The digital study of early modern news culture involves a geographical layer. News were sent from one city to another; they were circulating in different geographical regions and spreading information about remote locations. In order to study the geographical layer of information culture with computing tools, first we need to record and store geographical information: country of the location, geo-coordinates, etc. This however involves a number of challenges. In this post we will outline these challenges and discuss how we attempt to tackle them.
As a first step, the Euronews team transcribes a piece of news (or avviso in Italian) and records the geographical layer of the text. Consider the following news that arrived from London to Florence in 1667:
Giovedì mattina un fuoco si scuoperì a Southwarke suburbo di questa città dell’altro canto della riviera dove havendo consumato venti case, et abruciando sei persone, fu per la grazzia di Dio estinto.
Thursday morning a fire broke out in the Southwarke [Southwark] suburb of this city [London] at the other side of the river where it destroyed twenty houses; six persons were burned, thank God it was extinguished. [1]
At first glance the geographical layer of this news seems to be straightforward: Southwark suburb of London. This is the place that is discussed in the news. The Euronews team therefore records it in the following way.
<newsFrom>
<from date="04/08/1667">Southwark London</from>
<transc>Giovedì mattina un fuoco si scuoperì a Southwarke suburbo di questa città dell'altro canto della riviera dove havendo consumato venti case, et abruciando sei persone, fu per la grazzia di Dio estinto.</transc>
<newsTopic>Accident</newsTopic>
<wordCount/>
<position>5</position>
</newsFrom>
Next, as part of computational post-processing, we need to identify the geo-coordinates of Southwark. This is an important challenge. Even though Southwarke still exists today (it is a suburb of London), we cannot be sure that the geolocation of contemporary Southwarke is the same as it was three hundred years ago. Hence, we cannot just look up the current geolocation of Southwark, and insert it into our database. Furthermore specific resources (such as for instance a geo-database of historical locations) to identify the geo-coordinates of locations from the seventeenth century are scarce. To address this difficulty, we enter the coordinates of the larger geographic unit, i.e. London.
A similar problem arises in cases where a location does not exist anymore. Consider the following example:
S’intende che la nave croce che partì di qua per Constantinopoli sia stata ritenuta, et condotta a Messina da alcune galere di Napoli havendosi trovato [illegible] che andavano ad infideli.
It is understood that the boat that had left from here to Constantinople was detained and taken to Messina by galleys from Naples after they had found out that it was heading towards infidels.[2]
Constantinople was the capital of the Ottoman Empire until 1922. Today it is Istanbul. Hence a computer cannot just look up its geocoordinates in a database of contemporary places. In this case human intervention is needed. A human agent has to check where Constantinople is today and what its geo-coordinates are.
Finally, we would like to mention a last key difficulty related to the identification of historical places. Consider the following example:
Si crede almeno che sia per sortire al Generale la presa di Castelnuovo [Herceg-Novi] e di Risano [Risan] verso Cattaro [Kotor] stando quegl’habitanti con gran timore, e non si sente à quei confini motivo alcuno de’ Turchi, i quali però nel Varadino [Oradea] hanno fatto ultimamente alcune scorrerie con menar via più di mille coppi d’animali e condotti prigioni molti [proposed reading: Mulacchi], come seguiva di quando in quando per andar spargendo per quelle campagne à procacciarsi il vivere coll’uva, e fichi.
It is believed that the General is about to take Castelnuovo [Herceg-Novi] and Risano [Risan] toward Cattaro [Kotor], since the inhabitants are very fearful, and there is no reason to hear of the Turks on those borders, who, however, in Varadino [Oradea] have recently made some incursions, taking away more than a thousand pairs of animals and leading many prisoners [proposed reading: Mulacchi], as occurred from time to time, and then scattering themselves through the countryside to procure a living with grapes and figs.[3]
The news item above mentions a town named Varadino. This is the old Italian name of the town Oradea from Romania. When we store the information about geographical locations, we also try to identify – computationally – the country in which that location was. In the case of Oradea, a database of geographical locations can hardly help. When this news was compiled, Oradea was not part of Romania; in the early modern period it was mostly part of the Principality of Transylvania, a political entity that does not exist anymore.
To sum up, identification of historical places is a difficult task for computers. In fact, it is a task that computers cannot handle alone; humans and computers have to work together.
Footnotes:
[1] Archivio di Stato di Firenze, Mediceo del Principato, Volume 4207
[2] Archivio di Stato di Firenze, Mediceo del Principato, Volume 52635
[3] Archivio di Stato di Firenze, Mediceo del Principato, Volume 3088b
