There are many factors influencing the selection of a particular translational behaviour or its avoidance. Although there is no real list, this array of factors is essentially heterogeneous: some variables are cognitive, others cross-linguistic or socio-cultural. Due to this vastness and heterogeneity, there can be no deterministic explanation in Translation Studies. Firstly, there is no single factor that cannot be enhanced, mitigated or even offset by the presence of another. Secondly, there are always several factors interacting, hence influencing each other as well as the selected behaviour. This paper attempts to discard deterministic reasoning, presenting a different format of reasoning; conditioned, and hence probabilistic, and to define the aim of TS as gradually moving towards an empirically-justified theory consisting in a system of interconnected, even interdependent probabilistic statements. This paper aims to return to all these issues to ask whether such explanations qualify as “universals of translational behaviour” and, if not, whether there are any other candidates for universal-ship.