[Crm-sig] DigiPal: The Problem of Digital Dating: Online Survey

Øyvind Eide lister at oeide.no
Thu Mar 12 12:43:30 EET 2015

12. mars 2015 kl. 10:31 skrev Arianna.Ciula at roehampton.ac.uk:

> Agreed.
> There are however at least two references in the TEI to issues concerning the translation of verbal (and imprecise) temporal descriptions in the source text to interpretations of numerical values. See http://www.tei-c.org/release/doc/tei-p5-doc/en/html/CO.html#CONADA (in particular the reference to mechanisms connected to the expression of ' Certainty, Precision, and Responsibility') and also the reference to the possibility to link a temporal expression in the document being encoded to an explicit interpretation linking to a features structure mechanism outside of the encoding of that document (see http://www.tei-c.org/release/doc/tei-p5-doc/en/html/ND.html#NDDATER).
> I see another issue concerning the interpretation of numbers into dating and connected to what you say here. For example in the survey there are two expressions: 'circa 1100' and 'circa 1172'. The number 1100 in natural but also palaeographical language might denote a full century and not just the year 1100, while 1172 is most likely used to denote that year only. So 'circa 1100' could arguably be understood to mean '1091-1199' while 'circa 1172' would be taken to denote a narrower interval range, say '1170-1174’.

This is interesting and common across data types. For instance, ‘north-north-east’ will often indicate a higher precision than ‘east’ because one cannot know if the latter is an expression in a system of 16, 8, 4, or 2; whereas the former is signalling a system of 16. 

Leif Isaksen pointed out a similar point in Ptolemy’s data: Isaksen, Leif. "Ptolemy’s Geography and the Birth of GIS." Digital Humanities: Book of Abstracts, University of Hamburg, Germany, July 16--22 2012: 236-239.



More information about the Crm-sig mailing list