內發性注意力導向: 物體為基選擇機制的必要但非充要條件

Abstract

本研究探討注意力作用機制中的導向 (orienting) 與選擇(selection) 之間的關係。注意力導向分為外引式 (exogenous) 和內發性 (endogenous) 兩種，而注意力選擇也分為空間為基(space-based) 與物體為基 (object-based) 兩種。本研究旨在探討內發性注意力導向是否為物體為基選擇的必要和充分條件。本研究採用 Egly、Driver、與Rafal (1994) 的提示作業典範，因其具備下列兩項優點：可同時操弄「外引式」和「內發性」兩種注意力導向，並可同時測到空間效果與物體效果，二者分別為注意力選擇是以「空間為基」和「物體為基」的指標。實驗一重複Egly 等人 (1994)的實驗以作為後續實驗的比較基準。在實驗一中，同時有外引式與內發式的注意力導向，並同時觀察到空間效果和物體效果。實驗二使用沒有訊息量的線索使內發性注意力導向無法作用，而只有外引式注意力導向，結果只觀察到空間效果，故內發性注意力導向是物體為基選擇的必要條件。此外並觀察到空間位置上的回向抑制(inhibition of return)，這是以Egly 等人的作業所未曾發現過的。實驗三採用與實驗一同樣有訊息的線索，但讓物體在提示出現之後、目標出現之前消失，以進一步探討內發性注意力導向是否為物體為基選擇的充分條件。結果無法得到物體效果，顯示內發性注意力導向並非充分條件。綜合三個實驗可知，內發性注意力導向是物體為基選擇的必要但非充要條件。欲產生物體為基的選擇，至少還需物體的持續呈現。

In this study, we investigated the relationship between orienting (exogenous vs. endogenous) and selection (space-based vs. object-based) in attention mechanisms. Specifically, we examined whether endogenous orienting is the necessary and sufficient condition for object-based selection. The cueing paradigm of Egly, Driver, and Rafal (1994) was adopted because the location effect and the object effect, which are indicative of space-based and object-based selection, can be observed in the same task. Moreover, with the manipulation of cue validity, exogenous and endogenous orienting could be induced alone or simultaneously in this paradigm. In Experiment 1, we used an informative peripheral cue as in Egly et al. (1994) and replicated their results, showing that both the location and object effects were obtained. In Experiment 2, an uninformative peripheral cue was applied to induce pure exogenous orienting. Only an inhibition of return at the cued location was obtained, which has not been found in previous studies using the Egly et al.’s paradigm. The absence of the object effect when no endogenous orienting was involved thus lead to the conclusion that endogenous orienting is necessary for the object-based selection. Experiment 3 used the same cue as in Experiment 1 but the object disappeared after the cue display. No object effect was obtained when the object disappeared for 200 or 80 ms. Therefore, endogenous orienting is not a sufficient condition for the object effect. We conclude that endogenous orienting is necessary but not sufficient for the object effect.