In this paper we suggest that the cognitive mechanism of suppression attenuates interference in many language comprehension phenomena, and is particularly crucial when comprehension must share processing capacity with other cognitive tasks, as is manifestly the case in simultaneous interpreting. During lexical access, the mechanism of suppression attenuates the interference caused by the activation of other lexical information, such as the inappropriate meanings of homonyms. During anaphoric reference, the mechanism of suppression attenuates the interference caused by the activation of other potential referents. In this way, the referent to which the anaphor does refer becomes the most activated concept. During syntactic parsing, the mechanism of suppression attenuates the interference caused by a previous syntactic form. During metaphor comprehension, the mechanism of suppression attenuates the interference caused by a literal interpretation. During inferencing, the mechanism of suppression attenuates the interference caused by an initial but inappropriate inference. We propose therefore that suppression - a general, cognitive mechanism that attenuates interference - plays a crucial role in language comprehension and simultaneous interpretation.

Among spoken language interpreters, a long-standing question regarding directionality is whether interpretations are better when working into one's native language (L1) or into one's 'active' non-native language (L2). In contrast to studies that support working into L1, signed language interpreters report a preference for working into L2. Accordingly, we investigated whether signed language interpreters actually perform better when interpreting into their L2 (American Sign Language) or into their L1 (English). Interpretations by 30 interpreters (15 novice, 15 expert), delivered under experimental conditions, were assessed on accuracy (semantic content) and articulation quality (flow, speed, and prosody). For both measures, novices scored significantly better when interpreting into English (L1); experts were equally accurate, and showed similar articulation quality, in both directions. The results for the novice interpreters support the hypothesis that the difficulty of L2 production drives interpreting performance in relation to directionality. Findings also indicate a disconnect between direction preference and interpreting performance. Novices' perception of their ASL production ability may be distorted because they can default to fingerspelling and transcoding. Weakness in self-monitoring of signing may also lead novices to overrate their ASL skills. Interpreter educators should stress misperceptions of signing proficiency that arise from available, but inappropriate, strategies.