Logo DBNL
TS. Tijdschrift voor tijdschriftstudies. Jaargang 2015 (2015)

TS. Tijdschrift voor tijdschriftstudies. Jaargang 2015

(2015)– [tijdschrift] TS–rechtenstatus Auteursrechtelijk beschermd

Vorige Volgende

Digital Gatekeeper of the Past: Delpher and the Emergence of the Press in the Dutch Golden Age
Michiel van Groesen

m.van.groesen@hum.leidenuniv.nl

Abstract

Launched in 2013, Delpher has developed into arguably the most successful digitization project in the humanities in the Netherlands. However, success comes at a price. Because of the seemingly exemplary methodology of making historical newspapers openly accessible, Delpher has inadvertently encouraged several important misconceptions by scholars and general users alike. Few users realize that the database is not exhaustive, and that some of the choices underlying Delpher distort our current understanding of the role of printed newspapers. Common misconceptions are not actively discouraged by Delpher, and this should change. Based on the digitization of two seventeenth-century newspapers - both ‘pioneers of the Dutch press’ - this essay argues that Delpher is a database that should be used with care. Although it opens new avenues of research, it obscures potentially more fruitful strategies of gaining a better understanding of the emergence of the printed newspaper in the Netherlands.

Keywords

newspapers, Delpher, digital humanities, Dutch Golden Age

 

Before the days of social media - in fact not so very long ago - most journalists operating in a free democracy like the Netherlands considered themselves to be the ‘gatekeepers’ of news and information for the public.1 ‘The news of the day as it reaches the newspaper office is an incredible medley of fact, propaganda, rumor, suspicion, clues, hopes and fears’, observed the American newspaper editor Walter Lippmann in his book Liberty and the News (1920),

and the task of selecting and ordering that news is one of the truly sacred and priestly offices in a democracy. For the newspaper is in all its literalness the bible
of democracy, the book out of which a people determines its conduct. It is the only serious book most people read. It is the only book they read every day. Now the power to determine each day what shall seem important and what shall be neglected is a power unlike any that has been exercised since the Pope lost his hold on the secular mind.2

The gatekeeper metaphor has now lost much of its appeal, and is unlikely to reappear any time soon in the context of the on-going information revolution. In this essay, I will attempt to analyse how it continues to be relevant in our understanding of the past, or more precisely in the way that heritage institutions are using digitization to select and order sources from the past and determine what shall be deemed important enough and what shall be neglected. The case that I will be examining is the corpus of early seventeenth-century newspapers in the Delpher database that was built, and is still maintained, by the National Library of the Netherlands (KB) in The Hague, and to which this special issue of TS is devoted.3 Although my tone may occasionally be sharp, let me begin by stating that this essay is by no means intended to be overly critical of Delpher. I have used the newspaper database to my great benefit for both research and teaching, I know many colleagues who have done the same, and I think it is safe to say that there is broad support among scholars in the humanities (and occasionally beyond) for the observation that Delpher is already, only two years after its launch, an indispensable tool for the study of political culture in the Netherlands in the last four centuries.

For specialists of the seventeenth century, too, there is good reason to be cheerful. Since the foundational work done by the Swedish librarian Folke Dahl during World War II, presented to the KB in 1946 as Dutch Corantos, 1618-1650: A Bibliography, newspapers have long been ignored as sources for understanding the relative freedom of expression in the Dutch Golden Age.4 In the early 1960s, press historian Dirk H. Couvée wrote an important article on the flow of news in the first half of the seventeenth century which explicitly built on the work by Dahl, and which laid out the interpretative possibilities of the earliest newspapers.5 In more recent years, scholars like Craig Harline, Otto Lankhorst, Henk Borst, and Roeland Harms have examined printed newspapers from various perspectives to assess their relative weight in the seventeenth-century Dutch media landscape.6 Nevertheless, the most recent major survey of the Dutch Golden Age,

Jonathan Israel's The Dutch Republic (1995), does not even mention newspapers. Delpher is likely to change all that; it is a change which is long overdue.

In this essay, I will focus on newspapers published between 1618 and 1650, not only to pay homage to the exemplary work done by Dahl, but also because this chronology enables me to study the first two professional journalists in the Dutch Republic, Jan van Hilten and Broer Jansz, during their respective careers in Amsterdam. Both already published a weekly newspaper well before 1620, and both continued their enterprise - without making major changes to the format or style of reporting of the corantos - until the early 1650s.7 Elsewhere I have outlined some of the differences between the two newspapers, but for my purposes here, I will consider them together.8 An important additional reason for studying the newspapers of Van Hilten and Jansz is that Delpher's scholarly advisory board has recently argued that these two titles, their publishers being ‘pioneers of the Dutch press’, merit full digitization regardless of practical restrictions.9 By all accounts, they are the crown jewels of Delpher's early modern material. But after the excitement of the first two years AD (After Delpher), it is perhaps time to take a step back and analyse not only what the database has added to our understanding of the print media in the Dutch Golden Age, but also what it does not reveal or even what it obscures.

The quantitative approach: the survival rate of early modern newspapers

Both Jan van Hilten's Courante uyt Italien, Duytschlandt, &c. and Broer Jansz' Tijdinghen uyt verscheyde Quartieren appeared every Saturday, 52 (or 53) times a year, with only very occasional additions if news was deemed important enough to warrant a special issue. The first surviving issue, and thus the oldest Dutch newspaper still available to researchers, was published by Jan van Hilten on 14 June 1618, but there are indications that this issue is not the first Van Hilten printed. Jansz' first surviving issue (from 10 February 1619) suggests that he too had already been issuing weekly bulletins for some time. Both corantos invariably opened with a section of foreign bulletins, in which the oldest news was listed first. These reports were preceded by date and place of correspondence. Bulletins typically originated in European cities with a significant political function, such as Rome, Venice, Vienna, London, and Antwerp (but not, interestingly, Madrid or Lisbon), or in places where recent military action had taken place, such as the various German battlefields of the Thirty Years' War. This section was followed by information from the sources of the coranto itself, separated from the above by a horizontal line. On the whole, Van Hilten gave more space to exclusive stories,

whereas Jansz focused on continental news from a wider variety of places. Both newspapers were almost certainly read by a wide middle-class audience.



illustratie
Figure 1: Issues per year of early-modern Dutch-language newspapers available on Delpher.


Figure 1 shows the number of newspaper issues available on Delpher for the years between 1618 and 1650; just fewer than 950 issues in total. The number of newspapers available for later centuries of course dwarfs these numbers, when many thousands of issues have survived. For this figure, I made some small adjustments to the official figures mentioned in the Delpher database. Sometimes single issues are included twice on Delpher, because identical copies can still be found in two different collections - and both were digitized. Sometimes unique copies were replicated for no apparent reason. To give just one example, according to Delpher, there are fourteen surviving newspapers for the year 1630. Identical copies of both Jansz' and Van Hilten's issue of Saturday 25 May, however, have been included (and counted twice), so there are in fact only twelve unique issues on Delpher for that calendar year. It is this number that I have used to compose figure 1, which explains why my numbers occasionally differ from those provided by Delpher. This double counting of issues occurs quite frequently: the very first copy from 14 June 1618 for example is included (and counted) twice, while for the editions from 1642, this happens almost systematically. Still, even by counting several issues twice, the ideal number of 104 issues - 52 by both printers - remains well out of sight on Delpher for the entire period under scrutiny here.

Figure 1 reveals the success of the early modern Dutch newspaper. Although the preservation rate is of course unpredictable, the surge of surviving copies in the 1640s reflects the increasing variety in the Amsterdam newspaper market. From the late 1610s until the late 1630s, Van Hilten and Jansz managed to monopolize this market, but by the

early 1640s, perhaps because printing privileges were somewhat loosened, the competition between newspapers became much more fierce. The Ordinarisse Middel-Weeckse Courant (1638), the Ordinaris Dingsdaesche Courant (1639?), and the Europische Courant (1642) now also started to appear in Amsterdam every week.10 In response to the emergence of these new rivals, Van Hilten started producing several editions on the same day. Folke Dahl traced the first such case to 1632, but Van Hilten seems to have increased the number of versions he printed more systematically in 1642 and 1643, a time when the competition became more intense.11 It is important to point out that unlike Jansz, Van Hilten had two presses at his disposal, enabling him to better follow the daily flow of news. His ‘evening’ editions enabled Van Hilten to correct blatant errors and, more importantly, include information that arrived during the printing process. On Delpher, Van Hilten's different editions make the numbers rise even further in comparison to the 1630s.

The story of real interest, however, lies behind the façade of numbers provided by Delpher.12 For the years between 1618 and 1650, Delpher has not included two major collections of early-modern newspapers. When the KB started its digitization offensive in 2004, it first surveyed which copies of old newspapers were still preserved in European (and occasionally American) research libraries. For the seventeenth and eighteenth centuries, René Vos composed a long list of surviving newspapers that served as the starting point for the digitization. Vos used bibliographies of newspapers that already existed, and information he received from various heritage institutions in the Netherlands and abroad. He did his work exceptionally well, and managed to locate many more corantos than Dahl had done - and for a longer period - especially in German institutions like the Niedersächsisches Staatsarchiv in Oldenburg. For the first half of the seventeenth century, Vos agreed with Dahl that three institutions were head and shoulders above the others in terms of numbers of surviving newspapers: the Kungliga Biblioteket in Stockholm (the collection which had inspired Dahl to compose his bibliography in 1946), the Bibliotheque Mazarine in Paris, and of course the KB in The Hague.

If we compare Vos' long list to the contents of the newspaper database for the period between 1618 and 1650, it becomes clear that many hundreds of issues are missing from Delpher.13 The reason for this incompleteness is a mixture of oversight on the part of the digitizing team, and financial prudence. Since 1810, the KB is in the possession of a hefty volume containing more than 500 consecutive newspapers from the period between April 1626 and November 1635. Catalogued as ms. 341 A 1 - a manuscript - it is a unique collection of printed corantos, but in spite of its inclusion in

René Vos' 2006 long list, these newspapers are not included on Delpher. When I first encountered this volume in 2010, in preparation for an article on Atlantic news, I asked the newspaper curators in the KB why they had not digitized the most valuable asset from their own collection of seventeenth-century corantos.14 Their answer was that, somehow, they had overlooked it. The error is currently being corrected as part of the Metamorfoze project, but for now users of Delpher must do without a nearly complete run of Van Hilten's newspapers for the second half of the 1620s and the first half of the 1630s.

Other volumes of newspapers that are missing from Delpher can be found in the Bibliotheque Mazarine in Paris (Rés. 5028 *1-4). They contain even more corantos than volume 341 A 1 in the KB, and are arguably even more unique in terms of what they can tell us about the practice of newspaper production, distribution, and reading. These volumes were also listed by Vos in 2006, and were already known to earlier generations of archivists, but the steering committee of Delpher (quite understandably) refused to pay the exorbitant price demanded by the Parisian library. However, in the 1980s, the Stadsarchief Amsterdam microfilmed these French volumes, and they can be consulted and converted into jpg. files of very readable quality. Like the volume in The Hague missing on Delpher, the Parisian volumes also contain many consecutive newspapers, in this case for the years between 1637 and 1643.15 The added value of this particular collection is not only that it contains well over 400 unique Dutch editions of Jansz' Tijdinghen and (mainly) Van Hilten's Courante - as well as more than two dozen of the earliest newspapers printed by François Lieshout - but also well over 200 French editions of Jansz' Tijdinghen, published under the title Nouvelles de divers Quartiers. The steering committee of Delpher unfortunately decided that early French-language newspapers, even if produced by the ‘pioneers of the Dutch press’, would not be digitized.



illustratie
Figure 2: Early-modern Dutch-language newspapers available via Delpher compared with known surviving issues.


Figure 2 shows the difference between issues of Dutch newspapers that have and have not been included on Delpher. The latter number, including issues that have survived in libraries and archives and have not been digitized, is based on the long list made by René Vos, and further augmented by copies in libraries that have not been contacted by Delpher, or have not responded to their call. The most striking discrepancy can be registered in the year 1638, when uniquely, all 104 printed corantos have survived - 52 from the two leading journalists of Amsterdam, but not a single one of these issues can be accessed on Delpher. In other years, too, the difference is significant. In the Riksarkivet in Stockholm, for example, eighteen additional copies from the year 1626 were consulted and listed by Dahl in the 1940s.16 Whereas Delpher presents scholars with zero newspapers for this particular year, the overlooked volume in The Hague and the additional issues in the Stockholm archive ensure that we have in fact the very healthy number of 57 surviving issues. Most of the institutions that answered Delpher's call for digital reproduction of ‘their’ copies did a good job. The Kungliga Biblioteket in particular provided wonderfully clear photographs of what appear to be well-conserved newspapers. Here too, however, occasional errors were made. Inexplicably, and for

reasons unknown to me, Delpher does not include the 116 Stockholm issues of Van Hilten's Courante for the years between 1643 and 1650 - a strange omission, especially if we take into account that the Stockholm issues of Jansz' Tijdinghen for the same years are available on Delpher.17

The qualitative approach: Delpher and the advancement of scholarship

Arguably more interesting than basic number crunching is the question if the coming of Delpher has enabled historians to pose new questions or finally answer old ones. Here too, two years in, the verdict is mixed. The burgeoning field of early modern news is developing in multiple directions. In the last two decades, historians have gradually abandoned their attacks on Jürgen Habermas' paradigm of the public sphere after establishing that it already existed before the era of the coffee house, be it in different forms across early modern Europe.18 Elsewhere, old habits have proven to be persistent. Research in the last twenty years has somewhat predictably developed along linguistic lines. In order to break this proto-national examination of the emerging press, American historian Brendan Dooley launched the term ‘contemporaneity’ in 2010: the shared realisation across Europe that everyone followed the same stories at more or less the same time. Consequently, Dooley continues, comparing news bulletins is the desired scholarly approach, and he points to cross-boundary transmission and reception of news as an obvious new direction for research.19 Digitization of early newspapers can be an extremely helpful step in the direction of exploring a transnational information culture, and Delpher is leading the way here.

It is important to remember that despite the shortcomings of its newspaper database, the KB is pioneering the open-access offensive in Europe. In Belgium, the Abraham-database contains only newspapers from the nineteenth and twentieth centuries. In France and Germany, several projects are underway, but nowhere near as systematic and ambitious as the Netherlands with Delpher. In the United Kingdom, the Burney collection of newspapers contains only a fraction of the surviving seventeenth-century news publications, and it is not available in open access.20 Only the French Gazette by Théophraste Renaudot has been digitized in a way that comes close to what users of Delpher have come to expect.21 Yet here we arrive at the paradox of Delpher's success. Because it is so far ahead of its counterparts abroad, and because it contains close to 1,000 newspapers for the first half of the seventeenth century, it gives a false

impression of completeness to ordinary and professional users alike. My suggestion, at least for now, would be to include references to all the surviving issues that have not been digitized, in order to serve scholars who otherwise would consider non-digitized issues to be lost. In the context of Dooley's notion of contemporaneity, it is particularly regrettable that it has been decided not to include the French editions of Broer Jansz' Tijdingen, which would have allowed for a first systematic (and hence ground-breaking) comparative analysis.

There is another particular technological problem for digitized seventeenth-century newspapers. The gothic lettering of the early Dutch press cannot be analysed properly by Optical Character Recognition software (OCR), making a full-text search in this part of the database unreliable. This problem is of course well-known. Several letters in particular are often misread by OCR technology - of course the long ‘s’ (mistaken for an ‘f’), but also many of the capitals, and letters that were inked too heavily by the printers. The success percentage of OCR in newspapers with gothic lettering on average lies between 80 and 85 per cent, meaning that in most five-letter words at least one letter is ‘misread’. Delpher claims that many seventeenth-century newspapers are ‘totally illegible’ for OCR-software.22 But this is not the case. A truncated search for, say, the word ‘Brazil’ - using the search terms ‘Braz?’ and several variants - still yields a significant number of useful hits. It is the false impression of a complete search that is most damaging. While a search may yield several newspapers containing bulletins on Brazil, just as many will probably have been missed. Here, too, more explicit notifications of the shortcomings of OCR for seventeenth-century newspapers on Delpher would be desirable.

And finally, no matter how well many of the newspapers have been digitized, nothing beats reading the original version on paper. This is especially true for historians who are interested in the reading practices of the Dutch Golden Age. Physical collections of newspapers bring us closer to readers, not an implied or assumed readership based on a close-up analysis of the process of production, but ‘actual’ readers who touched newspapers, read them, kept them and perhaps consulted them at a later time, and then had them bound in order to create a personal historical record. The fact that reading a digitized newspaper does not come close to reading the real thing is of course not Delpher's fault. However, I fear that digitization might (in due course) facilitate laziness. How many scholars are prepared to go through the trouble of looking up the original sources if they can read digitized versions in their office or at home? Most historians probably still are, but it is imperative that we remind future generations that consulting primary sources in libraries or archives still trumps reading them from a computer screen.

Technology will continue to advance, and perhaps OCR deficiencies can be mended in due course. A crowd sourcing project by the Meertens Institute, with individual volunteers transcribing newspapers with gothic lettering, is already underway. But even then, consulting the original newspapers remains important to address some of

the key questions in the historiography of early modern news culture. The following example will demonstrate my point. Once again I return to ms. 341 A 1 in the KB, the volume that was collected and bound by a single reader, and that erroneously has not yet been included on Delpher. It becomes clear from the (anonymous) reader's collection of corantos, which he assembled between early 1626 and late 1635, that he preferred Jan van Hilten's Courante over Broer Jansz' Tijdinghen. This brings us to the intriguing matter of loyalty to a particular newspaper title - very common today, but never previously the subject of study in early modern times.23 The volume in The Hague begins in the spring of 1626 with sixteen consecutive issues of the Tijdinghen, but then switches to Van Hilten's Courante for the second half of the year. It includes four or five extraordinary issues of Van Hilten, which he printed when he wanted to be the first to bring news of certain events. Perhaps this was one of the selling points that made the reader switch to Van Hilten after initially reading Jansz. The next years, 1627 and 1628, all issues in the collection come from Van Hilten's workshop.

Something fascinating happens in 1629. From January to mid-May, our reader bought Van Hilten's paper as usual, but he switched, by all accounts deliberately, to Jansz' Tijdinghen. The first issue of Jansz' coranto (26 May 1629) contained news that came almost exclusively from 's-Hertogenbosch (Bois-Le-Duc). Troops of stadholder Frederik Hendrik had just started their long-awaited siege of the city after a succession of victories over Spanish troops in Gelderland and Wesel. Broer Jansz, in his imprint, fashioned himself as a ‘former Coranteer in the army of the Prince’, probably a reference to the previous stadholder Maurits: a claim of authority that newspaper readers in the early seventeenth-century United Provinces probably recognized. Indeed, we can witness that our reader continued to read Jansz' Tijdinghen throughout the summer and the fall, until 10 November, when the siege of Den Bosch had ended, and the main media event of the year gradually lost its news value. The last six newspapers of the year 1629 in this collection are once again issues of Van Hilten's Courante. Six years later, in the same volume, the same thing happens again, this time in the context of the siege of Schenkenschans. It appears that the nature and the origins of the news determined which newspaper ‘our’ readers, and probably readers in general, preferred.

Even if the KB had digitized this spectacular volume of newspapers, it would still not have been possible - based on the information currently given by Delpher - to draw these kinds of conclusions from a set of jpgs or pdfs. Patterns can occasionally be deduced from Delpher: for example, when studying the rhythm of the surviving newspapers from the Kungliga Biblioteket in Stockholm for the year 1642, a clear pattern emerges. It appears that both newspapers of Amsterdam only made it to Sweden every fortnight. Both Van Hilten's and Jansz' issues of 1 and 15 February survive in the collection, but none are available for the Saturday in between. The same is true for both newspapers throughout the year. Sometimes consecutive issues have survived, but always in pairs - Van Hilten and Jansz. Back in 2003, Marika Keblusek established that the Swedish agent in Amsterdam Michel Le Blon was responsible for sending news (and printed

newspapers) to Axel Oxenstierna at the court in Stockholm.24 It would be fascinating - but unfortunately beyond the scope of this article - to determine why he sent newspapers in certain weeks but not in others. This would in theory be possible with the data provided by Delpher, although consulting the original sources would be preferable here too. Also, for a reliable analysis, the database would need to be supplemented with the Van Hilten issues from Stockholm for the years between 1643 and 1650.

Conclusion

The coming of Delpher is a wonderful gift, and a very useful addition to the tool kit of historians of political culture of the Netherlands, including those who study the early modern Dutch Republic. It will enable scholars to ask new questions, although Delpher will not be able to provide answers to all of them. Most importantly, the significance of the early press will surely be recognized in new surveys of the Dutch Golden Age, thus generating even more attention from professionals and amateurs for seventeenth-century corantos. Nevertheless, scholars should be careful in delving the database. Using Delpher requires the same methodological precision as using the print collections of archives and libraries. It remains advisable for all researchers interested in seventeenth-century newspapers to study them in print as well as in their digitalized form. The haphazard blessings of OCR technology are an important reason for caution, as are some of the choices made by Delpher, such as the choice to exclude French-language newspapers printed in Amsterdam before 1650.

It is imperative to alert users to the fact that many remaining corantos for the first half of the seventeenth century are not available via Delpher. Van Hilten's Courante is the most obvious case: in the period discussed here, 199 copies of his newspaper can be found using Delpher, while more than 750 (!) surviving issues have not been digitized. In this sense, Delpher operates as a digital gatekeeper, and not in the ‘truly sacred and priestly’ sense that Walter Lippmann meant when he coined the metaphor. Delpher gives the impression of exhaustiveness by not explicitly mentioning those surviving newspapers that have not been included. Even the online publication of the long list composed by René Vos in 2006 alone would enable scholars to establish for themselves how many surviving issues have not been digitized. According to my own comparison of the lists by Vos and Dahl with the contents of Delpher, around 52 per cent of all surviving newspapers from the period between 1618 and 1650 have been ‘lost in digitization’, a number that is too high to disregard amidst the justified applause for Delpher's many merits.

•> MICHIEL VAN GROESEN is Professor of Maritime History at Leiden University. For his NWO-Vidi project Covering the Ocean, he studies Atlantic news in printed newspapers of the early modern Low Countries.

Vorige Volgende

Over dit hoofdstuk/artikel

auteurs