In this paper, we aim to challenge what we see as two misconceptions in the literature on sentential embedding. The first, due to Kiparsky & Kiparsky (1971), posits that the complements of factive verbs are structurally more complex than the complements of non-factive verbs, an idea that has been challenged in much recent work. The second, related misconception, is the idea that factivity itself is a concept that is active in syntax, determining the structural properties of embedded clauses. This idea has also been challenged through the years, but there has been no clear consensus on the way forward. We present arguments that the referentiality of complement clauses is what matters for syntax, as opposed to factivity (part of the semantic component) or givenness (part of the pragmatic component). We compare our proposal to a competing analysis (Kallulli, 2006, 2009), which argues for the Kiparskian view that factivity is represented by extra syntactic structure, and which blurs the line between factivity and givenness. We conclude that our account, building on the referentiality of the embedded clause itself as opposed to the lexical class of the selecting predicate, can potentially go a long way towards clearing up the conceptual, empirical and terminological confusion in the realm of clausal complementation.