Glass C, Bradshaw J L, Day R H, Umiltà C
Cortex. 1985 Dec;21(4):513-31. doi: 10.1016/s0010-9452(58)80002-7.
Laterally presented portraits were judged as familiar or unfamiliar (Experiments 1 and 2), or were processed in a target matching task (Experiment 3). Exposure durations (100 or 190 msec.) were varied between (Experiments 1 and 2) or within (Experiment 3) the experiments. Clear and degraded faces appeared either randomly intermingled (Experiment 1) or in blocked sequences (Experiments 2 and 3). In Experiment 3 familiar and unfamiliar stimuli also appeared in blocks. Stimulus degradation was achieved by coarse quantization of the image into blocks, so replacing relevant high-spatial-frequencies (the features) by spurious information, preserving only the original lower-spatial-frequency components. Left hemisphere mediation was strongest and most consistent with long exposures and random sequences of clear and degraded stimuli. Right hemisphere mediation tended to appear with shorter exposures and degraded stimuli presented in blocks. Though the interactions were often complex, the general pattern of results was consistent with the analytic-holistic processing dichotomy.
侧向呈现的人像被判断为熟悉或不熟悉(实验1和2),或者在目标匹配任务中进行处理(实验3)。曝光持续时间(100或190毫秒)在实验之间(实验1和2)或实验内部(实验3)有所不同。清晰和模糊的面孔要么随机混合出现(实验1),要么按块序列出现(实验2和3)。在实验3中,熟悉和不熟悉的刺激也按块出现。通过将图像粗略量化为块来实现刺激退化,用虚假信息替换相关的高空间频率(特征),只保留原始的低空间频率成分。左半球介导在长时间曝光以及清晰和退化刺激的随机序列时最强且最一致。右半球介导倾向于在较短曝光以及按块呈现的退化刺激时出现。尽管相互作用通常很复杂,但结果的总体模式与分析-整体加工二分法一致。