Translation could be defined as a textual process. Text has distinct features which differentiate it from non-text. These distinct features constitute standards of textuality, and when these standards are violated, the text sounds unnatural and foreign. Cohesion is one of the standards of textuality; it is mainly concerned with mutual connectivity of items in the surface structure of a text. Every language has its own set of cohesive devices and preferred means for creating cohesive harmony (Hasan 1985) and bonding patterns (Hoey 1991). As a result, shifts in cohesion inevitably occur in translation. The objective of this study is to examine potential shifts in cohesion and bonding patterns between a set of English source texts and their Korean translations and to explore the nature of motivations behind these shifts. Two types of shift are expected to occur: language-oriented cohesion shifts, motivated by systemic differences between the two languages, and translation-oriented cohesion shifts, caused by the translation process. Identification of language-oriented cohesion shifts presupposes the recognition of intrinsic cohesion differences between Korean and English. In the absence of comprehensive research documenting such differences, this study attempts to address the issue by comparing a number of original (i.e. non-translated) English source texts and a number of original Korean comparable texts. The study sets out to test a number of hypotheses. The main hypothesis is that there are cohesion and bonding pattern shifts between English source texts and their Korean translations. These hypothesized shifts may be motivated by: (a) intrinsic linguistic differences between English and Korean and translators' attempts to abide by target language textual norms and expectations: and/or (b) the nature of the translation process. The secondary hypothesis in this study is that translated Korean texts display patterns of cohesion which are different from those evidenced in a comparable collection of non-translated Korean texts. These hypothesized differences may be explained in terms of several features posited in the literature as being characteristic of the translation process, including source language interference (see, for instance, Toury's law of interference, Toury 1995: 275). The analysis reveals that there are cohesion and bonding patterns specific to each type of text under analysis in this study, i.e. English source, Korean target, and Korean comparable texts. These differences are potentially motivated by intrinsic linguistic and textual differences between English and Korean and/or by the nature of the translation process itself.