Current literature on preattentive detection of significant emotional faces in visual search tasks show that these results require an explanation within the theories devoted to explaining similar results obtained for features like orientation and color in the attention literature, the best known candidate for which so far is the feature integration theory (FIT) proposed by Treisman and Gelade. This paper attempts to bridge such a gap between the attention and the emotion theories by proposing that `emotional features' should be proposed within the FIT due to their resemblance with the other features of the theory like their preattentive detection. Within such a scheme the amygdala will act as a substitute for the cells in the striate cortex that are shown to be responsible for processing the visual features, as amygdala has been shown to process faces with significant emotional value preattentively. The paper also proposes an experimental paradigm that can provide empirical support for such emotional features.