I wrote:
>>True.  And I just came across a questionable-date format in an example
> under the 375 $s on the MARC21 site.
> <>
>>But I don't want to rely on a single example as support for applying
> EDTF date formats

Richard Moore responded:
> That's not an example of an EDTF date format. It's an example of the way
> RDA indicates that uncertain dates should be recorded (e.g. in

I read RDA <> as
applying only to those dates applicable to the elements under 9.3, and
any date formatting instructions in RDA as applying to those
instructions found within the context of the RDA text.  For instance,
nowhere in RDA is there instruction, explicit or implicit, that the
Profession or Occupation element (9.16) be further enhanced
(optionally?) with dates that provide the time period in which a
person had a job or was otherwise occupied in a particular field.  We
may, of course, spell this out in the Biographical Information element
(9.17) or the Source Consulted element (8.12), both mapped to
different MARC fields.  Date of Usage (8.9) is limited to the
preferred name.

I did a little poking around the MARC site and I found this in
Discussion Paper 2008-DP05/2

"3.2.2. Period of activity
RDA includes elements for period of activity (start) and period of
activity (end).[!!] To include specifically in MARC, the same options
are available as with birth and death dates: either define a new field
for period of activity or define a new field to accommodate additional
dates with one or two subfields for period of activity (depending upon
whether you parse the start and end into two subfields or encode as a
range in one subfield)."

(These elements must have existed before the November 2008 draft of RDA.)

Proposal 2009-01/1 codifies the dating qualification for Field of
Activity and other fields with the supportive statement that
characteristics of an entity may change over time.  Otherwise, there's
no RDA source I came across that specifies the application of these
dates for 37x elements.

Ultimately, the 37x $s and $t subfields strike me as a MARC layer we
apply on top of the RDA information in our authority records.  I'm
seeking the MARC instructions/NACO policies that tell me what kind of
date formatting is allowed or required for the subfields in question.

For the time being, I'm following the approach you mention in another
post in this thread:
1.) use both known and fuzzy dates in the 046 for those RDA elements
that are mapped to that field
2.) stick with known dates in the 37x $s and $t and give fuzzy dates
in the 670/678s until I hear of a policy set down by the higher ups in

> ISO 8601 and edtf dates are only recorded in 046. Part of the problem is
> that there would be no way to indicate what date standard you are using
> in 37X.

I wouldn't mind getting that resolved if we catalogers find it
necessary to spell out what standard we're using in these $s and $t

Mark K. Ehlert                 Minitex
Coordinator                    University of Minnesota
Digitization, Cataloging &     15 Andersen Library
  Metadata Education (DCME)    222 21st Avenue South
Phone: 612-624-0805            Minneapolis, MN 55455-0439