My understanding of the role of <did> is that it contains the information about the thing being described itself,
with all the elements contained within the <c0_> but outside the <did> being as it were meta information about
the contents of <did> - contextual matters such as the past history <custodhist> or wider context <scopecontent>
in which it exists.

As such the loss of <did> would only be a slight blurring of the issue, but it may complicate matters structurally as
it is unclear to what extent it underpins the enforcing side of the EAD DTD that controls the rigid format of the
document created, or how it would affect the way in which <c0_>'s contain their children.

Would the order of the elements formerly contained in <did> assimilated with those in <c0_> be dictated?  At
present once you have closed the <did> you cannot reopen it, which raises complications when describing
documents which occur in the middle of other documents (something which can only be overcome by redeclaring
the relevant <c0_> within the declaration subset thus:

<!ELEMENT c02            - -     ((head?, did*, (admininfo | bioghist | controlaccess | odd |
scopecontent | organization | arrangement | dao |daogrp | note)*,
                                   ( (thead?,c03+,did*,(admininfo | bioghist | controlaccess | odd |
scopecontent | organization | arrangement | dao |daogrp | note)*)* | dsc*))
                                  (drow+, c03*)  )

If <did> was to go it could solve one major flaw with the present DTD, the absence of an element for describing
the content of the item within <did>, where such a description clearly belongs.  It would then be possible to use
<scopecontent> among the relevant <did> contained elements, rather than floating outside beyond <did>'s pale,
where its only role can be to give contextual information.  If <did> stays, it will require an element within it to
take item description perhaps <scopecontent> itself, permiited in two locations, within the <did> for actual
description of the item concerned, and outside in its current role of providing contextual descriptions of groups of

One other thought: why select <c01> for retaining a <did>?  It is perfectly possible to do a finding aid where the
item level descriptions are at <c01> (I have several).  It seems against the spirit of the <c0_> elements to
differentiate between them in this fashion, as no doubt other lists will emerge which would require <did> at
<c02> or below.

Richard Higgins
Durham University Library