This paper argues: (1) that one of the major syntactic constraints adopted by many proponents of the Extended Standard Theory, namely the Specified Subject Condition (SSC), is empirically inadequate with respect to "unbounded" extraction phenomena; and (2) that the unbounded extraction data which the SSC purported to account for need to be accounted for in nonsyntactic terms. After a review of the research on the subject, Oerle's counterproposal is discussed and found inadequate. A nonsyntactic alternative to the extraction data to which Chomsky and Oerle addressed themselves is offered and the role of context with respect to them is stressed. Finally, it is argued that the SSC is insufficient and unnecessary with respect to unbounded extraction phenomena, and that it cannot be maintained within existing versions of the (Revised) Extended Standard theory, unless one is willing to countenance either an increase in theoretical power, or an unnatural characterization of the class of phenomena which the SSC is supposed to constrain. (Author/AMH)