Add Slander and Stir: Reprinting of Anti-Labor Stories in Early-Twentieth-Century American Newspapers

In the opening years of the twentieth century, the National Association of Manufacturers (NAM), an organization of businessmen vehemently opposed to labor unions, spearheaded a multifrontal and multi-industry campaign aiming to stop the rapid growth of labor unions. The NAM itself envisioned publicity and

propaganda - or a "campaign of education," as the NAM liked to call it - as a key aspect of this campaign. Newspapers, in turn, were to play a crucial role in the campaign. As the head of the publicity agency the NAM hired said, the "infantry" of the whole campaign consisted of "the constant repetition of small paragraphs or short articles throughout the newspapers of the country.”

The publicity bureau promised that getting this “infantry” into the field was going to be smooth sailing, claiming that "it is easily possible to bring together the vast majority of papers all over the country in a kind of syndicate form." With this, the bureau referred to syndication of the type now familiar to us, but also something specific to the early 20th century: boilerplate and readyprint. Because typesetting was still very expensive, large numbers of small country newspapers got part of their paper either preprinted (readyprint) or as boilerplate (stories that were already typeset and could be directly printed).

Archival sources show that the NAM, besides hiring a publicity bureau to badmouth labor unions, also created "newsworthy" events that it hoped to get into the press without direct payment: these included, for example, staged anti-union parades and Astroturf antiunion worker organizations. What we do not know with much certainty, however, is to what extent these efforts were successful.

This poster reports on a preliminary investigation into the extent to which stories reproducing the language or viewpoint of anti-labor organized employers - a language that is quite recognizable if one is familiar with it - were reprinted in early-twentieth-century American newspapers. It draws on the Chronicling America database of newspapers and makes use of a reprint detection algorithm developed for the Viral Texts project. Eventually the aim is to create as comprehensive an account of such reprinted texts as allowed by the geographical and other limitations of the Chronicling America collection.

Sources and methodology

The dataset used in this investigation consists of the articles (or rather, pages) published between 1897 through 1908 that matched the query (worker OR workers OR workmen OR workman) AND (union) in the Chronicling America collection. The time period selected reflects the years surrounding the launching and high point of the employer open shop campaign, including the NAM's hiring of the publicity bureau noted above.

The identification of reprints is performed using the passim tool developed by David Smith for the Viral Texts project with the default settings (which were developed on Chronicling America material). The results are formatted into an HTML form for easier manual checking; as the algorithm inevitably identifies irrelevant reprints, the most laborious stage is the culling of the set of identified repeated passages.

The early-twentieth-century employer anti-union campaign had a recognizable vocabulary and set of themes, and it is the author's expertise, acquired developed through research on these employer campaigns, that is the main “method” for recognizing stories potentially put out by employers.

Results

As of current writing, I have processed 1906, 1907, and, partly, 1908. For these years I have run the passim algorithm and manually examined the resulting identified reprinted clusters, which has turned out to be even more laborious than anticipated.

On the whole, I expected to see more stories clearly planted by NAM or its allies in readyprint or boilerplate. There are a number of candidates, but these are both less conclusive and have fewer reprints than I anticipated. There are, however, a few very widely reprinted clearly paid-for advertisements excoriating unions.

In some ways, the partly “negative” results in themselves are rather interesting (they seem not to be at

least entirely due to any failings of the algorithm -widely printed items like speeches by the President are also prominent among the reprints). They seem to indicate that despite publicity bureaus' interest in claiming that material could be easily and discreetly planted, the task of shaping public opinion was more challenging that the NAM perhaps anticipated. The very fact that readyprint and boilerplate had to be palatable to a broad cross-section of papers, in an environment where there still were many labor or proworker papers, meant that “syndicated” material needed to be bland and apolitical to sell. Perhaps the democratic role of the press was better protected than expected?

Bibliography

Ipsum, L. (2017) “Lorem ipsum dolor sit amet, consectetur adipiscing elit, sed do eiusmod tempor incididunt ut labore et dolore magna aliqua”. Lorem Ipsum Quarterly. 13.1: 27-45

Lorem, I. (2014) “Lorem ipsum dolor sit amet, consectetur adipiscing elit, sed do eiusmod tempor incididunt ut labore et dolore magna aliqua". Lorem Ipsum Quarterly. 7.1: 46-55

Amet, C. (1887) “Lorem ipsum dolor sit amet, consectetur adipiscing elit, sed do eiusmod tempor incididunt ut labore et dolore magna aliqua”. Lorem Ipsum Quarterly. 3.1: 56-71