Rod O'Donnell's (1999a, 1999b) and James Ahiakpor's (1999) contributions on the origin of the only diagram in J. M. Keynes's General Theory raise some interesting points regarding the use of documentary evidence for interpreting authors' ideas or specific passages in their writings.

In his recent article, Robert Dimand details for the profession the existence of four obscure papers by Irving Fisher (1935, 1936, 1937, 1940). He also brings to our attention two brief Econometrica contributions (Fisher 1946, 1947), neither of which is listed in the American Economic Association's Index of Economic Journals, as he points out. For this, and for his careful discussion of the content of those papers, we are indeed in his debt.

Heterodox economics today is alive and well, and will always be with us. It is generated by opposition to mainstream orthodoxy and performs the important function of preventing the mainstream from stagnating. Orthodoxy and heterodoxy feed off each other, and both are transformed in the process.

This topic is more in contemporary economics than in the history of economics, but in this case an historical perspective does help us to understand the present day. Roger Backhouse and Sheila Dow have approached the progress of heterodox economics mainly through the philosophy and methodology of the discipline. I take a more sociological approach. I see the issue as external to the discipline at least as much as internal.

My comments on the papers given by Bob Coats, Roger Backhouse, and Sheila Dow will focus on the history and prospects of institutional economics, as that is the part of the heterodox world I know most about.

According to the folk history of transaction costs, the concept is due to a seminal article by Ronald H. Coase, written in the 1930s. Failing to provide an operational framework, Coase's article was neglected for a long time, or so the story continues. In the 1970s, after the limits of the Arrow-Debreu paradigm had become obvious, several authors, including Oliver E. Williamson, Kenneth J. Arrow, Armen A. Alchian, and Harold Demsetz, took up the notion of transaction costs and turned it into a useful analytical tool. Partly responsible for this sudden upsurge of interest was an article by Coase in which he allegedly proposed the now famous “Coase Theorem.”