Scholarlink

<p>Human newborns preferentially attend to self-propelled objects, and three-month-olds expect them to act in a goal-directed manner. Dogs also prefer animate objects, and anticipate human actions to be goal-driven, but it is unclear whether perceiving an object as animate leads to the expectation that it has goals. Here, first dogs observed either (1) a human moving around the room, (2) a robot interacting with a human, (3) a robot displaying self-propelled motion, or (4) a robot moving ambiguously. Next, the human or the robot repeatedly approached one of two objects. Finally, the objects’ places were switched, and the actor approached the same-object or same-location as before. We measured dogs’ look toward the objects after switching their locations. Dogs looked at the goal object before the actor turned toward either object, and looked longer at the goal object when the actor moved in a route-consistent, instead of a goal-consistent manner. These suggest that irrespectively of the actor type, dogs anticipated it to approach the goal-object, and the actor violated this expectation when moving to the same location. Thus, dogs’ goal attribution is not uniquely tied to human action, but it can be elicited by broader features of animacy.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Perception of animacy leads to expectation of goal-directed behaviour in dogs

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年1月13日

查看全文
  • Zsuzsanna Gedai,
  • Ádám Miklósi,
  • Judit Abdai

摘要

Human newborns preferentially attend to self-propelled objects, and three-month-olds expect them to act in a goal-directed manner. Dogs also prefer animate objects, and anticipate human actions to be goal-driven, but it is unclear whether perceiving an object as animate leads to the expectation that it has goals. Here, first dogs observed either (1) a human moving around the room, (2) a robot interacting with a human, (3) a robot displaying self-propelled motion, or (4) a robot moving ambiguously. Next, the human or the robot repeatedly approached one of two objects. Finally, the objects’ places were switched, and the actor approached the same-object or same-location as before. We measured dogs’ look toward the objects after switching their locations. Dogs looked at the goal object before the actor turned toward either object, and looked longer at the goal object when the actor moved in a route-consistent, instead of a goal-consistent manner. These suggest that irrespectively of the actor type, dogs anticipated it to approach the goal-object, and the actor violated this expectation when moving to the same location. Thus, dogs’ goal attribution is not uniquely tied to human action, but it can be elicited by broader features of animacy.