Sight Translation as a Testing Ground for Default Linguistic Patterns

This presentation reports on an experimental study using sight translation, i.e. an original or source text in written form is rendered orally in a different language (Chen, 2015). In this experiment, 14 professional interpreters sight translated a series of caused motion events from Spanish (their B language or L2) into Danish (their A language or L1). The texts containing the relevant caused motion events were displayed on a screen in the form of running sentences, thus adding time pressure. The choice of Spanish and Danish as working languages was motivated by the fact that they differ radically in their prototypical lexicalization pattern (Talmy 1985, 2000, 2017) for expressing motion. In short, Spanish typically uses directional verbs (i.e., Saca bombones de la caja ‘She extracts chocolates from the box’), while Danish typically encodes the direction in a post-verbal particle and uses the main verb to express how the motion is performed (i.e., Hun hælder chokoladen ud af æsken ‘She pours the chocolate out of the box’). This has enormous implications for translation, since translating between typologically different languages requires a great deal of retextualization (Rojo & Valenzuela 2013). When translating from a Germanic into a Romance language, the process of rearranging how the relevant semantic components of a motion event are to map onto each linguistic form often results in a loss of information (see, e.g., Molés-Cases 2019, Molés-Cases & Cifuentes-Férez 2021, Ibarretxe-Antuñano & Filipovic 2013). This omission of information can be termed implicitation (Molés-Cases 2019). The present study, however, aims at investigating in the opposite direction (from Romance into Germanic) and gaining insights about explicitation, that is, cases where interpreters add or make explicit certain information that was implicit in the source text (Vinay & Darbelnet 1958, 1995). In the present study, explicitation is measured in types (specific, verbalized linguistic forms) and tokens (number of types). The general assumption is that there is a general tendency towards explicitation of certain semantic information when translating from Spanish into Danish. The results indicate indeed that explicitation occurs systematically, with sight translators making finer-grained distinctions that are not present in the Spanish source texts (i.e., specifying verticality and not just direction).