The relationship between Descriptive Translation Studies and Translation Theory, maintains Gideon Toury, is one of interdependency: the results of observational and experimental descriptive–explanatory research, which reveal what translation "DOES involve, under various sets of circumstances, along with the REASONS", verify, refute or modify theoretical assumptions about what translation "CAN, in principle, involve". In turn Translation Theory, on the basis of empirically established "regularities of behaviour", predicts what translation is "LIKELY to involve, under one or another array of specified conditions" (Toury 1995: 15–16). These predictive statements are expressed in terms of conditioned, probabilistic laws of translational behaviour whose basic format is: "The presence of 1, 2, 3, … ∞ enhances the likelihood that X (or: reduces the likelihood that no-X)" (Toury 2004: 26). The laws of translational behaviour explain the relations existing among a myriad of cognitive, cross-linguistic and sociocultural variables that influence a particular translational behaviour or its avoidance (2004: 15). Toury assigns to Descriptive Translation Studies a vital role in the evolution of Translation Studies (Toury 1995: 265) because, through the accumulation of findings concerning actual translational behaviour, it should be possible to move "gradually, and in a controlled way, towards an empirically justified theory which would consist in a system of interconnected, even interdependent probabilistic statements", this being the ultimate aim of Translation Studies (Toury 2004: 15). Although Toury prefers the term "laws" to denote the probabilistic pronouncements put forward by Translation Theory, he concedes that they qualify as universals of translational behaviour, whose value lies in their explanatory power rather than in their existence. He also recognizes that the notion of translation universal "is one of the most powerful tools we have had so far for going beyond the individual and the norm-governed" (2004:29).

In line with Toury, Andrew Chesterman (2000, 2004) views the quest for universal features of translations as the descriptive route through which scholars propose and look for generalizations about translation. These general regularities or laws, he explains, are explored by operationalizing abstract notions about universals and putting forward and testing, through a comparative model of translation, general descriptive hypotheses about the existence of similarities between different types of translation, without disregarding either the differences between them or the uniqueness of each particular case. Chesterman distinguishes between S-universals, which refer to "universal differences between translations and their source texts" and T-universals, which refer to “universal differences between translations and comparable non-translated texts” (Chesterman 2004: 39). If universals, which are essentially descriptive constructs, are supported by extensive empirical evidence, they can have explanatory force as regards the occurrence of a given feature in a particular translation (Chesterman 2000: 26). This means that if a universal is verified, it acquires the status of law, so one can say that a given translation exhibits a particular feature under given conditions because all translations tend to exhibit that particular feature under those same conditions (Chesterman 2005: 198). The causes of universals, on the other hand, are sought in neighbouring fields of scientific enquiry, such as human cognition. Discourse transfer, observes Toury (1995: 275), is an example of translation universal that can be explained by the basic mental processes involved in translation, particularly the rapid switching between source and target codes and vice versa. Therefore in Translation Studies, like in any discipline, general explanatory laws not only permit us to make predictions about future cases but also create vital interdisciplinary links (Chesterman 2000, 2004).