| 其他摘要 | The specific knowledge of manipulating objects, i.e., action representation of manipulable objects plays an important role in the process of object recognition. According to whether the aim of manipulation is to grasp or use the object, the action representations could be divided into structural action representation and functional action representation. Many empirical studies showed that the activation of structural and functional action representations is separated and they play different roles in the process of object recognition based on behavioral and physiological evidences. Recently, many studies have found that there is an interaction between structural action representation and functional action representation during object manipulation, but the mechanism beneath the interaction in the same action sequence remains unclear. Whether the structural (or functional) action representation activated in advance will facilitate (or inhibit) the subsequent using (or grasping) action, and whether the two action representations are equal during object recognition, including whether the action representations could migrate different objects, these questions need to be further explored.
This thesis systematically examines these questions through two studies. Study 1 included the evaluation of experimental materials and two behavior experiments using the variation of priming paradigm to explore the influence between the two action representations; At the same time, by strictly controlling the relationship between different objects and the same representation, the cross-object interaction between structural action representation and functional action representation was investigated. In study 1, we found that pre activated action representations could promote the subsequent grasping or using actions of the same object, and similar priming also exists in different objects with the same action representation. In Study 2, based on Study 1, motion capture technology was used to verify the stability of the results of Study 1 in multiple variables and dimensions. Also, by canceling the guidance of delayed response, subjects are required to make instant manipulation actions on the target object in order to explore the interaction between two action representations and the difference between object recognition stage and action execution stage. The results showed that the interaction between two action representations mainly occurred in the object recognition stage. The peak velocity, peak grip aperture, movement time and other kinematic data in the movement stage had no significant difference under different switching conditions.
In this paper, we found that structural and functional action representations can promote each other in the process of object recognition, and this promotion can occur not only within the same object, but also between different objects with the same action representation. This kind of promotion manifests in the object recognition processing stage, but does not appear in the action planning and execution stage. This paper argues that structural action representation and functional action representation have similar functions in object recognition and processing. They not only compete, but also promote. Furthermore, structural action representation can also promote grasping action performance, which suggests that object recognition and manipulation are not dominated by a single representation, but by two different but interrelated knowledge representations with similar characteristics. This representation of knowledge can be shared by different objects, rather than unique information based on object identity. |
修改评论