BKAS ESSAY SHORTS No. 2 (April 2018)

Preservation of Newspapers: The Vulnerability of Advertising

Peter Robinson (Japan Women's University)

Newspapers remain a critical source for the study of book advertising. This fact alone justifies consideration of the state of newspaper preservation globally, and reflection on some of the difficulties the researcher may encounter in using such material. 

In 2001, Nicholson Baker launched his now infamous "assault" on libraries and librarians¹ in a contentious, but much-cited attack on what he perceived to be their scandalously cavalier attitude towards the very items they were entrusted to preserve, namely books, periodicals and newspapers. Initially as a shocked observer, but later a leading newspaper activist and preservationist, Baker embraced the mantle of whistler-blower, sounding the alarm on a process of willful destruction that had commenced quietly two decades before behind the closed and inscrutable doors of America’s great knowledge sinks, but which was gathering pace and boldness. At stake was the fate of billions of words, photographs, illustrations, and data that we, collectively, thought was being kept safe for future generations. Since the mid-1970s, inside the quiet, insignificant storage rooms, and on the miles of bookshelves lining hallowed institutions, a scandal of monstrous proportions had been brewing: a sickening abuse of trust, a betrayal born of the fateful mix of blind zeal for technology, personal careerism, fund-grabbing, and the most dangerous factor of all, good intentions. The precise importance of each of these plot elements has been in some cases fairly contested, but the narrative (Baker is first and foremost a novelist) was one of conspiratorial proportions, worthy in fact, of the thousands of serialized page-turners that were also being denuded, degraded, and impoverished, by the mania for microfilming that swept through, firstly America’s libraries, but later spread like a contagion across the globe. In peacetime, the application en masse of this war technology –microfilming– had come to be seen as a panacea to the (mostly imagined) space and (mostly genuine) preservation issues surrounding printed matter in general, but owing to their size and vulnerability, specifically newspapers. Further fueling the momentum was a fiscal landscape of budget cuts, combining to form an institutional consensus that this activity was ‘okay’, and in the best interests of the fragile, defenseless papers themselves. 

In 1987, a slick film commissioned by the Library Association, Slow Fires: On the Preservation of the Human Record, directed by the some-time Hollywood producer Terry Sanders, gave the project greater legitimacy and, perhaps more importantly, urgency. According to the 58-minute film, deep within the bowels of the country’s leading institutions, millions of newspaper pages, printed on low-quality, photo-sensitive, highly acidic paper, were primed to self-destruct. It was a race against time to preserve, as the film’s subtitle implied, the "human record", before it was lost forever. That powerful message of fear created the sense of urgency that always accompanies a public scandal in which actions are taken hurriedly and uncritically, and when hindsight determines that one "ought to have known better". A raft of pseudo-science was also used, Baker implies, to secure the funding and attention necessary to support the massive logistics of the filming process. The award-winning film was a propaganda coup worthy of Leni Riefenstahl’s Triumph of the Will, creating a sort of collective frenzy which helped to sell the improbable idea of the near spontaneous self-destruction and "encrumblement" (my word) of millions of newspaper pages under little more than their own weight. A brittleness symbolized by the wildly inappropriate and much maligned "Double Fold" test². There was a conspiratorial silence about the relatively short lifespan of microfilm, the inherent image quality issues, and the highly imperfect user-end experience. What researcher has not cursed out loud at the ocular, to say nothing of the spinal, discomfort of whizzing through reams of microfilm, only to encounter an illegible black and white image?  (See Figures 1a & 1b).

Baker’s central point was that apart from the general myopia of this pell-mell rush to microfilm, hundreds of thousands of newspapers had suffered the ignomy of having their spines guillotined, were poorly photographed, and either outright discarded, or else left to rapidly degrade in their now unsupported and anemic state. Microfilming was accompanied by the fundamental idea of substitution, of the creation of film surrogates that led to the wholesale dumping, burning, and selling-off of a textual and literary source of unparalleled richness and complexity. When viewed as a complement to the real, Baker conceded the process had some merits, but not at the expense of the real. Newspapers were suffering a death by a thousand cuts. In highlighting the actions of libraries and institutional repositories, not only did Baker shine a powerful light on what was going wrong, but spotlighted the beauty, depth, scope and fundamental importance of newspapers as cultural mediators, offering a diversity of content found in no other source: from national "official" papers, to local hand-press newssheets produced by minority groups whose voices were seldom heard in other textual forms. He argued, persuasively, that retaining the original was essential for understanding the context of production, the technologies that were employed, and the very materiality of the paper which profoundly affected readers’ experiences. 

Where We Are Now 

In the twenty or so years since Baker’s well-placed salvo, the outlook for libraries has transformed beyond all recognition. In part as a response to the hoo-ha surrounding Baker’s attack, but also, paradoxically, because fast-shifting technological developments have made some of his critiques less relevant, even obsolete: libraries and archivists have a renewed sense of responsibility towards the items that have been entrusted to their care. Analogous to a long-running trend in medicine, there has been a general movement towards image reproduction practices that emphasize non-intrusiveness and tacitly adopt the Hippocratic dictum primum non nocere ("First, do no harm"). Millions of pages of newspapers have been scanned using flatbed and overhead scanners, employing software that automatically removes gutter shadow and avoids the gruesome (and unpopular) spectacle of systematic dis-binding. Millions more—perhaps the majority of newspapers—are available on vast publically accessible and subscription-based databases that use page images digitally scanned from the very microfilm rolls that beggared the storage vaults of what had been the bibliophile’s dream. Wholesale destruction has been, it seems—for now at least—consigned to the pantheon of humanity’s most shameful activities. Yet losses continue to mount. 

For the user too, access to newspapers has been transformed, with digital repositories allowing simultaneous and instant access to a bedazzling array of content, with full in-text searching through the deployment of Optical Character Recognition (OCR) technologies. In the domain of preservation and storage, the British Library has led the world by decommissioning its decades-old facility at Colindale in November 2013, and replacing it with a purpose-built, climate-controlled storage facility at Boston Spa (Yorkshire, UK), and a Newsroom at St. Pancras. This significant investment is an explicit and belated acknowledgment of the importance of preserving the original, providing "a final resting place for millions of pages where it is hoped optimal humidity, temperature and oxygen levels will prolong their lives". In tandem with this move, however, the British Library introduced a "surrogates first" policy, stipulating that digital or microfilmed copies or "surrogates" must be consulted where available, with the original only to be viewed as a last resort. In the wake of Baker’s campaign to rescue the print heritage of America, the National Endowment for the Humanities (NEH) and the Library of Congress partnered to create the Chronicling America project, a freely available, internet-based searchable database of American newspapers from 1789-c1930. It represents the biggest project of its kind anywhere in the world. In addition to institutional and state-led projects, private companies such as Gale Cengage Learning, Readex, ProQuest, BRILL, and other companies working on providing information targeted towards the genealogist have created huge subscription collections of newspapers. Despite these largely positive developments, new challenges and issues have emerged and are only gradually being addressed. The most obvious problem has been poor image quality (See Figures 1a & 1b), especially of newspapers digitally reproduced from microfilm surrogates (often where the originals have been destroyed or deaccessioned). Few digital newspaper databases have been able to offer pictorial images in anything like useable quality, and while text generally fairs better, OCR software has struggled to cope with inadequate input quality, complex non-English character-based fonts, and advertisements which combine images and artistic typographies in innovative and creative forms. A study of OCR accuracy conducted by consultant Simon Tanner in 2009 using the British Library’s digital newspaper databases, "The 19^th Century Newspaper Project" and the "Burney Collection" recorded a word accuracy rate of only 78% and 65% respectively.³ Indeed, the Library of Congress admitted that its digital exhibition: "Rotogravures of WWI", required the rescanning of the original material because the digital surrogates available on their database did not reflect the quality of the original. 

Several other important issues have also emerged since the Baker campaign. Firstly, there has been considerable interpretive laxity in labeling of newspapers as "surrogates", which are in fact not surrogates. This negligence principally takes two forms: the selection bias in scanning only the "late" or final editions of daily newspapers, and under-appreciation of variant editions. For example, the English tabloid newspaper the Daily Mail produced an Atlantic, Continental, Overseas, and even Braille edition at various points in its history, yet only the standard domestic edition and the Atlantic edition have been digitalized. Not only is the content and issue frequency of these targeted editions significantly different, but also the editorial tone and advertising is tailored to suite particular reading demographics. In 1926, the printers’ strike, which formed strike as part of the General Strike in Great Britain, had a minimal impact on the press because Lord Northcliffe defiantly broke the strike by having his papers printed in Manchester, but its counterpart, the weekly Daily Mail Overseas missed a week and was forced to issue a printed apology to subscribers. The following week a peculiarly sized paper was available. Librarians have been too quick to see what is available digitally, and to dispose of the original copies for what they think are perfect surrogates, but in reality, are wildly different. 

The ubiquitous single and multiple-page advertising wraps which simultaneously, enclosed, protected, and paid for the main body of the majority of nineteenth-century newspapers and some periodicals (see for example, The Manchester Guardian, see Figure 2) are, as David Finkelstein recently reminded us in a lecture on Blackwood’s Magazine https://bkas.org/wp-admin/post.php?post=529&action=edit, especially vulnerable to loss. Often thoughtlessly discarded when being institutionally bound –in the manner that a child throws away the Christmas wrapping paper– or else bear the brunt of enthusiastic readings, suffering from innumerable tears, abrasions, and soiling. 

(Reformatted and reposted 31 January 2023).

¹ Parts of this paper first appeared in the exhibition, "Novelists and Newspapers: The Golden Age of Newspaper Fiction 1900-1939", Komaba Museum, The University of Tokyo, (29 April – 25 June 2017). 

² A paper test designed to test the oscillatory endurance of a piece of paper, typically folded backwards and forwards to the point of failure. 

³ Tanner, S. ‘Measuring the OCR Accuracy Across the British Library’s 2 Million Page Newspaper Archive’, report presentation slides (2009).