The Gravitational Pull Hypothesis is an attempt to provide a cognitive account for features of translated language. It assumes that translated and non-translated texts in the same language exhibit distributional differences that can be regarded as translational effects. This book presents a number of studies aiming to test that hypothesis on five linguistic items: passive construal of events, diminution, verbal aspect, light verb constructions and adjective position. The studies draw on data from the COVALT corpus as well as elicitation and translation tasks performed by professional translators. The results shed light not only on the hypothesis itself but also on the mixed-methods approach adopted in the book.