从基本的感知觉到复杂的社会行为都属于人类的认知能力，这些认知能力通常被分为两阶层层级结构。元认知（metacognition）和心智化（mentalizing）属于两种高阶认知过程，它们都与元水平上的心理状态表征有关。高阶的认知能力通常涉及监控低阶的认知过程，例如元认知涉及监控自我决策过程，而心智化通常指监控他人的认知状态。目前对元认知和心智化进行直接比较的研究尚且不足。本文将通过三个研究并应用决策不确定性实验范式对元认知和心智化进行直接比较，从而研究二者在个体差异、神经表征和交互作用中的相关机理。
人类的认知能力都是先天与后天的复杂产物，确定影响认知能力个体差异的主要因素是一项重要的科学问题。行为遗传学的研究结果几乎都表明人类一阶认知能力主要受遗传因素影响。但目前元认知和心智化中所涉及的二阶认知能力的遗传环境作用尚不清楚，并且一阶和二阶认知能力个体差异的来源是否一致仍存在争议。在研究1中，我们利用经典双生子研究方法系统地比较了元认知和心智化任务中相关的一阶和二阶认知能力受遗传和环境的影响作用。结果表明，决策和判断中所涉及的一阶认知能力的个体差异主要由遗传因素贡献，而元认知和心智化中的所涉及的二阶认知能力则更偏向于受到共同环境因素的影响。这些结果提示我们高阶认知能力存在一定的可塑性，这可能依赖于大脑的皮层结构，尤其是前额叶皮层，在进化过程中对社会生活的适应性发展。因此，在接下来的研究中，我们尝试确定元认知和心智化在大脑中相互区分和交互作用的神经机制。
自我-他人区分（self-other distinction，SOD）是许多社会认知领域的一项关键机制，它通常指区分自我与他人的行为、感知、感觉和情绪表征的能力。虽然元认知和心智化被广泛的研究，但是很少有直接比较这两者对心理状态表征的自我-他人区分。此外用这种自我－他人二分法来描绘元认知和心智化也不充分，因为当表征自我曾经的心理状态时会出现歧义的情况。因此，研究2系统地研究当监控不同对象（当前自我，过去自我和他人）完成知觉决策任务时决策不确定性的不同水平（客体水平和元水平）的神经表征之间的差异。研究结果揭示了心智化过程中内部心理状态表征的多种形式。首先，所有心智化任务中的外部信息共同表征在右侧顶下小叶（inferior parietal lobe，IPL）。其次，归因于他人的客体水平和元水平心理状态在颞顶联合区（temporoparietal junction，TPJ）均有表征，但是背内侧前额叶皮层（dorsomedial prefrontal cortex，DMPFC）选择性地表征了归因于他人的决策不确定性的元水平心理状态。此外，当归因于过去自我的时候，决策不确定性的客体水平和元水平的心理状态分别表征在楔前叶（precuneus）和外侧额极叶皮层（lateral frontopolar cortex, lFPC）。另一方面，背侧前扣带回皮层（dorsal anterior cingulate cortex, dACC）表征元认知中当前经历到的决策不确定性并且监控不同心智化的评估的不确定性，但未发现对心智化中的评估本身的表征。
相对于自我-他人区分，在很多情况下人们也常常将自我和他人的思想混合在一起，即自我-他人融合（self-other mergence，SOM）。先前的研究发现心智化中SOM的发生依赖于社会情境，且选择性地发生在亲社会行为中。研究3则发现当被试同时经历相似的认知状态时，SOM可以在不同的社会情境中稳定地发生。这种独立于社会情境的SOM效应是通过从自我到他人的单向社会投射来实现的，但相反的方向SOM效应却并未发生。颞顶联合区（TPJ）的神经活动通过与表征自我认知状态的背侧前扣带回皮层（dACC）的有效功能连接选择性地追踪SOM效应。背侧前额叶皮层（DMPFC）则通过表征他人独特的认知状态选择性地参与SOD。此外，重复经颅磁刺激对TPJ神经活动的虚拟损伤可以因果性地减弱SOM效应。最后，SOM效应量大小受到人们对一般心智化能力的信念所调节。当一个人越觉得心智化很难实现时，就越可能利用社会投射来推断他人的认知状态。
综上所述，本文利用决策不确定性实验范式分别研究了元认知和心智化个体差异的主要来源，相互区分的神经表征以及交互作用的神经机制。研究结果表明，首先，共同环境中的经历可能是塑造元认知和心智化中的认知能力的主要原因。其次，元认知和心智化过程的神经特征存在明显的SOD效应，dmPFC唯一地表征了归因于他人的决策不确定性的元水平心理状态，dACC表征元认知中自我当前经历的决策不确定性并且监控不同心智化的评估的不确定性。最后，当自己同时唤起类似的认知状态时，个体可以通过TPJ和dACC的有效功能连接采用社会投射的方式产生SOM以促进心智化的实现。

Cognitive abilities in humans, ranging from basic perception to complex social behaviors, are typically organized into a two-order hierarchical structure. Metacognition and mentalizing are two high-order cognitive processes, both of which are related to meta-level representations of mental states. High-order cognitive abilities involve monitoring lower-level cognitive processes. For instance, metacognition entails monitoring one's own decision-making processes, while mentalizing typically refers to monitoring the cognitive states of others. However, there is currently limited research that directly compares metacognition and mentalizing. The three studies in this paper will employ a relatively consistent decision uncertainty experimental paradigm to directly compare metacognition and mentalizing. This will allow us to investigate the underlying mechanisms of both in terms of individual differences, neural representations, and interactions.
Human cognitive abilities are the complex result of both nature and nurture, and identifying the primary factors influencing individual differences in cognitive abilities is an important scientific question. The results of behavioral genetics studies consistently suggest that human’s first-order cognitive abilities are primarily influenced by genetic factors. However, the genetic and environmental effects on second-order cognitive abilities involved in metacognition and mentalizing are not yet clear, and there is still debate over whether the origins of individual differences in first-order and second-order cognitive abilities are consistent. Study 1 employed a classical twin study method to systematically compare the genetic and environmental influences on first-order and second-order cognitive abilities related to metacognition and mentalizing tasks. The results indicated that individual differences in first-order cognitive abilities involved in decision-making and judgment were primarily influenced by genetic factors, while second-order cognitive abilities associated with metacognition and mentalizing were more inclined to be influenced by shared environmental factors. These findings suggest that there is a certain degree of plasticity in high-order cognitive abilities, which might depend on the cortical structure of the brain, particularly the prefrontal cortex, which has evolved to adapt to social complexities. Therefore, in subsequent research, we aim to identify the neural mechanisms that distinguish and interact between metacognition and mentalizing in the brain.
Self-other distinction (SOD) is a key mechanism in many domains of social cognition, typically referring to the ability to differentiate between one's own behaviors, perceptions, sensations, and emotional representations and those of others. While metacognition and mentalizing have been extensively studied, there is limited research directly comparing the self-other distinction in the representation of mental states between these two processes. Furthermore, using this self-other dichotomy to characterize metacognition and mentalizing is not sufficient, as it can lead to ambiguity when representing one's own past mental states. Therefore, Study 2 systematically investigates the differences in neural representations between different levels of decision uncertainty (object-level and meta-level) when monitoring different targets (current self, past self, and others) while they engage in perceptual decision tasks. The research results reveal various forms of internal mental state representations during mentalizing. First, all external information was commonly represented in the right inferior parietal lobe (IPL) across the mentalizing tasks. Second, it’s dorsomedial prefrontal cortex (dmPFC), not the temporoparietal junction (TPJ), that uniquely represented the meta-level mental states of decision uncertainty attributed to others. TPJ also equivalently represented the object-level mental states of decision inaccuracy attributed to others. Furthermore, the object-level and meta-level mental states of decision uncertainty, when attributed to the past self, were respectively represented in the precuneus and the lateral frontopolar cortex (lFPC). In contrast, the dorsal anterior cingulate cortex (dACC) consistently represented both decision uncertainty in metacognition and estimate uncertainty during monitoring the different mentalizing processes, but not the inferred decision uncertainty in mentalizing.
In contrast to SOD, in many cases, people also often blend their own thoughts with those of others, known as self-other mergence (SOM). Occurrences of SOM in mentalizing have been previously found to be critically dependent on social contexts, selectively in prosocial behaviors. But Study 3 found that SOM stably occurred in different social contexts, whenever the subjects concurrently experienced their own similar cognitive states. This social context-independent SOM effect was sustained by unidirectional social projection from the self to others, but not along the opposing direction. The TPJ neural activities selectively tracked the SOM effect through an effective functional connectivity with the dorsal anterior cingulate cortex, which represented the self’s cognitive states, while the dmPFC selectively involved in SOD by representing others’ unique cognitive states. Further, disruptions of the TPJ activities by repetitive transcranial magnetic stimulation causally attenuated the SOM effect. Finally, the magnitude of SOM effects is regulated by people's beliefs about general mentalizing abilities. The more a person finds it difficult to achieve mentalizing, the more likely they are to use social projection to infer the cognitive state of others.
In summary, this paper utilized a decision uncertainty experimental paradigm to investigate the origins of individual differences in metacognition and mentalizing, the neural representations that distinguish them, and the neural mechanisms underlying their interactions. The research results indicate that, firstly, shared environmental experiences may be a primary factor shaping cognitive abilities in both metacognition and mentalizing. Secondly, there are distinct SOD effects in the neural characteristics of metacognition and mentalizing. The dmPFC uniquely represents meta-level mental states attributed to others' decision uncertainty, while the dACC represents metacognition of one's current experienced decision uncertainty and monitors the uncertainty of different mentalizing assessments. Finally, when individuals concurrently evoke similar cognitive states, they can generate SOM through effective functional connectivity between TPJ and dACC, employing a mechanism of social projection to facilitate the realization of mentalizing.