Metacognition and Mindreading: One system or two?

直觉上，我们意识到自己的想法的方式与我们意识到别人的想法的方式大不相同；我们每个人都有直接了解自己想法的特权，而我们必须根据他人的公开行为来“推断”他们的想法。监控自己思想的能力被称为“元认知”，而监控他人思想的能力被称为“读心术”。上述元认知的直观观点导致一些（所谓的“两个系统”理论家）声称，不同的心理机制支撑着元认知和读心术，并且元认知在进化和发展上先于读心术。这一说法受到另一组（“单一系统”）理论家的质疑，他们认为读心术和元认知实际上依赖于完全相同的核心心理机制（尽管有共同的直觉），而元认知实际上是读心术进化需要的副产品。双系统理论家使用两个潜在的有力论据来支持他们的观点。首先，他们认为非人类灵长类动物能够在元认知测试中表现良好，尽管这些灵长类动物几乎没有显示出读心术的能力。其次，他们认为患有自闭症谱系障碍（ASD）的人尽管读心术受损，但却表现出完整的元认知能力。如果这些论点得到证实，那么这将排除读心术和元认知之间关系的单一系统观点。然而，单一系统理论家对这些观点提出了质疑，他们声称(a)非人类灵长类动物擅长的所谓元认知任务实际上并不需要元认知意识，(b) ASD患者确实表现出元认知障碍。然而，很少有实证研究在这些论点和反论点之间做出决定。因此，我们提出的研究目的是通过在两组实验中采用多种元认知任务，以系统的方式测试这些理论。在第一组实验中，我们将以系统的方式操纵元认知任务的要求，以确定用于测试非人类灵长类动物元认知的任务类型是否真的需要意识到自己的想法。神经正常的成年人将参与这组研究。在第二组实验中，我们将评估ASD成人和在年龄和一般智力方面与ASD参与者非常相似的神经型成人的元认知监测。我们将使用各种各样的任务，包括那些用于测试非人类灵长类动物元认知的任务。总之，这些实验将使我们能够在元认知和读心术之间的关系的相互竞争的理论之间做出决定。除了支持理论发展之外，我们提出的研究对于理解ASD患者认知功能的本质也很重要。令人惊讶的是，尽管有数百项研究探索了这种疾病的读心术，但对ASD中元认知的研究却很少。事实上，这种“自我”意识的困难也许可以解释自闭症谱系障碍的行为特征，而读心术的困难很难解释（例如，行为不灵活）。因此，我们希望我们提出的研究最终会对自闭症患者有实际的好处，通过指导教学实践和干预措施来提高自闭症患者的学习和认知行为功能。

Intuitively, we are aware of our own thoughts in a way that is quite different to the way that we are aware of other people's thoughts; each of us is privileged in knowing only our own thoughts directly, whereas we have to *infer* others' thoughts on the basis of their overt behaviour. The ability to monitor one's own thoughts is termed "metacognition", whereas the ability to monitor others' thoughts is termed "mindreading". The intuitive view of metacognition described above has led some (so-called "two systems" theorists) to claim that distinct psychological mechanisms underpin metacognition and mindreading, and that metacognition is evolutionarily and developmentally prior to mindreading. This claim is disputed by another group of ("one-system") theorists who argue that mindreading and metacognition actually rely on exactly the same core psychological mechanism (despite common intuition), and that metacognition actually emerged as a by-product of the evolutionary need to mind-read others. Two-systems theorists use two potentially powerful arguments to support their view. First, they argue that non-human primates are capable of performing well on tests of metacognition even though such primates show little evidence of mindreading ability. Second, they argue that humans with autism spectrum disorder (ASD) show intact metacognitive ability despite impaired mindreading. If these arguments are validated, then this would rule out the one-system view of the relation between mindreading and metacognition. However, one-system theorists have challenged these arguments, claiming that (a) the supposedly metacognitive tasks that non-human primates perform well on do not, in fact, require metacognitive awareness, and (b) people with ASD *do* show metacognitive impairments. There is, however, very little empirical research to decide between these arguments and counter-arguments. Therefore, the aim of our proposed research is to test these theories in a systematic fashion by employing multiple kinds of metacognitive task across two sets of experiments.In the first set of experiments, we will manipulate the demands of metacognitive tasks in systematic ways to establish whether the kinds of task used to test metacognition in non-human primates really do require awareness of one's own thoughts. Neurotypical adults will take part in this set of studies.In the second set of experiments, we will assess metacognitive monitoring among adults with ASD and neurotypical adults who are closely similar to participants with ASD in terms of age and general intelligence. We will use a variety of tasks, including those kinds that are used to test metacognition in non-human primates. Together, these experiments will allow us to decide between the competing theories of the relation between metacognition and mindreading. In addition to supporting theory development, our proposed research will be important in understanding the nature of cognitive functioning in people with ASD. It is striking how little research there is into metacognition in ASD, given that many hundreds of studies that have explored mindreading in this disorder. Indeed, difficulties with such *self*-awareness might explain aspects of the behavioural features of ASD that difficulties with mindreading do not so easily explain (e.g., behavioural inflexibility). Thus, we hope that our proposed research will, in the end, have practical benefits for people with ASD, by informing teaching practices and intervention efforts designed to enhance learning and cognitive-behavioural functioning among people with this disorder.