Score contribution per author:
α: calibrated so average coauthorship-adjusted count equals average raw count
Communication through ordinary talk is very common. Often, the receiver of information has the option to investigate a sender’s type. We extend a standard cheap-talk model by giving the receiver this option. We show that when the receiver bears most of the investigation cost, the option to investigate drives away most cheap-talk communication. When the sender pays a substantial part of the investigation cost, the option to investigate disciplines the sender. Introducing an investigation device to a cheap-talk model demonstrates the vulnerability of cheap talk on the one hand, and shows how communication can be improved on the other.