Spatial History of the Encyclopédie

Katherine McDonough, Ludovic Moncla, Matje van de Camp

What’s happening here?

This site shares work-in-progress on space and place in Diderot and d’Alembert’s Encyclopédie. We are a research collective experimenting on geographic information retrieval in reference texts from early modern France. We are one historian, one computer scientist, and one computational linguist.

page image

Our work focuses on using new methods in Named Entity Recognition (NER) to interpret the spatial horizons of the Encyclopédie ou Dictionnaire raisonné des sciences, des arts et des métiers, par une Société de Gens de lettres (First Paris edition 1751-1772, in 17 volumes of text and 11 volumes of plates). There are 44,632 total text entries and 14,445 entries classified by the original editors as Géographie.

Initially we focused on developing a method for annotating the geographical information in the Encyclopédie. This work resulted in the GeoViz platform (created by Matje van de Camp). This annotation interface (not yet available publicly) permits researchers to view and edit geoparser XML output. Editing features include on-the-fly lookups of Geonames and DBpedia records matching place names in the text, differentiation of location types (real, historical, mythical/literary, sacred, extraterrestrial) and location positions relative to an article headword (“Paris is contained within France”).

GeoViz image

We used GeoViz to complete a set of gold standard annotations that helped to evaluate NER systems, since there is little to no information on the performance of these tools on early modern French texts. With this process now complete, we are both immersed in the texts of the Encyclopédie geography articles and aware of the current limitations of natural language processing for historical contexts. Foremost among these is the underlying assumption in NER that place names are entirely distinct from other entity types (people, for example) and that single- or multi-word place names can be extracted from their context without losing substantial meaning.

We wish to challenge these assumptions. In doing so we not only reshape the goals of automatic geographic information retrieval, we also experience historical geographical texts in a new environment.

Using a customized version of the NER and georesolution tool PERDIDO, we set out to answer questions like:

Early Modern Geographical Information

“Questions of spatiality – the geography of and in things, be they places, peoples, books, ideas or representational procedures such as writing and mapping – are, like questions of temporality and embodiment, ontological fixtures. What may matter more in understanding questions of geography in relation to the eighteenth century is the form taken by geography the subject – geography as the practices by which the world was encountered and inscribed in word and image and made real to its audiences (and the relationships between these themes and the explanation of them). We can then also address the interpretative significance that comes with thinking geographically about intellectual history and the lived experience of the eighteenth century.”1

This 2011 quotation from Charles Withers and Robert Mayhew captures two sides of the coin that often structure historical inquiries into geography - the geographical distribution of human activity and the invention of geographic science. Research often picks one or the other. In fact, as Withers and Mayhew suggest, these two approaches inform each other. Considering the ways that geography was publicized in printed media also requires understanding how knowledge was pieced together to make up these works. Our project examines the ways that space (as in, not only geographical places) is present in geographical texts, or, if you will, the spatiality of geographic writing during the early modern period.

Extended Named Entities

In this stage of the project, we apply the concept of Extended Named Entities (Moncla 2017) to collect contextual information about places named in the Encyclopédie. This development in Named Entity Recognition allows us to capture place-related information such as toponyms 1) embedded within other types of entities (official titles, institutions, objects, events), or 2) details extending from a named place that characterize or re-define a place (“la ville de Paris” vs. “une petite ville de France”).

The examples of Extended Named Entities (ENE) below demonstrate our preliminary work to visualize first-order embedded entities. Before more complex plans to capture adjacent verbs, adverbs, and adjectives, this fairly simple technique already provides new insights into the functions of geographical information in the Encyclopédie.

Examples

place or person extended named entity place or person (embedded) named entity unclassified entity

The examples below reflect variations on the different types of Geography-classified articles in the Encyclopédie. - Solid fill green or red boxes represent place or person entities. - Green and red outline boxes represent extensions of entities (which then become known as embedded entities). - Boxes with a black dotted line represent entities that have not been classified. NB: These results are preliminary.

1 The Little Town in France

Denat

DENAT 43.84656 2.20602 43.84656 2.20602 43.84656 2.20602 , petite ville de France 14.60365 -61.07418 46.0 2.0 46.0 2.0 au diocèse d' Alby 59.2335 17.8538 59.2335 17.8538 62.49713 15.4745 dans le Languedoc 43.66667 3.16667 44.0 4.0 44.0 4.0 ,sur l Assore 0.36667 9.8 0.36667 9.8 5.97342 -4.27073 ,à trois lieues d' Alby 59.2335 17.8538 59.2335 17.8538 62.49713 15.4745 .

2 Cases in contested attestation history (part 1)

Brassaw

BRASSAW ,ou GRONSTAT ,( Géog . anc . et mod . ) ville forte de Transilvanie 46.5 25.0 46.5 25.0 46.5 25.0 . 44 10 46 30 Les uns la prennent pour la Proetoria augusta de Ptolomée ,et d' autres la nomment Corona 40.74705 -73.86014 40.74705 -73.86014 -9.35802 14.64918 et Stephanopolis .

3 Cases in contested attestation history (part 2)

Zimara

ZIMARA 39.47199 38.34307 ,(Géog.anc.) ville de la grande Arménie 40.25 45.040.25 45.040.25 45.0 ,selon Solin 43.54331 16.4906543.54331 16.4906545.75528 1.27563 ,qui la place au pié du mont Capotes 4.4582 -74.533094.4582 -74.533094.31583 -74.47221 , l' Euphrate 30.57889 47.76572 prend sa source . On lisoit ci - devant dans les exemplaires imprimés de Pline 46.80278 -76.6869446.80278 -76.6869454.48456 23.5427 ,l. V. c. xxiv . Zimyra ,ou Zimira ;mais comme l' a remarqué le P. Hardouin ,c' étoit une faute insigne :car Simyra est une ville de Syrie 35.0 38.035.0 38.035.0 38.0 au bord de la mer Méditerranée 44.02722 1.63559 . La correction que ce savant religieux a faite ,est appuyée sur les meilleurs manuscrits qui lisent Zimara 39.47199 38.34307 . C' est ainsi qu' écrit Ptolomée ,l. V. c. vij . qui marque Zimara 39.47199 38.34307 dans la petite Armenie au bord de l' Euphrate 30.57889 47.76572 ,mais assez loin de la source de ce fleuve . Tout cela s' accorde avec les itinéraires . (D. J. )

4 Classical Urban Types

Nicopolis

NICOPOLIS 43.70255 24.89645 ,(Géog.) ce mot signifie ville de la victoire ,ville fondée à cause de la victoire . Romulus 42.22226 -83.396642.22226 -83.396647.42417 -72.06139 , Bacchus 40.663 -112.0960540.663 -112.0960555.39455 -66.87872 ,et Castor 52.21792 -111.9077652.21792 -111.9077619.42469 -72.51156 bâtirent des villes dans les lieux ils avoient triomphé ,ou établirent des colonies dans les lieux dont ils avoient chassé les anciens habitans ;c' est ce que Pompée , César , Auguste , Titus 45.52118 -71.9848945.52118 -71.9848945.52118 -71.98489 , Trajan 33.46381 75.3466433.46381 75.3466446.63333 27.03333 et autres empereurs imiterent ,en donnant aux villes qu' ils éleverent le nom de Nicopolis 43.70255 24.8964539.00933 20.7335339.00933 20.73353 . C' est pourquoi nous trouvons dans l' histoire plusieurs villes de ce nom . Nous allons tâcher de les distinguer avec exactitude .

5 The Homeland

Nangis

NANGIS 48.55535 3.01306 ,(Géog.) petite ville de France 14.60365 -61.0741846.0 2.046.0 2.0 dans la Brie 48.69247 2.610945.73804 0.2410745.73804 0.24107 , diocèse de Sens 48.2 3.2833348.2 3.2833348.2 3.28333 ,avec titre de marquisat :elle est à 14 lieues de Paris 48.85341 2.348848.85341 2.348848.85341 2.3488 . 20 58 48 33 C' est la patrie de Louis Carré ,fils d' un bon laboureur . Son pere vouloit qu' il fût ecclésiastique ,pour le sauver de l' indigence ,mais il aima mieux tomber dans l' indigence que de se faire ecclésiastique . Le P. Mallebranche le prit pour écrire sous lui ;il devint métaphysicien ,géometre ,et de l' académie des Sciences 33.70283 73.05329 49.5 150.0 . Il a donné le premier corps d' ouvrage qui ait paru sur le calcul intégral ;il est vrai qu' il y commit plusieurs fautes ,mais il les reconnut sans détour . Il mourut en 1711 ,âgé de 48 ans ;il fit l' académie sa légataire universelle ,c' est -à -dire qu' il lui laissa quelques traités qu' il avoit composés sur des sujets de Physique et de Mathématique . (D. J. )

6 Credibility on Display

Montagniac

MONTAGNIAC ,(Géog.) ville considérable d' Asie 29.84064 89.2968829.84064 89.2968829.84064 89.29688 ,en Natolie ,dans la province de Bec -Sangil ,sur la mer de Marmora 44.53342 -77.6661944.53342 -77.6661938.34954 16.27567 40.66667 28.25 . M. Vaillant prétend sur des inscriptions authentiques ,trouvées sur les lieux ,que Montagniac est l' ancienne Apamée 35.13604 36.75437 . Pour se refuser à cette conjecture ,il faut dire que les inscriptions qui l' autorisent ,ont été transportées à Montagniac de quelque endroit voisin . Quoi qu' il en soit ,le golfe sur les bords duquel est bâtie Montagniac ,s' appelloit autrefois Cianus -6.43222 106.50222-6.43222 106.50222-6.43222 106.50222 sinus ,de l' ancienne ville de Cium -6.88091 107.60383 ,dont on voit encore quelques ruines . Par le moyen de ce golfe ,cette ville a commerce avec Constantinople 41.01384 28.9496649.53477 -67.4244449.53477 -67.42444 ,dont elle est à 24 lieues ,et avec Bursa 40.19559 29.0601340.19559 29.0601340.16667 29.08333 ,dont elle est à 5 lieues . 46 30 40 10 (D.J. )

7 Mythical, but also real

Colchide

COLCHIDE 46.26749 7.00516 ,L' ancienne Colchide ,aujourd'hui la Mingrelie ,est au fond de la mer Noire -4.77609 11.86352 ,entre la Circassie ,la Géorgie 41.99998 43.4999 41.99998 43.4999 32.75042 -83.50018 ,et l' Aladulie . . Ce pays passoit autrefois pour être fertile en poisons ;de - vient qu' Horace 39.26255 -85.56414 39.26255 -85.56414 46.98583 -74.72306 parle souvent des poisons de la Colchide 46.26749 7.00516 46.26749 7.00516 46.26749 7.00516 , venena Colcha -18.29623 -66.38259 -18.29623 -66.38259 -17.83333 -66.41667 ou Colchica . Médée 46.20167 -76.13889 ,si fameuse par ses vénéfices ,étoit de la Colchide :en falloit -il davantage pour donner lieu aux fictions de la Poésie ? . Mais ce qui n' est point une fiction poétique ,c' est l' étrange et réelle différence qu' il y a entre la Colchide de nos jours ,et cette Colchide d' autrefois si riche et si peuplée ;différence qui n' a point échappé à l' auteur de l' esprit des lois . . and ;A voir ,dit -il ,liv . XXI . ch . v. aujourd'hui la Colchide 46.26749 7.00516 46.26749 7.00516 46.26749 7.00516 ,qui n' est plus qu' une vaste forêt , le peuple qui diminue tous les jours ne défend sa liberté que pour se vendre en détail aux Turcs et aux Persans ;on ne diroit jamais que cette contrée eût été du tems des Romains 47.14972 0.17418 pleine de villes le commerce appelloit toutes les nations du monde :on n' en trouve aucun monument dans le pays ;il n' y en a de traces que dans Pline 46.80278 -76.68694 46.80278 -76.68694 46.80278 -76.68694 et Strabon and ;. . Art . de M. le Chevalier de Jaucourt .


Recent and forthcoming publications

(under review) Katherine McDonough, Ludovic Moncla, and Matje van de Camp, “Named Entity Recognition Goes to Old Regime France: Geographic Text Analysis for Early Modern French Corpora,” International Journal of Geographical Information Science, Special Issue: Spatial Computing for the Digital Humanities, ed. by Patricia Murrieta-Flores and Bruno Martins.

Katherine McDonough, “Putting the Eighteenth Century on the Map: A Proposal for Early Modern French Geospatial Data Development,” in Digitizing Enlightenment, ed. by Glenn Roe and Simon Burrows (Oxford Studies in the Enlightenment, 2019).

Katherine McDonough and Matje van de Camp, “Mapping the Encyclopédie: Geographic Text Analysis for the French Enlightenment,” Proceedings: ACM SIGSPATIAL Workshop on Geospatial Humanities, 7 November 2017.

Cite this work

McDonough, K., Moncla, L., & van de Camp, M. (2018). Spatial History of the Encyclopédie. Retrieved [insert the date you retrieved the information here without brackets] from http://kmcdono.com/enc.

Acknowledgements

None of this work would be possible with the assistance of ARTFL & the Textual Optics Lab at the University of Chicago. Thank you!

Contact us

Contact Katherine McDonough by email
Contact Ludovic Moncla by email

Footnotes

  1. Charles Withers and Robert Mayhew, “Geography: Space, Place and Intellectual History in the Eighteenth Century,” Journal for Eighteenth-Century Studies 34, no. 4 (2011): 449.