This article examines the history of the first quantitative content analysis of a newspaper. John Gilmer Speed, a former New York World editor, used this research method to compare the content of four New York dailies published in 1881 and in 1893. He concluded that "new journalism" had injected high levels of gossip and scandal into newspapers during the twelve-- year interval. The new material adversely affected readers in two ways, he believed: It displaced useful news that readers needed to function in a democratic society, and it provided examples of poor behavior that readers might imitate. His study served as a foundation for later academic "muckrakers," who used content analyses to critique newspapers' interactions with other social institutions.
As Joseph Pulitzer planned in the spring of 1893 to celebrate the tenth anniversary of his purchase of the New York World with a mammoth, 100-page edition, a former World editor prepared a shorter document that he hoped would demonstrate the danger that Pulitzer's "new journalism" represented for readers. Others had filled scores of magazine and newspaper columns in the past with critiques of Pulitzer's methods, but the contribution of John Gilmer Speed, a civil engineer by training, was different. Rather than base his argument only on logic and anecdotes, he used numbers to demonstrate that Pulitzer and his imitators cluttered their newspapers with accounts of scandal and private gossip. The finished project, published in Forum magazine, earned Speed a footnote in scholarly literature as the author of the first quantitative content analysis of a newspaper.1 Analyses of written communication similar to Speed's were at least a century old before he employed the method. In the late eighteenth century, for example, Europeans used content analyses to detect heresy in religious hymns, and an American publisher used it to reveal class bias in the federal constitution.2 And while Speed used the method to explore the content of newspapers, a Shakespeare scholar at the University of Nebraska published a manual in 1893 demonstrating how to use content analyses to evaluate trends in literature.3
Speed, however, deserves more than a footnote. Late nineteenthcentury contributions to quantitative social research by European scientists sometimes overshadow Speed and the academic "muckrakers" who followed him.4 The Europeans, however, were approaching communication studies differently; they had begun to consider audiences to be masses of passive readers who could be manipulated by the media. But Speed was the precursor of a research line that tried to explain how audiences, mass mediums, and other institutions interact. The American scholars believed sensation represented a pathology that prevented newspapers from fulfilling their appropriate role in a democratic society.
Sensation was not new to the metropolitan press in 1893,(5) but "new journalism" used it with a frequency made possible by new technology and with an urgency driven by competition for more readers and greater advertising profits. Faster presses and typesetting machinery, along with increased advertising revenues, enabled publishers to print larger newspapers more quickly than in the past. To fill the larger editions and attract readers, news workers were tempted to expand the definition of legitimate news. Contributors to magazines and trade journals critiqued or defended the results while laboring to answer fundamental questions about the press.6 What should a newspaper's relationship be to its readers and other social institutions? Could news be defined? Did the "new journalism" benefit or harm readers? This article examines how Speed attempted to answer similar questions, where his study fits in the context of the 1893 news debate, and how it furnished a foundation for Progressive Era researchers.
Charles A. Dana, the "pope" of New York newspaperdom, called the technological developments employed in the 1880s end 1890s to transform newspaper size and content a "revolution" of "a very radical and remarkable nature. …