Against the root analysis of subject contact relatives in English

•Evaluation of den Dikken's (2005) topic-comment analysis of English SCR.•The predictions of the topic-comment analysis for the internal syntax are incorrect.•The predictions of the topic-comment analysis for the external syntax are incorrect.•Structurally, SCR and V2 relatives pattern differently.

Based mainly on a number of interpretive considerations, Henry (1995) and den Dikken (2005) elaborate an analysis of English SCR which mirrors their discourse function. However, while the proposed topic-comment representation may well reflect the information structural properties of SCR, its predictions for the internal and external syntax of SCR are incorrect. In contrast, the predictions would be correct for Germanic V2R, undermining the basis for den Dikken's (2005) assimilation of the two patterns. On the basis of the discussion above I thus conclude that a topic-comment representation for English SCR, though attractive on interpretive grounds, cannot be maintained. The various empirical data discussed in this paper show that a relativization analysis along the lines of that in Doherty, 1993, Doherty, 1994 and Doherty, 2000, reviewed in the light of our current understanding of the articulation of the left periphery (see also Sistrunk, 2012), is to be preferred.