‘Mere narrative’, ‘lapses into narrative’, ‘a narrative answer that fails to answer the question set’. These phrases flow in the blood of history teachers, from public examination criteria to regular classroom discourse. Whilst most of us use narrative in our teaching methods, we have demonised narrative in pupils’ written answers. Sean Lang persuades us to rethink our demonology. He suggests that it is not really narrative that we are objecting to – it is unthinking regurgitation, mere reproduction of content, uncritical story-telling, rambled ‘pourings out’ in the same discourse form that pupil received it from teacher or text. In other words, it is actually ‘knowledge-telling’ that we are rightly objecting to, not narrative itself. Our assumption tends to be that in order to carry out the opposite of ‘knowledgetelling - the difficult ‘knowledge-transforming’ – pupils must break out of narrative, break out of chronology and switch to thematic analysis. That thematic analysis remains demanding to do and essential to teach is not questioned. But is it not also possible to ‘transform knowledge’ by writing a narrative? A narrative is not a given. It is a construct. A good one is incredibly difficult to write. It can be written in multiple ways for multiple purposes and audiences. It is also remarkably versatile as a tool for communicating an historical point. Narrative can be used to carry out some historical practices that we too easily assume can only be done by a non-chronological arrangement. Narrative can explain, narrative can argue, narrative can reveal tight relationship with sources of evidence. Sean makes the case for teaching all pupils the difficult of art of narrative. He suggests that we should elevate it into a goal for serious study and that we should reward expert practice in narrative in our schema for progression and our criteria for assessment.