In France, numbered <c0#> are not in use in libraries, and as far as I know, are not  much used in archival repositories either. We recommend the use of <c> elements and the @level attribute. The @otherlevel attribute is also used, especially when converting legacy finding aids where descriptions don't match a series/file/item pattern or something similar. A large number of these legacy finding aids date from the late-19th or early-20th century and their descriptions were not structured in the way they would be today. The collections they describe include both individual manuscripts and archival fonds so the use of numbered <c0#>, where each number would always correspond to a pre-defined description level, would not have been satisfying. When these descriptions are being revised today the <c> structure can easily be changed by modifying the value of @level or @otherlevel. It saves us a lot of re-numbering <c0#>.
I understand the need for human-readable cues but I don't believe that <c level="[value]"> is more difficult to read than <c01>, <c02>, etc. It also depends on the software you use to edit your EAD data. Most people over here use XML editors, where the tree structure, and hence the description level, is easy to see.

Florent Palluault
Coordinator of the upcoming EAD Guidelines for French libraries

----- Message original -----
De : Michele R Combs <[log in to unmask]>
Pour : [log in to unmask]
Envoyé le : 15/04/2010 11:34:25 PM +0200
Sujet : Promoting Container Levels

[log in to unmask]" type="cite">

Well, the numbers mean something in the sense that it’s easy to see at a glance how far down in the intellectual hierarchy a given c0# is – a c03 is 3 levels down – and you know that a given c03 will be intellectually (and XML-ly) a child of the first c02 encountered above it.  Of course one can do the same thing by counting ancestral c0’s, but that requires a lot more thinking.  I guess in the end it depends on whether you’re likely to have human eyes looking at the EAD at some point.  If so, the numbers are a nice human-readable cue.  If not, then as several people have said, it doesn’t really matter because computers don’t care, so as long as you a) do it consistently and b) write, or choose, your output handling (XSLT or whatever) to deal correctly with whatever option you choose to go with, you’ll be good.


My personal preference is to include the human-readable cues because you never know when the automagic processing will crap out on you and you’ll need a human being to fix it ;)   Also, if you go with unnumbered, doesn’t that mean you have to set the level attribute for every single <c>?


Either way, I wonder how often Nathan’s situation of needing to promote/demote c0#s comes up?  Seems like it would be a fairly rare occurrence.  We’ve created or converted about 2000 finding aids so far and I can only think of one occasion when I had to do it.




From: Encoded Archival Description List [mailto:[log in to unmask]] On Behalf Of Ethan Gruber
Sent: Thursday, April 15, 2010 4:15 PM
To: [log in to unmask]
Subject: Re: Promoting Container Levels


I think a problem lies in the perception that the numbers in components have some sort of "meaning."  A c01 can be a series or an item.  It is semantically incorrect to place emphasis on the numbering system at all.  Suppose you were to process your collection into two outputs for comparison.  For the first dataset, you strip out all the level attributes and only have numbered components.  For the second dataset, you replace all the numbered components with unnumbered <c>, but retain the level attributes.  Clearly the second dataset retains intellectual meaning.

I can kind of see how <c level="series"> may be a little more difficult to read than <c01 level="series"> if you are using a simple text editor like NoteTab, but it seems like renumbering a collection is a nightmarish scenario.  It seems as if there are not many institutions that use unnumbered components.  The EAD cookbook stylesheets don't work out of the box on a collection with unnumbered components, I think.  It assumes top level components are c01 and that template calls c02.  Any sort of machine processing is greatly simplified by getting rid of numbered components, and I think the community is trending towards using software applications to create EAD rather than a slew of workflows in which guides are manually made in NoteTab Pro or other such text editors.

Ethan Gruber

On Thu, Apr 15, 2010 at 3:14 PM, Jordon Steele <[log in to unmask]> wrote:



We do <c>.  Archivist Toolkit’s default output is <c>, not numbered components (although there is an option).


The main problem for me is not technological at all—it’s really hard for a human being to read the code if you don’t have cues like <c01>, <c02>, etc.  Otherwise, it works fine.  When our tech guy was creating our stylesheet, I asked him if having non-numbered components was an issue, and he said that it was not.  I believe, like you indicate, he just wrote the XSLT with the level attributes to differentiate.






Jordon Steele


Biddle Law Library

Penn Law School

3460 Chestnut Street

Philadelphia, PA 19104-3406

(215) 898-5011


From: Encoded Archival Description List [mailto:[log in to unmask]] On Behalf Of Creighton Barrett
Sent: Thursday, April 15, 2010 2:21 PM

To: [log in to unmask]
Subject: Re: Promoting Container Levels


This could be different from what is under discussion here, but I've

seen "false levels" at multiple institutions while cleaning up batches
of EADs. As far as I can tell, the main reason component padding has
been done historically is for display purposes. Some processors in the
past may not have understoond CSS or didn't have access to the finding
aid CSS files.

For us, it was not a matter of CSS padding, it was problems that occurred during XSL transformation.  The XSLT we were using was a tweaked EAD cookbook stylesheet that couldn't handle a collection that had series with sub-series and series without sub-series.   For example, we were working with many collections intellectually structured like:

c01     Series: Vessel Papers

         c02     Sub-Series: M/V "O.K. Service X"

                c03      File: Manifests

                c03      File: Captain's Logs

c01    Series: Financial Records

        c02      File: General Ledger

        c02      File: Annual Reports

As I understood the problem, the XSLT we were using was not testing individual components to see if the attribute was series, sub-series, file, etc.  It thought all c02s were sub-series and would not correctly transform c02s that were files.  That might be a simple fix, but it was beyond our XSLT knowledge.  In the absence of a dedicated XSLT person to write the XSL test for us, we added "false" c02 layers to our series that did not contain sub-series so all files would be encoded as c03s.  We then uploaded static HTML files into the university's content management system, which controlled all the CSS.

It was less than ideal, but it got the job done.  I am especially interested in Ethan's comment about just  using <c> and @level.  Is anyone else doing this?