[Moims-rac] Re: Thoughts on RAC and Provenance

[Cal Lee]

I agree with Don on various points he addressed in his message. 
Particular statements I would note are:

"The goal of the RLG-NARA Task Force on Digital Repository Certification 
has been to develop criteria to identify digital repositories capable of 
reliably storing, migrating, and providing access to digital collections."

"Note that this does NOT say the goal is to develop criteria that meet 
every need of any repository to establish that it is performing all of 
its mission in a certified manner."

I think this is essential to remember, in order to avoid scope creep in 
the RAC work.

Repositories should be comfortable with the fact many of their core 
functions and activities might not fall within the scope of this 
particular set of criteria.  For example, many archives have major 
responsibilities for:

- Working with people in various recordkeeping environments to ensure 
that they create and manage their records responsibly.

- Engaging in very labor-intensive and intellectually challenging 
appraisal decisions about what needs to be documented and what sorts of 
resources the repository will select for inclusion.

- Providing education and outreach to various stakeholders.

- Answering reference questions and providing other value-adding 
services for those making use of the collections.

In many cases, the above responsibilities might be even more important 
to a repository's mission than the things that RAC is addressing. 
However, that just means their own criteria for success will include 
many things beyond those in RAC.  In my view, that's very natural and 
acceptable.

> In other words, TRAC is not requiring provenance to reach back beyond  
> ingest.  Although it is not clear exactly what 'beyond ingest' means,  
> this certainly does not violate the OAIS definition of provenance.   
> Therefore I find that TRAC is not in conflict with the OAIS  
> definition of provenance.

I think this is also the right approach.  It's important to recognize 
that many repositories will have very stringent requirements for 
documenting and working with those in the Producer environment.  But 
doing so should not be mandatory for all repositories that are certified 
by RAC.

> In summary, I believe cases 1 and 2 above constitute valid cases  
> where no 'pre-submission' provenance is available as a practical  
> matter for the repository.  Nevertheless, there is good reason to  
> want such repositories to have been been certified as to their  
> ability to preserve and make their received information available in  
> a reliable manner.

Once again, I agree.  The reality is that many repositories will have 
little control over or information about things that happened before 
ingest.  I like the idea of specifying something a long the lines of:

To the extent possible, a repository should capture/create and preserve 
information about the life of a digital object before it was ingested. 
When little or no information about its pre-ingest life is available to 
the repository, it should document the absence of such information, so 
that future users can make informed assessments about the context and 
authenticity of the objects.

The above language isn't quite right, but I hope it reflects the basic 
notion I'm trying to get at.

- Cal Lee