At the New York Public Library we are in the process of mulling over
     exactly these issues and this discussion has been very valuable. I,
     too, wished to create a template that would put all the boxes at a
     given <c__> level. The solution that I am contemplating would fix the
     number of organizational (intellectual) levels and also the container
     levels. For instance:

     <c01>Primary organizational level (e.g. Series)
        <c02>Secondary organizational level (e.g. Sub-series)
           <c03>Tertiary organizational level
              <c04>Primary container (e.g. box)
                 <c05>Secondary container (e.g. folder)
                    <c06>Tertiary container

     All levels would *always* be used even if they didn't contain
     anything. For instance, if there was no tertiary organizational level
     the <c03> component would still be used. Thusly:


        <C01><HEAD>Series 1</HEAD>

           <C02> <HEAD>Correspondence</HEAD>


                 <C04> <DID> <UNITLOC CONTAINERTYPE="BOX">1</UNITLOC>
                 <UNITTITLE>1925-1936</UNITTITLE> </DID>

                    <c05> <DID><UNITLOC CONTAINERTYPE="FOLDER">1<UNITLOC>






     There are two questions that this usage raises:

     First of all, is it legal? I don't see why empty components can't be
     used--the DTD allows them--but it does feel like cheating. On the
     other hand I've always felt that the only reasonable way to deal with
     computers was to lie to them...

     Secondly, if this is legal, how many organizational levels are enough?
     As a naive non-archivist I assumed that three intellectual levels were
     adequate. That's one of the things that we'll be hashing out in our
     upcoming discussions.

     There are a few advantages to creating such a schema: 1. Ease of
     entry--such a setup would simplify markup to the point where
     professional involvement in the keyboarding process could be kept to a
     minimum. 2. Uniformity of display instructions for any browser or
     search engine. 3. Uniformity of searching protocols. 4. It doesn't
     interfere with the intellectual organization being the controlling
     principle of the structure as in Steven Mandeville-Gamble's example
     (that started this discussion).

     I fully realize that such a rigid model will not fit all our
     collections but I'm hopeful that we can come up with a structure that
     will be designed so well that it will fit enough of them to make it a
     useful template.

     Robert DeCandido               *  "Skating away on the thin ice of
     Preservation Division          *   a new day."
     The New York Public Library    *                 Jethro Tull