Rad Dana, Costache-Colareza Corina, Paraschiv Ruxandra-Victoria, Gavrila-Ardelean Liviu
Centre of Research Development and Innovation in Psychology, Faculty of Educational Sciences, Aurel Vlaicu University of Arad, 310032 Arad, Romania.
Facultatea de Știintele Educației, Comunicare și Relații Internaționale, Universitatea Titu Maiorescu, 040051 Bucharest, Romania.
Brain Sci. 2025 Aug 23;15(9):909. doi: 10.3390/brainsci15090909.
The scientific study of emotion remains fraught with conceptual ambiguity, methodological limitations, and epistemological blind spots. This theoretical paper argues that existing paradigms frequently capture synthetic rather than natural emotional states-those shaped by social expectations, cognitive scripting, and performance under observation. We propose a conceptual framework that distinguishes natural emotion-spontaneous, embodied, and interoceptively grounded-from synthetic forms that are adaptive, context-driven, and often unconsciously rehearsed. These reactions often involve emotional scripts rather than genuine, spontaneous affective experiences. Drawing on insights from affective neuroscience, psychological measurement, artificial intelligence, and neurodiversity, we examine how widely used tools such as EEG, polygraphy, and self-report instruments may capture emotional conformity rather than authenticity. We further explore how affective AI systems trained on socially filtered datasets risk replicating emotional performance rather than emotional truth. By recognizing neurodivergent expression as a potential site of emotional transparency, we challenge dominant models of emotional normalcy and propose a five-step agenda for reorienting emotion research toward authenticity, ecological validity, and inclusivity. This post-synthetic framework invites a redefinition of emotion that is conceptually rigorous, methodologically nuanced, and ethically inclusive of human affective diversity.
对情感的科学研究仍然充满概念上的模糊性、方法上的局限性和认识论上的盲点。这篇理论文章认为，现有的范式常常捕捉到的是合成而非自然的情感状态——那些由社会期望、认知脚本和观察下的表现所塑造的情感状态。我们提出了一个概念框架，将自然情感——自发的、具身的和基于内感受的——与适应性的、情境驱动的且常常是无意识排练的合成形式区分开来。这些反应通常涉及情感脚本而非真实、自发的情感体验。借鉴情感神经科学、心理测量、人工智能和神经多样性等方面的见解，我们研究了脑电图（EEG）、测谎仪和自我报告工具等广泛使用的工具如何可能捕捉到情感的一致性而非真实性。我们进一步探讨了在经过社会筛选的数据集上训练的情感人工智能系统如何有复制情感表现而非情感真相的风险。通过将神经差异表达视为情感透明度的潜在场所，我们挑战了主导的情感常态模型，并提出了一个五步议程，以使情感研究重新转向真实性、生态效度和包容性。这个后合成框架要求对情感进行重新定义，这种定义在概念上严谨、方法上细致入微且在伦理上包容人类情感多样性。
