CONSTRAINTS ON UNDERSTANDING PHYSICAL DYNAMICS

点击翻译按钮获取中文摘要

DESCRIPTION (Applicant's Abstract): The perception and representation of object motion is of primary importance to animals generally. It appears to be the case, however, that the types of perceptual processes that serve motion analysis and ultimately motion understanding are not invariant over the different types of motions that rigid bodies execute. There appears to be a fundamental distinction between motion transformations that represent net displacements in space and rotations which are inherently cyclic. It is suggested that perceptual competencies in motion analysis reflect the underlying ecological importance of displacement and not the formal complexity of optic flow or the logic by which neural architecture renders early motion analysis. Studies are proposed to investigate the three main optic flow patterns that objects present as proximal stimulation; translations, expansions/contractions, and rotations. Assessments will be made in the areas of attention, recognition memory, implicit memory, and perceptual organization. It is proposed that the first two types of flow fields will reveal common processing protocols by virtue of their common distal interpretation as translational displacement in three dimensional space. We propose that the protocols governing rotation analysis will be distinguishably different and demonstrably inferior in terms of the quality and amount of information that is processed. Four sets of experiments are proposed to test and refine this theoretical framework. (1) Allocation of attention will be assessed using several search paradigms and by analyzing the nature of speed-accuracy trade-off. (2) Perceptual organization will be assessed in contexts in which hierarchical frame of reference effects can occur. (3) Recognition memory for direction will be assessed in standard acquisition/test phase methodologies. (4) Implicit memory will be assessed in both indirect and direct priming paradigms.