John Tone Young wrote:
> I agree with the broad consensus that this is certainty rather than
> precision, but I'm wondering what you're supposed to do if (for instance)
> you're absolutely certain that it's 'notBefore' 1598 but only making an
> educated guess that it's 'notAfter' 1615.
>   
Sebastian wrote:
> does this happen in practice? more likely to be
> things like
> <birth notAfter="1598" cert="low"/><death when="1623-08-01" cert="high"/>

I think this sort of thing really does happen all the time.
Whether you are dating texts, people, or events, one is 
always having to say (e.g.) "composed certainly not before 1615, 
since the events of that year are mentioned on line 6, but 
perhaps as late as 1650, or even later. The absolutely 
latest possible date is 1945, the date of the now 
lost transcript." Or "Could not have been born 
in Kirk Blisholm before 1470, since the town was not 
founded till that year, but may have been born as late 
as 1510, if he is really the same John Doe as appears 
in the University register of 1525." 

Am I right in thinking that the W3C date formats can
indicate precision only when the precision is relatively
high? I.e., there is no way designate dating precision 
only to the decade (like MARC's clumsy "198u") or century 
("19uu")?

pfs
-- 
Paul Schaffner  Digital Library Production Service
[log in to unmask] | http://www.umich.edu/~pfs/