What happens if a MapReduce job uses one reducer?

A. No shuffle occurs B. All intermediate keys are sent to a single reducer and final output can be globally sorted by key C. Every mapper writes one shared file concurrently D. The job cannot aggregate data