1) Modeling historical processes and temporality Like the astronomer, the historian is engaged upon the portrayal of time [..] both transpose, compose and color a facsimile which describes the shape of time, George Kubler (1962, 19)

More and more the data of academic and cultural heritage institutions and individual projects are shared as Linked Open Data. The project (2017-2022) for example, aims at developing a sustainable infrastructure to study relations and interactions between 1) the various branches of the cultural industries and 2) between producers and consumers of creative goods across the long Golden Age of the Dutch Republic. The project will link distributed, heterogeneous resources (both existing and new) on the production of the creative industries in the Dutch Golden Age from heritage institutions such as the Rijksmuseum, KB National Library of the Netherlands, The Netherlands Institute for Art History and of academic institutions. Consumption remains an under-investigated topic with regard to the creative industries in the Dutch Golden Age. The digitisation of the enormously rich collection of the notarial acts (more specifically the probate inventories) in the Amsterdam City Archives, will provide data on the possessions of cultural goods by the inhabitants of all layers of society in Amsterdam what was a global city in the 17th Century.

The Golden Agents research infrastructure enables analyses of interactions between various heterogeneous (un)structured datasets by using a combination of semantic web solutions and multi-agent technology supported by ontologies. One of the challenges is the modeling of ontologies for the historical processes of the interactions between various branches, and between the production and consumption of the creative industries of the Dutch Golden Age. These processes are described as multiple narratives for which we use the concept “storifying data”. These multiple stories developed over time in parallel orders, for instance the order in the making of an object (from idea to final product), the order of an object in the artistic life or oeuvre of their maker, the order between the original object and copies and transformations hereof and finally the order of the object within history.

For that reason the problem of representing time in linked data goes beyond the common practice in linked data projects of mapping a historical event in a given place to the right (Georgian, Julian, Chinese etc.) calendars. Important is the PeriodO initiative to create a gazetteer of period definitions. However, it is not sufficient to map vague period names to more precise chronological coordinates as confined events. Before building upon on well established ontologies based on standards such as CIDOC-CRM, FOAF and FFBR that are developed in a certain domain it might be useful to take step back to see whether those are applicable in different contexts as well. Historians often analyse encompassing processes and interactions in networks, rather than focusing on single events or objects (compare Shaw 2010). However, although these standards from the cultural heritage domain support the structuring of the metadata of cultural or bibliographic objects, they are not always suitable for modelling (meta-)data for historical research. To overcome that problem the DataforHistory.org consortium (created on 23/24 November 2017 in Lyon on initiative of Francesco Beretta and George Bruseker) took the initiative to develop ontologies for history that are not only fully compliant with standards such as CIDOC-CRM, but also compliment them. Within that consortium a working group brings project together that concentrate on the modelling of storylines. The group consists of Charles van den Heuvel (Huygens ING/UvA Amsterdam, Veruska Carretta Zamborlini (University of Amsterdam. Vanessa Bartalezi Lenzi and Carlo Menghini (CNRS-Pisa) Alex Butterwoth (University of Sussex), Karl Pinneau (UTCompiegne) and Regina Varnierne-Janssen (Vilnius University). It includes the very interesting Narratives in Digital Libraries project (Bartalezi Lenzi e.a. 2017) that models and connects narrative events in literature, but unfortunately does not allow for representing multiple time-sequences. Promising is the multiple strata (material, cultural, institutional) approach of life cycles of cultural goods that Karl Pineau presented at the 3rd Data for History meeting (Pineau 2019).

The Storyfying data model for the Golden Agents project not only describes multiple storylines of objects but also of ideas (Carretta Zamborlini, Betti, Van den Heuvel, 2017). Our inspiration came from George Kubler’s of 1962. We believe that his approach of temporality in the history of art as linked historical sequences of works instead of descriptions of styles and genres over time is still relevant for modeling ideas expressed by historians about the representation of time and historical processes in ontologies and standards.

2) Kubler’s Shape of Time. Remarks on the History of Things Revisted

The “history of things” is intended to reunite ideas and objects under the rubric of visual forms: the term includes both artifacts and works of arts, both replicas and unique examples, both tools and expressions -in short all materials worked by human hands under the guidance of connected ideas developed in temporal sequence From all these things a shape in time emerges. (Kubler, 9).

The history of things as ideas and objects are crucial concepts in the Golden Agents project as well. Herein, we create so to speak a ‘historical internet of things’ of (im)material cultural objects and events of the Dutch Golden Age. However, Kubler’s description of the history of things as parallel sequential orders is not only relevant for a better understanding of the lifecycles of objects and idea in the Dutch Golden Age. Kubler’s discussion of art history as a “system of formal relations” enables critical reflections on the historiographies of the concept of time of other disciplines as well. In Kubler’s morphological analysis of duration in series and sequence, resonates Braudel’s conceptions of serial history in human agency (micro time (events); meso time (cyclic processes) and ‘longue durée’ (structural change) that dominated the French historiography of the Annales School (Daley 2012). Recently, the theoretical physicist Carlo Rovelli (2018, 103) in The Order of Time argued that there is no need to choose a privileged variable and call it time. It would suffice to have a theory of dynamic relations that tells us how the things we see in the world vary with respect to each other.

While Rovelli is in search of a physical theory to understand the dynamic relations between the earth and the universe, our perception of historical relations within the world might be best captured by graph theory. It is revealing that the only image in Kubler’s art-historical analysis of long before the digital era is a tiny visualisation of a directed network. ( note p. 34; see small insert left in figure 1). In that respect recent timeline tools such as the one developed by Matt Jensen for NewsBLIP might express Kubler’s idea better:

Figure 1: Visualizations from and after Kubler’s statements about the

This paper explores and visualizes the potential of Kubler’s vision of (analog) multidimensional directed networks in which formal relations between ideas and objects build upon each other sequentially over time in a Semantic Web framework. In particular we are interested in Kubler’s concept of indiction that measures multiple distances in historical duration. The measures are drawn from experience and are used as module to connect things with individual lives and the generations of man (Kubler, 105). We use this concept to model the storylines of Artists, Things (artists and ideas) and their Replicas. Although we have to leave the alignment of the model with we here try to visualize our first ideas using the case of the many histories of Rembrandt and his famous .

3) Of Creators and Copyists: Storylines of Artists, Objects and Replica’s

While the Golden Agents project tries to break with the canon of the art history by analyzing the consumption of cultural goods in all layers of society instead of in elite culture, for the modeling the most famous painter of the Dutch Golden Age, Rembrandt and its most famous painting The Nightwatch of 1642 was chosen. We opted for a painter with many pupils and copiists, for a work of art with many multiple sources of ownership, with a rich material history of production, re-use and restoration and with contemporary copies and later derivates in other formats to make a rich model that includes as many stories and perspectives as possible. Rembrandt and his Nightwatch meet those requirements for an inclusive model. The Nightwath changed ownership from the Militia of Frans Banning Cocq to the City of Amsterdam. As an object the Nightwatch was cut in 1715 to move it from its original location from the Kloveniersdoelen to the Townhall of Amsterdam. A drawing, print and painting by the contemporary copyist Gerrit Lundens is needed to understand the original depiction. The painting was overpainted, damaged and restored several times.

Figure 2: Aspects of the material history of the Nightwatch: Detail of damage, etching Claesses after Lundens’ copy and reconstruction that illustrates cut areas of original for its relocation in 1715.

Apart from this material history of the painting, the story of Nightwatch lived on in many other media. It inspired for instance, Peter Greenaway to make a film, Mikhail Dronov and Alexander Taratynov to cast the arcquebusiers in freestanding bronze statues and finally a theater company to bring to live amidtst the shopping public in Amsterdam as a part of a commercial of a Dutch bank.

4) Storylines as input for a future ontological model for historical processes

Figure 3: Related Storylines of Rembrandt and the Nightwatch: Detail sources of evidence of its production and alterations.

The sources of evidence of the very rich story of the production, re-use and restoration of the Nightwatch with its many copies and derivates in other media is just one of the many stories of the history of this painting that allows us to storify data in related, partially overlapping timelines as input for modeling historical processes in knowledge graphs. However, as mentioned above we aim at ontologies than can structure the interactions between various branches, and between the production and consumption of the creative industries of the Dutch Golden Age. In 2017 we presented a poster at the DH Benelux in Utrecht visualizing the potential impact of connecting relations between sources of evidence in the notarial acts and the databases of Ecartico and Onstage that make part of the Golden Agents project. It shows the relation between the painter Emmanuel de Witte and the playwright Joris de Wijse. From notarial deeds we know that the latter was a lodger of the painter and that he was married to Adriana van Heusden. The Ecartico database informs us that De Witte portayed Adriana at the Fishmarket (1667) in Amsterdam, while the OnStage database provides information about the play Voorzigte Dolheit by De Wijse presented in Amsterdam, published by Jan van Hilten and translated by the Spanish playwright Felix Lope de Vega.

Figure 4 Storifying data of notarial deeds of City Archives Amsterdam and Ecartico and Onstage databases in Golden Agents infrastructure with relations between a painter and a playwright in Amsterdam. DH Benelux Utrecht 2017 Poster H Nijboer, C van den Heuvel e.a.

In short, just by combining a few archival sources and metadata in this example we were able tp create an intricate network that informs us about the interconnections between a painter, playwright, printer and translater active in Amsterdam around 1667. By connecting databases about the production and the contents of two millions scans of probate inventories that inform us about the consumption of cultural goods in Amsterdam in linked data we hope to tell multiple stories of Amsterdam in the Dutch Golden Age. However, its potential re-use goes fay beyond that. By modeling and implementing an ontology of parallel sequential orders of time that expresses historical processes, extending but interoperable with the vocabularies of the most commonly used standards in cultural heritage and digital humanities, such as CIDOC-CRM and FRBR, not only the data of the Dutch Golden Agents project but of multiple linked data projects in these domains could be interwoven in the Semantic Web.

Note of the authors and Acknowledgement

This paper reflects work in progress in modeling the ontologies of the Golden Agents project in mulitple storylines presented in public at the workshops Semantics 2017 Events in Amsterdam and the first Data for History.org workshop (Lyon 2017) and published as an abstract for the EADH conference in Galway 2018. We combined this abstract with idea’s presented at the 3rd Data for History.org meeting in Leipzig on the 4th of April 2018. A final publication wil include an implemented model of the ontology of parallel sequential orders of time that expresses historical processes compatible with common standards in cultural heritage and digital humanities. We thank Chiara Piccoli and Bart Reuvekamp (University of Amsterdam) of the Virtual Interiors project to share their information about Gerrit Lundens, copyist of Rembrandt’s Nightwatch.

Citations: Valentina BARTALESI LENZI, Carlo MEGHINI, Daniele METILLI (2017) . In: , vol. 12 (1) pp. 35 - 46. Inderscience Publishers, 2017. George KUBLER (1962) , New Haven and London: Yale University Press Harm NIJBOER, Charles van den HEUVEL e.a. (2017) https://dhbenelux2017.eu/wp-content/uploads/sites/187/2017/06/Abstracts_DHBenelux_Posters.pdf Karl PINEAU (2019) (in collaboration with Data for History org Meeting in Leipzig 4-5 April 2019. Carlo ROVELLI (2018) , New York: Riverhead Books http://perio.do/en/ [last accessed 1 December 2018] Ryan SHAW (2010) , Berkeley: University of California Dale W. TOMICH, (2012) . In . Edited and with an introduction by Richard E. Lee, Albany: State University of New York, 9-34. Gabri van TUSSENBROEK (2018) , Amsterdam, Uitgverij Prometheus. Veruska CARRETTA ZAMBORLINI, Arianna BETTI, Charles van den HEUVEL, SEMANTiCS 2017 workshop proceedings: EVENTS September 11-14, 2017, Amsterdam http://ceur-ws.org/Vol-2063/events-paper1.pdf