We model countersignaling (i.e., very high types refraining from signaling) arising from the tradeoff between persuasion and learning in a signaling game. We assume that the agent has imperfect private information regarding his/her productivity, which the signaling action provides additional...