Is mentalizing for social evaluation? Studies with human infants.

典型的发育中的人类会敏锐地调整到精神状态，很容易思考别人的意图、了解和信仰。事实上，心智化对人类物种的关键特征至关重要，包括有效的社会学习和维持长期的合作关系。鉴于心智化在人类生活中的普遍性和适应性，认知科学一直致力于阐明其潜在性质和发展起源。在这个研究项目中，我们将通过对学龄前婴儿的三个研究流来检验一个关于人类心智化系统的特定假设，这些研究流源于过去nserc支持的研究，这些研究表明，婴儿可能特别愿意和/或能够在社会评价与非评价的环境中进行心智化（例如，行动者的心理状态是亲社会的或反社会的，而不是中立的或非社会的），以及最近在发展心智化文献中复制关键结果的臭名昭著的失败。具体来说，我们假设心理化的发展首先是为了促进有效的社会评价，允许高度依赖的人类通过容易地区分好与坏的合作伙伴来有效地合作。从这个基本功能出发，人类开始运用他们的心智化能力来推理拥有各种心理状态的各种行为体。本研究通过3个研究流，使用包括婴儿注视时间、预测注视、脑电图/ERP和近红外光谱（fNIRS）在内的多方法方法，检验了心智化是为了社会评价这一总体假设下的4个子假设。每个流将探索婴儿是否以及在多大程度上更有可能在社会评价环境中进行心理化，通过仔细控制的实验，比较与社会评价相关的环境或多或少。具体来说，我们将探讨(1)关于社会心理状态的推理是否比关于生理心理状态的推理更有优势；(2)对有价社会心理状态的推理优于对中性社会心理状态的推理，尤其是对消极社会心理状态的推理；(3)如果一个行为人之前被认定拥有有价值的社会心理状态，尤其是消极的，那么这个行为人将来就特别有可能参与心理化系统，即使是在中立或非社会行为的情况下。建议使用多种不同（但互补）的方法将为各种高素质人员提供尖端研究方法的培训，并最大限度地提高我们为假设获得关键收敛证据的机会。总之，这些研究将有助于更好地阐明人类心理化系统的潜在本质，使我们对我们如何以及为什么进行心理化有了基本的了解。

Typically developing humans are exquisitely tuned into mental states, readily thinking about what others intend, know, and believe. Indeed, mentalizing is imperative for key features of the human species including effective social learning and maintaining long-term cooperative relationships. Given the ubiquity and adaptiveness of mentalizing in human life, significant effort in the cognitive sciences has been dedicated to elucidating its underlying nature and developmental origins. In this research program, we will examine a particular hypothesis about the human mentalizing system via 3 streams of studies with preverbal infants, which stems both from past NSERC-supported work suggestive that infants may be particularly willing and/or able to mentalize within socially evaluative versus non-evaluative contexts [for instance, contexts where agents' mental states are prosocial or antisocial versus neutral or non-social], as well as from recent notorious failures to replicate key results within the developmental mentalizing literature. Specifically, we hypothesize that mentalizing develops first and foremost to facilitate effective social evaluation, allowing highly interdependent humans to cooperate effectively by readily distinguishing good from bad cooperative partners. From this primary function, humans come to apply their mentalizing capacities to reasoning about all kinds of agents in possession of all kinds of mental states. The proposed research examines 4 sub-hypotheses under the overarching hypothesis that mentalizing is for social evaluation via 3 research streams, using a multi-method approach including infant looking time, predictive looking, EEG/ERP, and fNIRS. Each stream will explore whether and to what extent infants are more likely to mentalize in socially evaluative contexts, through carefully controlled experiments comparing contexts more versus less relevant to social evaluation. Specifically, we will explore (1) if reasoning about social mental states is privileged relative to reasoning about physical mental states; (2) if reasoning about valenced social mental states is privileged relative to reasoning about neutral social mental states, especially negative ones; and (3) if once an agent has previously been identified as possessing valenced social mental states, especially negative ones, that agent will be particularly likely to engage mentalizing systems in the future, even while acting neutrally or non-socially. The proposed use of multiple distinct (but complementary) methods will provide training in cutting-edge research methods to a variety of Highly Qualified Personnel, and maximize our chances of acquiring crucial convergent evidence for our hypotheses. Together, these studies will serve to better elucidate the underlying nature of the human mentalizing system, informing our basic understanding of how and why we mentalize.