A word about context: The discussion on the Western Canon mailing list had been concerned with whether Voltaire's Candide is or is not a great book. Adler's remarks were prompted by this question.

As Editor in Chief of GBWW's second edition I worked with an editorial board that consisted of the following persons: Douglas Allanbrook, Senior Tutor and Associate Dean, St. John's College, Annapolis, Maryland; Jacques Barzun, Provost Emeritus, Columbia University, and literary adviser, Charles Scribner's Sons; Norman Cousins, Professor of Medicine, University of California at Los Angeles; John Kenneth Galbraith, Professor of Economics, Harvard University; Heinz R. Pagels, Director, New York Academy of Sciences; Lord Quinton, former Chairman, The British Library Board, London, and also former President, Trinity College, Oxford. [1]

In addition to these associates, we formed an international committee of consultants to whom the nominations made by the editorial board would be sent for approval or disapproval, as well as for comments and recommendations.

When we had completed a second draft of the nominations for inclusion in or omission from the second edition, the next and final step in constituting the contents of the set involved submitting this second draft for consideration to the Board of Editors and to the University Advisory Committees.

The members had been provided in advance with various lists of nominees for addition to the set, especially twentieth-century authors and titles. I opened this meeting by stating the criteria for selection that Hutchins and I had employed in the 1940s when we met with a similar editorial board to decide on the authors and titles for inclusion in Great Books of the Western World. Now as then, considerations of space played a critical role. Some things had to be rejected or eliminated to prevent the set from becoming economically unfeasible to produce, distribute, or purchase and use.

At the end of a long and, on the whole, pleasantly harmonious session, we came up with our first draft of authors and titles. Before this draft was submitted to our international committee of consultants and other groups, a footnote had to be added to it stating the authors and titles in the first edition that we proposed to drop from the second. There were four: the "Conics" of Apollonius of Perga; Joseph Fourier's "Analytical Theory of Heat"; Henry Fielding's "Tom Jones", and "Tristram Shandy" by Laurence Sterne. The first two of these were thought to be mathematical treatises of unusual difficulty for most readers to comprehend.

We then proceeded with submitting our first draft to our international committee of consultants, to Britannica's Board of Editors, and to our University Advisory Committees. This process involved much discussion and correspondence over many months, at the end of which we came up with a final draft of the second edition's table of contents, including a list of the new and much better translations that we sought to acquire from their publishers.

Before I state the three criteria that our editorial board employed for its first draft selections and that we asked all the persons we consulted to keep in mind in judging what we submitted to them, let me say that at no point did we attain unanimity. One hundred percent agreement is too much to expect in proceedings of this kind. However, where there were unresolved disagreements, these did not exceed more than 10 percent; i.e., the items about which such disagreement occurred were less than 10 percent of the whole. That, it seems to me, is remarkable, and also sufficient to rely upon.

When all these preliminaries were completed and after the work of editorial production had begun, I found myself dissatisfied with three decisions we had made (much less than 10 percent of the whole). I regretted dropping the "Conics" of Apollonius, which was not much more difficult than Euclid's "Elements", which we retained in the set. I thought we were wrong in dropping Fielding's "Tom Jones", as the frequency in the Syntopicon of references to its contents attested, indicating its substantial presence in the great conversation. And I thought we were wrong in adding Voltaire's "Candide." Voltaire is a great author and one of enormous influence, but by our three criteria for selection, "Candide" is not a great book. [2]

What were those three criteria of selection? The first was the book's contemporary significance -- relevance to the problems and issues of the twentieth century. The books were not to be regarded as archaeological relics -- monuments in our intellectual tradition. They should be works that are as much of concern to us today as at the time they were written, even if that was centuries ago. They are thus essentially timeless -- always contemporary, and not confined to interests that change from time to time or from place to place.

The second criterion was their infinite rereadability or, in the case of the more difficult mathematical and scientific works, their studiability again and again. Most of the 400,000 books published each year are not worth carefully reading even once; many fewer than 1,000 each year are worth reading more than once. When, infrequently in any century, a great book does appear, it is a book worth reading again and again and again. It is inexhaustibly rereadable. It cannot be fully understood on one, two, or three readings. More is to be found on all subsequent readings. This is an exacting criterion, an ideal that is fully attained by only a small number of the 511 works that we selected. It is approximated in varying degrees by the rest.

The third criterion was the relevance of the work to a very large number of great ideas and great issues that have occupied the minds of thinking individuals for the last twenty-five centuries. The authors of these books take part in the great conversation, not only by reading the works of many of their predecessors, but also by discussing many of the 102 great ideas treated in the "Syntopicon". In other words, the great books are the books in which the great conversation occurs about the great ideas. It is the set of great ideas that determines the choice of the great books.

In a book entitled "The Great Conversation", which is not a part of the set's second edition but which accompanies it as an introduction to the set and as a guide to its use, we have demonstrated this point by two devices. One is something that we called the Author-to-Author Index, which shows how many of each author's predecessors that author has cited in his work. The other is the author-to-Idea Index, which shows in how many of the 102 great ideas treated in the "Syntopicon" readers will find references to that author's work on one or more topics, usually many. These two indices, along with the "Syntopicon" itself, are clear evidence of the reality of the great conversation, in which the great authors and the great books have participated.

By this criterion, the difference between great books and good books is not a difference in degree, but a difference in kind. There is not a continuum that has poor books on the far left, average books in the middle, good and very good books on the right, and a few Great Books on the far right.

As I have recently written elsewhere, the adjective "great" in the phrase "great books" derives its primary meaning from its use in the phrase "great ideas." There are many other criteria by which people make up diverse lists of the books they wish to honor by calling them "great books." But from the primary significance of the adjustive "great" as applied to the great ideas is derived the significance of that adjective as used in the phrase, "the great conversation."

In other words, we chose the great books on the basis of their relevance to at least 25 of the 102 great ideas. Many of the great books are relevant to a much larger number of the 102 great ideas, as many as 75 or more great ideas, a few to all 102 great ideas. In sharp contrast are the good books that are relevant to less than 10 or even as few as 4 or 5 great ideas. We placed such books in the lists of Recommended Readings to be found in the last section in each of the 102 chapters of the "Syntopicon." Here readers will find many twentieth-century female authors, black authors, and Latin American authors whose works we recommended but did not include in the second edition of the Great Books.

To complete the picture of the criteria that controlled our editorial process of selection, it is necessary for me to mention a number of things that we definitely excluded from our deliberations.

We did not base our selections on an author's nationality, religion, politics, or field of study; nor on an author's race or gender. Great books were not chosen to make up quotas of any kind; there was no "affirmative action" in the process.

In the second place, we did not consider the influence exerted by an author or a book on later developments in literature or society. That factor alone did not suffice to merit inclusion. Scholars may point out the extraordinary influence exerted by an author or a book, but if the three criteria stated above were not met, that author or book was not to be chosen. Many of the great books have exerted great influence upon later generations, but that by itself was not the reason for their inclusion. [3]

In the third place, a consideration not operative in the selection process was the truth of an author's opinions or views, or the truth to be found in a particular work. This point is generally misunderstood; many persons think that we regard the great books as a repository of mankind's success in its ever-continuing pursuit of the truth. "That is simply not the case". There is much more error in the great books than there is truth. By anyone's criteria of what is true or false, the great books will be found to contain some truths, but many more mistakes and errors.

Mortimer Adler
Center for the Study of The Great Ideas
　

[Back to text]

[Back to text]

[Back to text]