Edmund Lee suggested a possible data structure for a thesaurus.
I would say that qualifying the period term is essential, incorporating all
the elements suggested:
(1) Whether continuous/intermittent
(2) Derivation of the date range (known historical dates or other authority
used)
(3) Confidence in the date range
There may be a little confusion when dealing with specific date ranges for
distinct geographic areas. Using Romans as an example once again, 43CE (SE
England), 47CE (SW & Midland England), 55CE (? Wales), 70CE (N England),
whenever whatsisname moved into Scotland, and the foundation of the Roman
fort in Ireland could all be used as starting dates for the beginning of the
Roman Empire period in different databases. That means at least six
different potential start dates for one term alone. Maybe one term could
have several different date range options governed by a qualifier. If the
qualifier is not used in a search, then the broadest possible date range
would be used.
Maybe also a proportionate indicator of the length of the period would be
helpful (although this could probably be computer generated). If a single
once-only activity is to be described, it might be helpful to know how this
would fit within 3000 years (Early Mesolithic: 0.00033), as opposed to 30
years (Flavian: 0.033).
Duncan
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|