[I’ve been invited to speak at an expert seminar prior to the launch of Welsh Newspapers Online at the Pierhead in Cardiff today (follow along on Twitter: #papur). When asked to give the paper I was thinking about genre quite a bit, particularly the print genre of the newspaper and the emerging (but surprisingly stable) genre of the digital archive. As I’ve argued in previous posts, I think genre is a key term for understanding how media operate and this applies to both historical and newspapers and the resources that provide access to them today. I was also thinking a bit about the aura of old newspapers in hard copy and what happens when this is remediated digitally. Like many bloggers interested in the nineteenth-century media, I have been sent some papers by Thomas Walker on behalf of Historic Newspapers. Bob Nicholson has written a great post about the pleasures to be had when confronted with original papers. The success of large digital archives of historical newspapers testifies to the pleasures that lots of people find in researching the past online, too. I’m interested in the extent to which these remediated newspapers reproduce aspects of their materiality in print, particularly with regards to their seriality. People often remark on those facets of materiality that digital resources cannot reproduce (smell, texture etc); but what about those aspects that are also missing in print? No newspaper really exists in isolation and, as serials, are always embedded in a complex dynamic in which they represent, partially, their predecessors while asserting that there will be more issues to come. It’s not about recovering this – and I’m not making an argument about some sort of deficiency in the resources we have. Rather, I think we (and by that I mean scholars of the media mainly) need to take seriality (and so genre) seriously as an object of study. Simulation plays a part here, but so might more abstract models too.]

Digitization returns newspapers to us, but differently. What I want to focus on today is how we might leverage this difference to help us learn more about how newspapers actually work. The benefits of searchable newspaper archives to historical research are, I think, well understood. In fact, if anything I think we take them for granted, with the considerable work that goes into creating archives such as Welsh Newspapers Online often overlooked. Yet I think that because newspapers are so familiar to us, we assume that reading newspapers from the past is just the same as reading them in the present. This afternoon, I’m going to say a little bit about old newspapers. What I want to know is what happens to these objects, so oriented towards a present that has passed, when they are digitized and rendered searchable according to the words they contain. My argument is that users of historical digital newspaper archives treat what they find online much as they would had they read it in a current issue of the paper. This is not to say that users naively assume the content is somehow up-to-date or mistake it for the news; rather, users apply their existing interpretive practices, honed through a lifelong engagement with the news, uncritically to the news from the past. In doing so, users overlook the formal and the repetitive, the features that allow the news to both identify the new and novel while reconciling it with what has gone before

Reading newspapers from the past is pleasurable as it returns us to a moment when the news was new. The contingency of events, before they have become accommodated into broader, over-arching narratives, allows us to see them differently, to see them as new once again. However, newspapers do not provide direct access to events as they occur, but carefully situate them within interpretive frameworks that turn novelty to their own ends. For instance, in order that readers can make sense of the new, newspapers present it as an example of something that they are already familiar with. The news is always situated ideologically, told in such a way that the disruptive potential of newsness is mitigated. In this way, newness becomes about difference, with the latest turn of events becoming another example of what has gone before. However, the mediation of novelty must be carefully judged. Newspapers are commodities that sell timely information and so the newness of their contents is both an important part of their identity as a genre and an important way in which publications compete against one another. And, of course, the affective shock of novelty has long provided a welcome spur to sales.

One of the paradoxes of the news is that we’ve heard much of it before. Newspapers are testament to the predictability of events. They deal in a cast of characters already known to readers and draw from a known repertoire of scenarios; they allocate their contents into a set of predetermined sections, each attributed a regular place in the issue in advance; they are constituted from a set of pre-existing textual genres, many of which also appear in other forms of publication; and the newspaper itself is a recognizable genre, easily identifiable both by the way it looks and what it contains. The appeal of the newspaper is predicated on the promise of novelty, but the newspaper itself consists of a complex set of recurring forms. Repetition is key to the way newspapers opertate and, because they are serials, is encoded into their DNA. All serials create a sort of contract between publisher and reader: publishers attempt to anticipate the demands of their readers by giving them more of what they have already demonstrated they want; readers repeatedly spend their money on the understanding that they will not be disappointed. Each issue of a newspaper attempts to narrate the present, assimilating events into a set of pre-existing structures that assert the individual publication’s identity. In other words, the single issue does not exist in isolation, but re-presents, through its form, the issues that precede it. Newspapers might be oriented towards the present and contemporary, but they do so by establishing continuity with the past and promising it into the future.

It is repetition that permits mediation, allowing readers to differentiate between the formal features that delineate the mediating object and those that mark the content. As readers pick out the content, the repetitive formal framework that sustains it fades into the background. This dynamic, where content is identified as what changes while form becomes what stays the same, underpins the logic of print. Although the newspaper has a history that dates almost to the invention of movable type, it is most associated with industrial print culture, its focus on contemporaneity and the commodity identifying it as the medium of modernity. Readers learn to subjugate form to content every time they read; but this process is particularly important for types of content that appear to be unmediated such as the news. Reading makes form become invisible, enabling content to appear distinct from whatever it becomes marked, and then forgotten, as its media.

Digital newspaper archives foreground a different type of reading. Most users want to find articles about something and most newspaper archives conceive of their content as a repository of information, waiting to be found. The rhetoric of the digital newspaper archive is that by using this particular interface the user will be able to find the article they want, even if they do not know what, exactly, this might be. This is search exploiting the differences within processable text strings to find the one particular article as opposed to all the others. Given that the newspaper archive is characterized by its fragmentation and abundance, subjecting it to some sort of order and then making this accessible to a wide range of users is some achievement. Nevertheless, most resources are based on two sets of elisions: they encourage users to overlook their own mediating role, presenting themselves as transparent gateways rather than publications in their own right; and they circumvent the various levels of structure that situate the article within the newspaper. These elisions are encoded features: deliberate attempts to reproduce the effect of a particular type of reading focused on the verbal information contained within articles. However, different uses produce different objects, so by changing the way we interact with the configurations of hardware and software that produce such effects, we can generate different representations of the newspaper. Rather than use processing power to differentiate between articles in order to find the ‘correct’ one, we might use it to map the generic and the repetitive.

One of the very powerful things that newspaper archives do is reveal the connections between different publications. Searching for a particular figure or event will reveal a range of perspectives. It will also, probably, reveal some surprising treatment of the subject in other parts of the newspaper: advertising tie-ins; correspondence; satirical references etc. Computers, in other words, are good at making visible repetition and, because they are not so susceptible to fatigue and, dare I say it, boredom, they can detect patterns over datasets that are too large to work on by people alone. The genre of the digital newspaper archive, however, is oriented towards the specific rather than the repetitive. I think this is right. One of the interesting things that the digitisation of the press has revealed is the appetite amongst all sorts of people for this material and digital newspaper archives rightly devote processing power to detecting significant differences between the verbal content within articles or that appended as metadata so that users can find the articles they want. I am certainly not criticizing existing newspaper archives here, nor the way they are designed. For scholars, the gains offered by searchable transcripts – challenges presented by ocr notwithstanding – are tremendous. The bibliographical control offered by searchable transcripts alone is sufficient, I think, to redefine historical research in many areas, my own period in particular. In addition, the use of page images as reading texts – a way of compensating for errors in the transcripts – means that it has never been easier to see what historical newspapers looked like, making available many of the repetitive features that I have been discussing. Instead, what I want to suggest is that some scholars – especially those working in my area, media history – need to start thinking more seriously about form and genre, both of the digitized material and the way in which it is digitized. One of the good things about the process of digitization is that it forces us to rethink what we thought we knew about newspapers. One of the things that stayed with me from working on a newspaper digitization project, the Nineteenth-Century Serials Edition, was watching experts question the extent to which they really understood their material. The radical transformation that digitization entails means that almost everything about the source material must be defined and modelled in digital form. I do not want scholars to stop using digital resources to access content – that is, after all, what they are designed to do – but instead start to think about how they might be used to study form. If questions of form and genre, so integral to the way the press operates, do not seem that important, that is because of the history of the disciplines as they stand, and of reading more broadly, remain grounded in print. The digitized press gives us the tools, intellectual and technical, to look differently at print and the culture we have built with it. It also furnishes data that can reveal how the news, however it may be mediated, depends upon what we have seen before.