التحولات الناجمة عن الترجمة في كثافة أدوات الربط في اللغات الهدف: التماسك النصي والإيضاح

The using of conjunctions is one of the most sensitive indicators for showing how translated texts differ from both their source texts and even more so from native target language texts. In this paper, we consider the seemingly well-established claim largely associated with the explicitation hypothesis that translation significantly increases the frequency of overt conjunctive markers compared to a comparable non-translated text in the same language. Based on cohesion theory, corpus-based translation studies and contrastive work (on Arabic and English discourse organisation), the paper synthesises the theoretical mechanisms that have been proposed to account for this tendency, evaluates the empirical data that support and complicate it, discusses conjunction density as one type of explicitation which is obligatory, optional, pragmatic or inherent in translation and situates conjunction density within a general typology of explicitation. We focus especially on the Arabic–English pair, where a kind of close but ambivalent relationship exists between the source language own preference for coordination and the explicitating pressures translation is said to introduce, which confers a dual yet disabling effect. To validate the hypothesis across corpora and languages, we outline a method that scholars interested in operating at the level of empirical testing can draw on in their future work: one combining methods to generate normalised frequency counts of conjunctive relations with frameworks for taxonomic classification of those relations among comparable-corpus triangulation. It ends up with the conclusion that conjunction density shift is real but qualified—not a robust translation universal, but dependent instead on a rich interaction between universally operative processing pressures on translators in combination with structural distances between source and target languages, and an array of genre -conventions and translator-normalized preferences.