Methodology

This electronic archive is dedicated to creating, preserving, organizing, and publicizing the best available scholarly information and digital resources on early American bestsellers. In order to delineate the horizons of this archive, I have defined each of those words--early, American, and bestsellers--below.

EARLY: The Oxford English Dictionary attests that the word bestseller was first used 0n April 25, 1889 by a writer for the Kansas Times & Star, who listed six 'highbrow' books that sold well in Kansas City as evidence that the region's literary taste was improving. Frank Luther Mott explains that the first "bestseller" list appeared six years later in 1895, when Harry Thurston Peck launched a New York edition of the older, monthly, London periodical, The Bookman (204). Because Peck's list and its successors provide a guide to sales figures throughout the twentieth century, this archive will limit itself to identifying and creating resources for American bestsellers published and distributed in and before the year 1900.

AMERICAN: While the adjective American has been applied to most of the Western Hemisphere, this archive uses it to describe the behaviors of peoples who lived on lands now belonging to the United States. Because the study of bestsellers is not a study of literary production (how a book was written) but literary consumption (how well a book sold, and how widely it was read), the adjective American in the phrase "early American bestsellers" refers to American consumption and not American production. This means that books consumed by readers living in the United States or its preceding colonies but produced by authors living in Europe or other locations--Sir Walter Scott's Ivanhoe or Emanuel Swedenborg's Heaven and Hell, for example--would qualify as American bestsellers even though they were written outside the boundaries of the United States (and in Swedenborg's case, another language!).

BESTSELLERS: When Mott compiled the first book-length study of American bestsellers, he defined a bestseller as a book "believed to have had a total sale equal to one percent of the population of the continental United States (or the English Colonies in the years before the Revolution) for the decade in which it was published" but omitted "bibles, hymnals, textbooks, almanacs, cookbooks, doctor-books, manuals, and reference works" (303). His study also includes a list of "better sellers" which did not meet the one percent standard but came close. The archive's editors generally approve of Mott's criteria, but we have modified Mott's criteria in the following ways:

1. Mott uses the date of a text's first American printing to determine the requisite one percent sales figure, which means that some early British works (like John Calvin's Instititutes of the Christian Religion, which was first published in the United States in 1813 but which was most popular in the seventeenth century) which arrived and sold very well as imports have been excluded from his list. This archive uses the date of a text's first publication in English or the year 1650, whichever comes later, to establish the necessary one percent sales figure.
2. We reject Mott's excision of nonfiction and reference texts as an arbitrary imposition of modern sensibilities. While modern readers may not regard the Bible or cookbooks as "literary texts," early Americans often consumed (pun very much intended) these texts with as much or more gusto than they displayed for fiction, and we will accordingly treat bestselling texts without regard for genre.
3. Because Mott included all sales of a given text from the date of its first American printing until he wrote his study in 1947, Mott produces several "posthumous bestsellers," texts that were NOT popular with contemporary readers whose sales figures were artificially inflated by the recovery efforts of literary scholars and schoolteachers in the early twentieth century. To wit: Mott sets the "one percent" sales figure for texts written in the 1850s at 225,000; based on that criteria, he includes Moby-Dick (1851) as a bestseller, even though the book sold fewer than 3,000 copies (the original press run) until it was reprinted after Melville's death in 1891. This archive will, as far as possible, weed out these "posthumous bestsellers;" texts whose cumulative sales until 1947 that were not steady sellers within the lifetime of the author or fifty years after the text's original publication will be omitted.
4. As a supplement to Mott's list we will include some of the "popular" books identified by James Hart, generally books which Mott has omitted either because of his insistence on using the American publication date, his inattention to imported books, or for reasons of genre.