The paper broadens the scope of the Comp-trace effect with additional phenomena, like the ban on scrambling -ga subjects in Japanese and ECM in French, arguing for a reduced clause approach to the lack of Comp-trace effects in who do you think left, where a null C induces Comp-trace effects. This leads to fine-tuning the position of various subject and wh-phrases (within a Split IP perspective), and a new contextual conception of the EPP that is put into a broader perspective regarding a more general theoretical move toward contextuality in syntax. In particular, it is argued that there is an EPP domain, with the highest projection in this domain functioning as the locus of the EPP (on a par with similar approaches to phases and phasal edges). To address all these issues, the paper explores low left periphery and the possibility of multiple subject positions. It is argued that the landing site of subject wh-movement (who in who left) is lower than non-subject wh-movement (who in who did Mary leave), but higher than regular subjects (Mary in Mary left). Quirky subjects, on the other hand, are argued to be lower than regular agreeing subjects, with a new generalization proposed regarding their crosslinguistic availability. From this perspective, the paper addresses the issue of the proper analysis of a number of other constructions, including Japanese raising-to-object, Germanic V-2 clauses, existential constructions (a non-uniform analysis is argued for there), locative inversion, and clausal subjects.