[Personal_archives] Literary types and Arrangement

Thu Nov 19 12:45:41 EST 2009

Thanks, Susan, for pointing to the final report and for enumerating the benefits to digital forensics techniques. I'll respond to your questions, Susan, since I'm not sure I'll have another chance this week. I agree that there is much to be gained from sharing case studies, though I suspect most of this will be old news to you.

We made minor revisions to the Paradigm records survey (at http://www.paradigm.ac.uk/workbook/appendices/records-survey.html), tailoring it more to authors (see the attached draft, though we may have sent a different version). We emailed the survey to the author, with an accompanying letter, prior to making a site visit, and then, on the day of the survey, we worked through the document with him. Some first-time field observations: it was a challenge to get the author to focus on the survey (the document and his machines) and to get accurate information and, upon later review, I felt I didn't get enough detail in my survey notes. I have found myself going back to the author to fill in gaps and/or clarify things and, thankfully, he has been very accomodating.

One lesson learned: though we tweaked the survey for authors who are most likely working independently, outside an institutional/networked environment, I think we need to go into future surveys prepared to respond to more complicated environments.

As far as the transfer is concerned, we largely followed Paradigm protocol for removable media, minus the encryption software/check. We sought to copy files preserving the directory structure, to run a fixity check, and to identify a non-invasive method (since we were taking copies from the author's working machines). We used an external hard-drive with USB connector, a terminal/command line code for a recursive md5, and a terminal/command line code for a hash comparison. Though we haven't sufficiently investigated digital forensics options yet, we did also try to get a disk image. Interestingly, it didn't work. One obvious lession here is that the work requires collaboration. In our case, between curatorial, archival, and IT staff. The work also requires some technical testing/troubleshooting. I would be remiss if I didn't acknowledge that a programmer at another institution (Peter Hornsby, Emory) provided the command line code for the recursive md5.

We also copied the author's Entourage email files (the "Main Indentity"). We've had only preliminary discussion about how to preserve it, including an exchange with Riccardo Ferrante of CERP (see http://siarchives.si.edu/cerp/parserdownload.htm).

Regards,
Michael

From: personal_archives-bounces at mailman.yale.edu [mailto:personal_archives-bounces at mailman.yale.edu] On Behalf Of Susan E Thomas
Sent: Wednesday, November 18, 2009 5:51 AM
To: Personal_archives at mailman.yale.edu
Subject: Re: [Personal_archives] Literary types and Arrangement

Hi Michael,

Great to hear some more about the Beinecke's experiences.

In answer to your first question, not all of the Paradigm collections have been fully processed. Those exemplars acquired from working politicians were generally quite recent in date and would not be suitable for full processing because we could not yet provide access to much of the material. The archive that has been most fully processed is that of the politician Barbara Castle (those not familiar with Castle might like to take a look at her obituary in The Guardian at http://www.guardian.co.uk/news/2002/may/04/guardianobituaries.obituaries). Castle formed a rather different case study for the project; her papers were at the Library before the project began, but her computers and disks were transferred in the first year of the project, once the Library had already appointed a cataloguer - Matthew Neely - to catalogue her archive.  The catalogue is complete and our curator of modern political papers is currently determining what may be released and when, and 'open' material is gradually being made available to researchers through this process. One of futureArch's developers - Pete Cliff - has been working to develop a researcher interface for the 'open' component of Castle's digital items. While we have made some digital materials accessible to researchers in the past, we have yet to do this via a proper interface. Once Pete's work goes into reading rooms, we will have our first real opportunity to see how scholars interact with this kind of hybrid archive and I hope to feed this interaction into the development process for refining our researcher interfaces.

Question number two. Yes, we did make a number of snapshot accessions during the project and the final report made to our funder (JISC) provides some information on those case studies (see http://www.paradigm.ac.uk/projectdocs/jiscreports/ParadigmFinalReportv1.pdf). For one creator, this amounted to three snapshots in just over a year. In practice the Library would be unlikely to take snapshots quite so frequently, but this is what a project timeframe produces. The Library is now working with the literary press Clutag, from which we also take snapshot hybrid accessions - these come in roughly annually and the timing tends to be tied to the publication of the Press' Archipelago magazine. Under our current futureArch project, we are processing many of our legacy hybrid archives for access and Clutag is one of those we're working on. As part of that process, we have recently captured* all the Clutag digital material received in the past three years and will be surveying and cataloguing this alongside the papers received with it.

For surveying born-digital material, we are making more and more use of 'digital forensic'  techniques, more typically associated with the police, FBI, fraud office, etc., rather than archives and libraries. The individual who has pioneered their use in the context of personal archives is Jeremy John, of the British Library, who has been a terrific collaborator (see his 2008 iPRES paper at http://www.bl.uk/ipres2008/presentations_day1/09_John.pdf). Forensic analysis software is designed to discover relevant evidence among vast quantities of data and supports a number of techniques useful to archivists and curators. So, how is it useful? Well, it helps with a number of things:

1. Provides an infrastructure within which to explore materials while maintaining authenticity - these environments have to produce evidence that can stand up in court
2. Identifying duplicates across different media or snapshot accessions
3. Identifying similar files (not something we've tested much yet, but could have huge potential for literary drafts)
4. Availability of bookmarking and annotation tools to support arrangement
5. Ability to preview a wide range of file formats
6. Ability to filter data, including the option to ignore non-archival data (such as clipart or application software)
7. Ability to recover 'deleted' data (yes, there are ethical issues here that need to be addressed)
8. Ability to export files and metadata in CSV format
9. Pre-configured data filters that allow discovery of certain file types that might be in unexpected locations on a hard disk drive

Moving away from the issue of appraisal and analysis for a moment and returning to the issue of transfer. I'm pleased to hear that the Beinecke has been able to adapt the Paradigm survey and transfer protocols; are you able to say some more about what changes you made, what the experience was like, and perhaps tell us something about the email system you encountered, and how you acquired the email from it? If others have experiences of 'digital transfer' that they are willing to share, I'm sure many of us would find that useful.

Best,

Susan

* By 'captured' I mean we have created 'disk images' of the material given to us on CD and floppy disk. These disk images can be used by forensic tools and can be mounted on computers as if they were the disk in question.

Susan Thomas
Digital Archivist/Project Manager
Bodleian Library

Web:  http://futurearchives.blogspot.com<http://futurearchives.blogspot.com/>
Tel:    +44 (0) 1865 283821
Post: Oxford University Library Services
         Osney Mead, Oxford, OX2 0ES

________________________________
From: personal_archives-bounces at mailman.yale.edu [mailto:personal_archives-bounces at mailman.yale.edu] On Behalf Of Forstrom, Michael
Sent: 17 November 2009 20:37
To: Personal_archives at mailman.yale.edu
Subject: [Personal_archives] Literary types and Arrangement
For starters, I'd also like to publicly congratulate Susan and her colleagues on the Workbook.

I'll respond to Susan's call for "other experiences" with personal archives containing digital materials and move on to discussion relating to arrangement. At the Beinecke Library we've been acquiring manuscript collections with what I've indentified as fugitive digital content (small numbers of disks) for roughly a decade. At Beinecke these disks tend to reside in the papers of writers and in the records of small presses and literary journals and they tend to contain word processing files, such as drafts of writings or letters.

We've processed several personal archives containing disks and discovered that, in general, small numbers of disks and files lend themselves to item-level analysis and arrangement by content and larger numbers of disks and files lend themselves to arrangement by format. There are of course other considerations, but these are the two models we've pursued to date.

We recently acquired our first real hybrid archive--a writer in mid-career. We made a macro/functional appraisal and a snapshot accession, copying files, incuding email, directly off two working machines in the writer's home. In doing so, I should add, we tested a revised Paradigm records survey and transfer protocol. We may be able to respect the original order in arrangement, though processing will be complicated by the fact that there is duplication between the machines and variations in the directories. Assuming additions to this collection, a preliminary proposal is to leave well enough alone and not integrate the digital and paper components at this time. We need to revisit this, but there is an argument for tackling this now, doing some additional appraisal, testing appraisal tools, in general, doing more so that some of the complexities of the accession, say, context provided through dialogue with the author, isn't lost or difficult to interpret at the point of future processing.

So, a couple questions for Susan. First, whether any Paradigm project collections have been fully processed? And second, if you conducted any snapshot accessions as part of the project, what if any additional work was  done on them in house. Other comments welcome of course.

Best,
Michael Forstrom
Manuscript Unit
Beinecke Rare Book & Manuscript Library
New Haven, CT   06520
203.432.7215

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mailman.yale.edu/pipermail/personal_archives/attachments/20091119/1d3bdd52/attachment-0001.html 
-------------- next part --------------
A non-text attachment was scrubbed...
Name: Paradigm survey revised.doc
Type: application/msword
Size: 36352 bytes
Desc: Paradigm survey revised.doc
Url : http://mailman.yale.edu/pipermail/personal_archives/attachments/20091119/1d3bdd52/attachment-0001.doc 