An SFL-based model for investigating explicitation-related phenomena in translation: two case studies of English–Arabic translation