Objects processing in perception and action: From location and action coding to affordances

对象是人类与环境联系、修改和创造环境的手段。单个和成对物体的属性为我们的感知和运动系统提供线索，影响我们的表现效率。单个物体的空间方向就是这样的线索之一：当它们对应于物体方向时，左/右手的运动反应通常更快更准确（基于物体的对应效应- ce）。基于这种空间取向的来源，对这种效应提出了两种不同的解释。位置编码说明，物体的不对称性，由突出的部分产生，如手柄或目标导向的侧面（如左或右），对观察者来说变得突出，并自动激活相应手的一般运动反应。相反，功能激活帐户声明对象激活与手柄对齐的手的适当操作动作，无论视觉上是否显著。最近，由于位置编码不足以解释观察到的效果，我专门为配对对象获得的ce制定了一个动作编码帐户。因此，对于一致的对象对（例如，壶+杯）形成了一个动作方向代码，但是当对象对没有任何关系（壶+球）时就没有了。因此，ce只在前一种情况下产生。我的研究通过支持位置编码帐户的证据，有助于减少功能激活的影响。然而，究竟是哪个过程控制着我们与对象的交互，仍然是一个有争议的问题。此外，我的新动作编码方案需要进一步的研究。我的项目预计在三个WPs中并行测试ce的当前账户及其潜在过程：WP1，在两个行为实验中，处理准确选择的对象图片之间的动作关系的影响，有和没有语义关系的贡献。WP2通过一个功能磁共振实验，研究了大脑处理物体之间的动作和语义联系的方式，以及它们的激活模式在多大程度上重合或不匹配。为此，结合了创新的计算方法（多变量模式、预测模式分解分析）。在WP3中，进行了五项行为实验，包括：真实物体及其图片（单个和成对），远端和近端反应（按按钮和抓取），有物理障碍和无物理障碍时的伸手和抓取反应。在这里，我将评估日益现实的环境对与对象交互的影响，并测试位置-动作编码为提供过程提供方式的可能性。总而言之，我的项目旨在找到详尽的答案：在证据相互矛盾的情况下，功能激活账户是否仍然有效；位置-动作编码说很可能占上风；这两种进程在某种程度上是共存的，必须建立一条分界线。

Objects are means by which humans relate to, modify and create their environment. Properties of single and paired objects provide cues to our perceptual and motor system that affect the efficiency of our performance. Spatial orientation of single objects is one of such cues: motor responses of left/right hands are typically faster and more accurate when they correspond to objects orientations (object-based Correspondence Effects-CEs). Two alternative accounts are proposed for this effect, based on the source of such spatial orientation. The location coding account states that object asymmetries, generated by protruding portions like handle or goal-directed sides (e.g. left or right), become salient to the observer and automatically activate generic motor responses of the corresponding hand. The affordance activation account instead states that objects activate appropriate manipulative actions of the hand aligned with the handle, whether visually salient or not. Recently, I formulated an action coding account specifically for CEs obtained with paired objects, since the location coding was insufficient to explain the observed effects. Thus, an action-direction code is formed for consistent object pairs (e.g. jug+cup), but is missing when object pairs share no relation at all (jug+ball). Consequently, CEs are only produced in the former case. My studies contributed to reduce the impact of the affordance activation through evidence supporting the location coding account. However, which process precisely governs our interaction with objects is still a matter of debate. In addition, my new action coding proposal calls for further investigation. My project foresees to test in parallel the current accounts of CEs and their underlying processes in three WPs: WP1, in two behavioral experiments, tackles the effects of action relations between accurately selected objects’ pictures, with and without the contribution of semantic relations. WP2, with one fMRI experiment, investigates the way action and semantic links between objects are processed in the brain and to what extent their activation patterns coincide or mismatch. To do so, innovative computational approaches (multivariate pattern, predictive pattern decomposition analyses) are combined. In WP3, five behavioral experiments are run including: real objects and their pictures (single and paired), distal and proximal responses (button press and grasp), reach-and-grasp responses with and without physical obstacles. Here, I will assess the effect of increasingly realistic contexts on the interaction with objects, and test the possibility that position-action coding gives the way to the affordance process. In sum, my project aims at finding exhaustive answers as to whether: the affordance activation account is still valid, net of conflicting evidence; the location-action coding account is likely to prevail; the processes coexist to some extent and a dividing line must be established.