I will first establish a simple framework for considering the information needs of various types of reader. Then in the rest of the paper I will use the framework to consider first the role that journals do and should play in the dissemination of economic ideas, and then the tasks that necessarily are left for books to perform. This program, of course, gives me many opportunities to criticize others and to exhort economists to write the books that I want to publish.

My topic is not dissemination of ideas to the intelligent lay reader, but dissemination within the discipline, that is, from the highbrow to the highbrow and from the highbrow to the “middlebrow.”

What does your reader need to be told?

When we communicate with others we vary the form of the message to suit the nature of the recipient. With some a very terse, even telegraphic, message will suffice. With others a fuller but still demanding and terse statement will communicate effectively. Yet a third case requires that our message provide a substantial amount of background information if it is to be understood. The essential factor deciding which of these message types to use is the amount of relevant information that the recipient already has to help him understand the message. Given that most messages reporting research results are to be conveyed to a variety of recipients, more than one form of the message may well be needed. If one form only is to be provided, it needs to be designed to convey its information to the largest feasible part of its total potential audience.