The present list is a thorough revision of the version originally posted in 1999, essentially constituting a new edition of this work. It utilizes an improved ranking system that more faithfully represents and reconciles the rankings of the source lists. My appreciation to John Young for demonstrating the deficiencies in the system previously used.

Together the four source lists feature 253 different titles. Thirty of these are excluded from this composite version for various reasons (published prior to the 20th century, not fictional works, not originally written in English), but this still leaves an impressive 223 works to represent the best of the century. The sources provide different selections of titles as well as rating differently those they share in common, but also display a broad degree of consensus. Twenty-four titles in the composite list are common to all four sources, thirty-one are common to three and sixty-one are common to two. Thus, there are 116 "best" works on which there is at least some agreement, as opposed to 107 that stand alone.

Part of the apparent consensus may be due to copying. Undoubtedly the Modern Library list, the first to appear, had some influence on the later lists, all of which were issued in response to it. Yet there is also evidence of agreement on a deeper level. For instance, the Library Journal and Koen lists, apparently compiled independently, show broad agreement in their top titles and rankings.

This is not to imply that the sources lack flaws and biases. Indeed, their failings prompted the compilation of the present work. Details on these are provided in individual discussions of each source. Generally, the flaws consist of

• fuzzy inclusion standards (inconsistency as to what constitutes a single work, resulting in the inclusion of short story collections, trilogies and series).
• divergent inclusion standards (exclusion of children's fiction by one list, inclusion of translations by another).
• inclusion errors (admission of works that fail to meet the lists' own inclusion criteria, such as nineteenth century works in the Koen and Radcliffe lists, translations in the Koen and Modern Library lists, and nonfictional works in the Koen, Library Journal and Radcliffe lists.

The current composite list attempts to reconcile the ratings in the sources and cancel out their biases. In this regard, I believe it an improvement over its predecessors. I have not presumed to add additional works, feeling the sources' large-scale efforts will have picked out the truly worthy better than I or any other single person could.

My listing includes most titles from the source lists, ranked according to their cumulative weight in the sources. As noted above, works translated from other languages, non-twentieth century works, and nonfictional works are excluded. Other divergent works, such as series, trilogies and collections, were allowed to remain, as some at least can validly be regarded as single works.

Qualifying titles from the source lists were re-ranked in a manner that both reflects the sources' rankings and rewards appearance in two or more sources. Details about the method used are provided in the explanation of the ranking system.

A prime feature of the composite list is to show what source lists a given work appears in and how each ranked it, as well as its ranking in the current list. For further information on the data included, see the column key.

I owe particular thanks to

• Esther Erman and Geoffrey Skinner for assiduously identifying errors in draft versions of this work prior to the initial posting,
• Mitchell Marks for observing that there should be no apostrophe in the title of Finnegans Wake,
• Jon Skavlan for spotting a math error (four years after posting!) requiring the re-ranking of Fahrenheit 451,
• Elese Veeh for pointing out that the author of Bonfire of the Vanities is Tom Wolfe, not Thomas,
• Patrick O'Daniel for observing that the final "r" had been left off The Outsider,
• John Young for demonstrating fundamental flaws in the system I originally used to reconcile the source lists,
• Jim Smith for finding typos in the entries for Ford Madox Ford, The Killer Angels and The Magus and misdatings of Animal Farm, The Good Soldier, Studs Lonigan, and U.S.A., and for helping in the resolution of a complicated problem regarding The Wapshot Chronicle,
• J. Michael Andresen for an impressively extensive statistical critique of the current rating system, which in due course will result in another revision thereof,
• and Ross Jallo for discovering I had left the initial article off The Satanic Verses.