罗马的 Santacittārāma（泰国森林传承 Thai Forest Tradition）中，检验了 12 位资深佛教僧侣。研究人员在 Chieti-Pescara 的实验室使用脑磁图（magnetoencephalography, MEG），在僧侣练习 2 种经典技巧时，以高精度记录大脑电活动：Samatha（持续、狭窄的专注，常以呼吸为对象）与 Vipassana（开放、平等心地监测感受、想法与情绪）。该研究以神经振荡、讯号复杂度与「临界性（criticality）」来建构冥想的框架；临界性是从统计物理改写而来的神经科学概念，约 20 年来被应用，用以描述在秩序与混沌边界附近运作，并被认为可支持高效率的资讯处理。

在实验中，僧侣在 Samatha 与 Vipassana 之间交替进行，两次练习之间有短暂的休息期；团队以进阶讯号处理与机器学习方法分析 MEG 记录，推导出多个动力学指标。发表于 Neuroscience of Consciousness（2025）的结果指出，两种冥想形式相较于休息都提升了脑讯号复杂度，意味著是一种更丰富、资讯更密集的状态，而不只是单纯的平静。同时，研究也观察到某些与神经活动全局组织相关的参数出现广泛下降，显示较高的局部或讯号层级复杂度，可能会与大尺度协调性的变化并存。

一个关键的区分指标是「临界性偏离系数（criticality deviation coefficient）」，它能将 Samatha 与 Vipassana 区分开来，暗示即使两者都提高复杂度，仍会到达不同的动力学配置。作者解读 Vipassana 会让大脑更接近平衡的临界区域（稳定性加上弹性），而 Samatha 则更趋向稳定、紧密聚焦的模式；理论上，接近这种临界平衡应可支持更具回应性的功能，例如更好的任务切换与资讯储存。证据受限于样本数偏小（12 位僧侣）与族群特定（高度受训的出家修行者），但研究为冥想与振荡调节、复杂度提升、临界性改变，以及压力与与焦虑、忧郁相关症状降低的连结，提供了量化支持。

A new international study argues meditation is not a “resting” brain state but a dynamic, mind-altering one, and it examined this in 12 experienced Buddhist monks from the Thai Forest Tradition at Santacittārāma (near Rome). In a lab in Chieti-Pescara, researchers used magnetoencephalography (MEG) to record brain electrical activity with high precision while monks practiced 2 classic techniques: Samatha (sustained, narrow attention, often on breathing) and Vipassana (open, equanimous monitoring of sensations, thoughts, and emotions). The work frames meditation in terms of neural oscillations, signal complexity, and “criticality,” a neuroscience concept adapted from statistical physics that has been applied for about 20 years and describes operation near a boundary between order and chaos that is thought to support efficient information processing.

During the experiment the monks alternated between Samatha and Vipassana, with brief rest periods between sessions, and the team analyzed the MEG recordings with advanced signal-processing and machine-learning methods to derive multiple dynamical indicators. Results published in Neuroscience of Consciousness (2025) reported that both meditation forms increased brain-signal complexity relative to rest, implying a richer information-dense state rather than simple calming. At the same time, the study observed widespread reductions in some parameters tied to the global organization of neural activity, indicating that higher local or signal-level complexity can coincide with shifts in large-scale coordination.

A key differentiator was the “criticality deviation coefficient,” which separated Samatha from Vipassana, suggesting the 2 practices reach distinct dynamical configurations even when both raise complexity. The authors interpret Vipassana as moving the brain closer to a balanced critical regime (stability plus flexibility), whereas Samatha trends toward a more stable, tightly focused mode; in theory, proximity to this critical balance should support more responsive function such as improved task switching and information storage. The evidence is constrained by the small sample size (12 monks) and the specific population (highly trained monastics), but the study adds quantitative support for links between meditation, modulated oscillations, increased complexity, altered criticality, and reported reductions in stress and symptoms associated with anxiety and depression.