Proposal by BL: ”Between”, “Before” and “After” dates (Revision of RDA

1 July 2014

”Between”, “Before” and “After” dates (Revision of RDA

British Library (BL)

6JSC/BL/18/ALA response:  “Between”, “Before” and “After” dates (Revision of RDA (August 27, 2014)

Print Friendly, PDF & Email

Tagged , , . Bookmark the permalink.

21 Responses to Proposal by BL: ”Between”, “Before” and “After” dates (Revision of RDA

  1. John Myers says:

    It is an intriguing possibility, but the treatment of dates in the “before/after” and “not before/not after” options displays an arbitrariness of their selection indicating they are made up. Real examples would bolster the case for the necessity of the proposed change. (That is, what is the “literary warrant” for such a change?)

    • Kathy Glennan says:

      I took a look at some real examples, and I think there are some challenges in applying the revised instructions.

      LCCN: nr2002025102
      current 100 $d: active 1492-approximately 1511
      One ref. source: printer active in Bologna 1492, 1499-1510
      One ref. source: first book printed alone in 1499; retired in 1511
      — Question: would you still use approximately here?

      LCCN: n 83014261
      current 100 $d: active 1735-1757
      One ref. source: birth date unknown, died not before 1757
      One ref. source: flourished 1735-1757
      — Question: if trying to be more specific about 1757, and option 1 is used, does that make the ending date “after 1758”? Note that “not before” and “after” do not necessarily mean the same thing.

      LCCN: no 00094733
      current 100 $d: active approximately 1685-approximately 1732
      one ref. source: first mentioned before 1685, last mentioned before 1732
      –Question: would this make the revised $d: active before 1685-before 1732

      LCCN: nr 97019784
      current 100 $d: -approximately 1599
      one ref. source: his second wife is known to have been a widow in 1599
      –Question: would you still use “approximately”, or would you convert to “- before 1600”? Note that if you take this course, 1600 has to be supplied by the cataloger; it’s not in any of the 670s cited.

      • Kathy Glennan says:

        I do think that at least one example should be added to the “known to be” before/after instruction that illustrates period of activity.

        • Dominique Bourassa says:

          If this is accepted, should we will also add examples in the section Date of Birth and/or Death to show the impact of these new instructions on the construction authorized access points?

  2. Kathy Glennan says:

    The structure of the revised instruction implies a hierarchy in applying the various options; however, this is not explicitly stated. I assume a clause along the lines of “apply the following instructions (in this order)” would be a valuable addition to the revised instruction.

  3. Kathy Glennan says:

    The addition of two new paragraphs raises the question about the actual preferred hierarchy of applying these instructions (as proposed):
    – Probable year
    – Uncertain date, one of two years
    – Uncertain date, between two years
    – Uncertain date but outer limits known (before/after a particular year)
    – Approximate year

    By presenting the instructions in this order, I wonder how often “approximately” would be used over a before/after construction.

    I also wonder if this limits the applicability of “period of activity” vs. using a before/after construction for birth or death dates.

    • Kathy Glennan says:

      This raises an interesting question — what’s the difference between a probable year and an approximate year? RDA normally just uses “probable”….

      EDTF gives the following examples:

      uncertain: possibly the year 1984, but not definitely

      “approximately” the year 1984
      the year is approximately 1984 and even that is uncertain

    • Dominique Bourassa says:

      Maybe we could suggest that the order should follow EDTF. The result would be:
      – Probable year
      – Approximate year
      – Uncertain date, one of two years
      – Uncertain date, between two years
      – Uncertain date but outer limits known (before/after a particular year)

  4. Kathy Glennan says:

    I think that the wording in the 2nd new paragraph (both option 1 & 2) should include the term “particular” before “year”. The revision for option 1 would then read:

    If the year is uncertain but known to be before a particular year or after a particular year, record the date in the form before [year] or after [year]

  5. Kathy Glennan says:

    While not addressed specifically in this proposal, I wonder why the two instructions that only apply to “period of activity” are included in this main instruction. Should they be moved to, “Recording Period of Activity of the Person”?

  6. Kathy Glennan says:

    I’m having a hard time deciding which option I prefer. Reference sources will use whichever terminology suits the situation better.

    While consistency within RDA is generally good, if there are good reasons to prefer one treatment here for identifying persons (vs. the treatment for identifying manifestations in, I think that’s OK.

  7. Dominique Bourassa says:

    This is interesting and I can see how this type of information could be useful to users. There seems to be a need for this as there are already some RDA records in the NAF (Giecy, Ignacy Łukasz, 1745-after 1802; Masi, Giovanni, approximately 1730-after 1776).

    If these changes are accepted, will we have to add examples under Families which follows the same instructions as personal names ( and

    Finally, will there also be a need to add simlar instructions to dates of corporate bodies (11.4 and

  8. John Myers says:

    Is the solution to reference the Extended Date/Time Format (EDTF) as the specification for recording dates when they are distinct elements? And determining a corresponding equivalent notation amenable for use in AAPs? The transition under AACR2 from 1868- to b. 1868 and now back to 1868- is tedious enough. Now we are adding even more complexity to date notations. Will the whole edifice collapse under its own weight and bring about “naked” preferred names, differentiated under the hood by the URI and in public displays by a presentation of relevant elements (a la a Wikipedia disambiguation page)? Forgive me for wandering off-topic here. I guess I am troubled by the prospect of ever expanding boundaries regarding the recording of dates. Will the benefit derived from this addition sufficiently offset the resulting complexity in the rules?

  9. Kathy Glennan says:

    I’ve had some additional thoughts on this proposal.

    When it comes to the instructions in, why shouldn’t you record any/all dates as appropriate (based on cataloger’s judgment)? This would allow for recording the following, all for the same person [some of this is made up, but I think you’ll get the point]:
    period of activity: active 1320
    birth date: 1280?
    death date: after 1322
    birth date: 1280 or 1281
    death date: before 1325
    [note: I’ve essentially done this in a few NARs, using separate 046 fields when the situation warrants.]

    I think there’s no reason to be restrictive in choosing to use “after YYYY” vs. “not before YYYY” — you should use the information available in reference sources, etc.

    The problem with all of this is selecting the “correct” date when building the AAP. How important is it for people applying 9.19 to make the same decision about which date is recorded? (This dovetails into 6JSC/BL/20.) In any case, some sort of hierarchy is likely needed there — and I’m not sure we all agree on what that should be.

    As I said in my BL/20 comments, RDA prefers birth & death dates both in (as core elements) and in 9.19 (over any other addition to distinguish one access point from another). Thus, in my example above, I would end up using “1280?-after 1322” instead of “active 1320”. I assume that we don’t want to contest this, even though I could see an argument to use the most precise/certain of these kinds of dates instead.

    [Please let me know if you can think of a case where you would choose the “active” date in this situation in the AAP!]

    • Kathy Glennan says:

      If we take this approach, I think we should recommend adding a paragraph in the early part of (maybe the 3rd or 4th paragraph?) that says something about recording as many dates as are applicable to the person, using information from reference sources [etc.]

      I would welcome wording suggestions for how to convey this!

      • Kathy Glennan says:


        If different information is found in sources for the same date attribute, record the most specific date. Record additional dates according to the policy of the agency creating the data, or according to the judgment of the cataloguer.

  10. Kathy Glennan says:

    Would it be worth trying to add a sub-instruction in to deal with the variety of uncertain dates? Something like: Uncertain Dates
    If the exact date cannot be identified, record an uncertain date, using one of the following methods:
    a) For a probable date, record the date followed by a question mark.
    1816? — Probable year of birth
    b) For a date known to be one of two years, record the date in the form [year] or [year].
    1666 or 1667 — Year of birth uncertain; known to be one of two years
    828 or 829 — Year of death uncertain; known to be one of two years
    c) For a date known to be between two years, record the date in the form between [year] and [year].
    between 1310 and 1319 — Year of birth uncertain; known to be between two years
    d) For a date known to be before or after a particular year, record the date in the form before [year] or after [year].
    before 1685 — Year of birth uncertain; known to be before a particular year
    after 1802 — Year of death uncertain; known to be after a particular year
    e) For a date known to be not before or not after a particular year, record the date in the form not before [year] or not after [year].
    not before 1799 — Year of birth uncertain; known to be not before a particular year
    not after 1606 — Year of death uncertain; known to be not after a particular year
    f) If the year can only be approximated, record the date in the form approximately [year].
    approximately 931 — Approximate year of birth
    approximately 680 — Approximate year of death

  11. Steve Kelley says:

    Call me crazy, but I’m detecting some discomfort with this proposal as written. 🙂 I’m inclined toward Kathy’s idea of a sub-instruction for handling all varieties of date ambiguity. One problem I have with the current proposal is that “before” and “after” mean something different from “not before” and “not after” and it seems strange to eliminate one pair
    as an option while leaving the other.

  12. Robert L. Maxwell says:

    As has been noted, this proposal doesn’t take note of how this might affect the access point. is really only about the *element*, not the form in the access point, and we are already able to handle this in 046 (the MARC field for the element)–that is, we already ignore and use either the ISO or the EDTF format, which covers this problem. If this is meant to be for the *access* point, then the instruction needs to be in 9.19 (however, I wouldn’t like to see constructions like “Last name, first name, between 1310 and 1319-not before 1430”–VERY confusing). This is an endemic problem in RDA–the authors wrote the element sections in ways that make it clear that they were thinking of pieces of the access point, not recording information as attributes (elements).

    Bob, SAC

Leave a Reply