In this study, a novel version of the spatial cuing paradigm is used to directly compare 100% valid, spatial and non-spatial symbolic cues. A variety of setup times, SOAs, and target displays are utilized in order to thoroughly examine any differences in how space, color, and form symbolic cues direct attention from a cue to a target. This investigation includes both linguistic and non-linguistic symbolic cues. Costs for non-spatial cues when compared to spatial cues are best accounted for by a binding hypothesis that posits a necessary binding of location information when non-spatial cues are being used to direct spatial attention. Alternative explanations that are addressed include the nature of space (Experiment 1), perceptual feature encoding (Experiment 2), cue processing (Experiment 3), attentional guidance (Experiment 4), and object-based attentional selection (Experiment 5). The reported findings are also discussed with respect to their significant contributions to research areas in attentional control, visual selective attention, space-based vs. object-based attention, and the nature of space.