In an important paper on the English “double-object”, or ditransitive, construction, Richard Hudson proposes a hypothesis that conflates the ditransitive direct object, or O2, and the monotransitive direct object, or OO, into the same syntactic functional category. While making important departures from a number of unfortunate assumptions within mainstream formal theories of linguistics at the time, the OO = O2 hypothesis itself is problematic in the perspective of contemporary cognitive linguistics. This paper addresses the hypothesis from the perspective of usage-based construction grammar. Applying simple collexeme analysis and multifactorial heatmap analysis to instances of OOs and O2s in ICE-GB, this paper shows that the usage-patterns of both are far too complex, displaying cross-register variation, for the OO = O2 hypothesis to be tenable. The findings provide support for a usage-based variationist account in defining syntactic functional categories.