Other peoples' (imagined) visual perspectives are represented perceptually in a similar way to our own, and can drive bottom-up processes in the same way as own perceptual input (Ward, Ganis, & Bach, 2019). Here we test directly whether visual perspective taking is driven by where another person is looking, or whether these perceptual simulations represent their position in space more generally. Across two experiments, we asked participants to identify whether alphanumeric characters, presented at one of eight possible orientations away from upright, were presented normally, or in their mirror-inverted form (e.g. "R" vs. "Я"). In some scenes, a person would appear sitting to the left or the right of the participant. We manipulated either between-trials (Experiment 1) or between-subjects (Experiment 2), the gaze-direction of the inserted person, such that they either (1) looked towards the to-be-judged item, (2) averted their gaze away from the participant, or (3) gazed out towards the participant (Exp. 2 only). In the absence of another person, we replicated the well-established mental rotation effect, where recognition of items becomes slower the more items are oriented away from upright (e.g. Shepard and Meltzer, 1971). Crucially, in both experiments and in all conditions, this response pattern changed when another person was inserted into the scene. People spontaneously took the perspective of the other person and made faster judgements about the presented items in their presence if the characters were oriented towards upright to them. The gaze direction of this other person did not influence these effects. We propose that visual perspective taking is therefore a general spatial-navigational ability, allowing us to calculate more easily how a scene would (in principle) look from another position in space, and that such calculations reflect the spatial location of another person, but not their gaze.

中文翻译：

---

将视角视为虚拟导航？对他人所见事物的感知模拟反映了他们在空间中的位置，而不是他们的凝视。

他人（想象的）视觉视角以与我们自己相似的方式在知觉上表示，并且可以以与自己的知觉输入相同的方式推动自下而上的过程（Ward，Ganis和Bach，2019年）。在这里，我们直接测试视觉角度拍摄是否由其他人的注视驱动，或者这些知觉模拟是否更一般地表示他们在空间中的位置。在两个实验中，我们要求参与者确定字母数字字符是否以垂直放置的八个可能方向之一呈现，还是以其镜面倒置的形式呈现（例如“ R”对“Я”）。在某些场景中，一个人会坐在参与者的左侧或右侧。我们操纵试验间（实验1）或受试者间（实验2），插入者的注视方向，以使他们要么（1）朝着要判断的物品看，（2）将目光从参与者身上移开，要么（3）向参与者注视（实验2只要）。在没有其他人的情况下，我们复制了公认的心理旋转效应，即物品的识别速度越慢，则物品越远离直立放置（eg Shepard and Meltzer，1971）。至关重要的是，无论是在实验中还是在所有条件下，当另一个人插入场景时，这种响应方式都会改变。人们会自发地从另一个人的角度出发，如果角色是直立在他们面前的，他们会更快地判断他们面前的物品。另一个人的视线方向不会影响这些效果。