This paper examines the key linguistic arguments underpinning Basil Bernstein's theory of "elaborated" and "restricted" "codes". Building on a review of selected highlights from the collective critical response to Bernstein, the paper attempts to clarify the relationship of the theory to "deficit" views and to explore the conceptual roots of Bernstein's position as well as of the linguistics which informs it. The paper finds that Bernstein's theory qualifies as a "deficit" position on a number of counts, most particularly due to the alleged cognitive implications of the "codes". However, the paper argues that no convincing evidence or rationale for the existence of such "codes" has ever been provided. The paper gives support to the argument that Bernstein's "code" theory has its roots in a particular model of literacy and, as such, is best understood as a variant of the more traditional ethno- (and socio-) centric "great divide" perspective. The paper further argues that the relatively recent re-working of Bernstein's position in the work of Ruqaiya Hasan does not succeed in overcoming its principal theoretical and methodological failings. (Contains 22 notes.)