Tone Merete Bruvik, Aksis - The Research Group for Text Technology, UNIFOB - University of Bergen
TEI Members Meeting, Victoria, Canada
27 - 28 October 2006
When studying the development of a city, various sources like maps, protocols, census and other textual sources
might be linked together in order to build a graphical view of the city over time. In the project Urban Landscape
historians at the University of Bergen do this, using a map from 1888 and some protocols from the 17th and 18th century
over building sites in Bergen. The aim is to reconstruct a map of the town maybe 200 years before the first good
property map of the town was made.
This is implemented as a GIS (Geographic Information System), with links to text sources encoded using TEI P5.
The project is led by Arne Solli and Geir Atle Ersland, Institute of History, University of Bergen. The encoding in TEI is done by Tone Merete Bruvik, The Research Group for Text Technology - Aksis (The Department of Culture, Language and Information Technology), Unifob - University of Bergen.
Project page at Urban Landscape (Norwegian only).
Project period: 2005-2007
We have a complete survey of every plot in the town in 1686. About 55% (Ersland 1989) belonged to the Crown, and paid an annual plot rent to the "Manufakturhuset". And to keep records for these rents, protocols were set up, of which we have two.
The sites in this first protocol do not have a proper address system, but we know the part of the town where they are situated, and we know the size of about 1/4 of the plots.
The 1753 survey is in its core a copy of the 1686 survey, but with new entries as late as the 1770s. This survey has the address system that was implemented in Bergen towards the end of the 18th century. Using this together with the information about the size of the plot, makes it posible to map most of the site in the protocol to the 1888 map.
The protocols are encoded in TEI P5, with the modules analysis, core, header, msdescription, namesdates, tei, textcrit, textstructure, transcr.
My view is that documents should be encoded and structured as much as possible as they are and in it own right, and not with a particular use in mind.
The protocols might look like a table, but I found it more convinient to encode each entry of the protocols with a div element, where the list of the years the rent was to be paid in a child div element, and the description of the plot in an other child div element.
Person names and building site sizes are to be explicitly encoded.
Some paragraphs in the text contains more than one plot:
Giertrud Michels \Jurgen Laußen/ hauge, bred paa vestre ende 34 1/2 al:,
og østre ende 20 1/2 al:, lang paa søndre side 33 1/2 al:,
og nordre side 29 al:. Noch it lidet støche grund paa det sydoest
hiørne, lang 13 1/2 al: og bred 7 1/2 al:. Endnu et støche grund,
lang 16 all: og bred 8 1/2 al:, effter grundebref dat: 1628, gifuer 1 ort 12 ß
English: Giertrud Michels \Jurgen Laußen/ plot, wide at the western side 34 1/2 al:,
and on the eastern side 20 1/2 al:, long on to the south side 33 1/2 al:,
and to the north side 29 al:. Another small plot on the south eastern
corner, long 13 1/2 al: and wide 7 1/2 al:. Yet an other piece of land,
long 16 all: and wide 8 1/2 al:, measured in plot letter: 1628, gives 1 ort 12 ß
Giertrud Michels \Jurgen Laußen/ plot, wide at the western side 34 1/2 al:,
and on the eastern side 20 1/2 al:, long on to the south side 33 1/2 al:,
and to the north side 29 al:. Another small plot on the south eastern
corner, long 13 1/2 al: and wide 7 1/2 al:.
Another small plot on the south eastern
corner, long 13 1/2 al: and wide 7 1/2 al:..
Yet an other piece of land,
long 16 all: and wide 8 1/2 al:, measured in plot letter: 1628, gives 1 ort 12 ß ,
measured in plot letter: 1628, gives 1 ort 12 ß
The element s (section) is here used to divide the different plots from each other.
The protocols can not only be a source in this project, but in other studies as for instance language studies, genealogy, etc.
Our aim is to make a critical edition of the protocols online, where we do not normalize the spelling nor do we expand the abbreviations.
The protocols are online at Urban Landscape:
We know that building sites are very stable over time in general, and this is also the case in Bergen (see Ersland and Solli 2005).
We like to add information from written sources before 1888 on building sites in Bergen to investigate the development of the city, the history of each site and the social development of the city, as well as for instance photographs of the same plots.
Bergis is a way of addressing and mapping large amount of written, pictorial and archaeological data collected with a time span from medieval to present day.
The methods which have been developed and applied are thought to be applicable to other towns with similar sources. The GIS-format makes it possible to create maps or layers describing certain stages in the development of the town, and date addressed in this layer can be used for any demographic, economic or other kind of analyses.
A second use will be to investigate one or a few properties through time, maybe it will be a favourite pastime for house owners and others but this is a byproduct.