A controversy regarding relative-clause processing was raised by Hsiao and Gibson (2003), who showed a processing preference for object extractions in Mandarin Chinese. This article argues that Hsiao and Gibson's claim of an object preference was invalid as their experiment was confounded. Two sets of experimental evidence from Mandarin—the self-paced reading of regular relative clauses and that of possessor relative clauses—showed a preference for subject extractions. As the controversy is removed, the Incremental Minimalist Parsing theory is discussed to account for this universal parsing preference.