Third code research has documented the distinctiveness of translated language and singled out recurrent tendencies framing them as translation universals. This paper aims to identify the interaction between interference, explicitation, implicitation and normalization and their potential relationship(s) with other variables, such as register. The focus of the study is on Spanish discourse markers (DMs) translated from English. This study uses interference, explicitation, implicitation and normalization as methodological tools to unveil these patterns. Evidence comes from a bilingual parallel corpus (P-ACTRES 2.0), a corpus of translated Spanish (CETRI), and a reference corpus of contemporary Spanish (CORPES XXI). We select the input DMs according to two criteria: first, we focus on DMs showing cross-linguistic formal correspondence, indicating the possibility of grammatical interference; second, we consider different procedural meanings for the DMs to anticipate potential regularity distortions. Results indicate that DM underuse in the target texts generally co-occurs with explicitation. Register is an important variable: implicitation is more frequent in non-fiction and, together with normalization, affects the majority of DMs. Evidence also points to the DMs' semantics influencing implicitation and explicitation.