In this latest update on the Palladium email archive project, I explore an additional appraisal experiment, carried out with the intention of creating a small sub-collection which could be made available via ePADD’s Delivery module, (which allows for the display of full-text email at a designated computer terminal in a reading room) but was not intended for online metadata display on the Discovery module.
I discovered that this makes a difference to the way in which the collection needs to be appraised, as it was found that the distinction between online access requirements, and those which we would allow in a reading room environment with the copyright exemptions for research and educational purposes, allowed for a larger amount of email to be made open to access. It is also possible to retain a higher level of control over the access that is granted to the material. This experiment was carried out with the permission of Michael Schmidt (Managing Editor of Carcanet Press), and on the understanding that no access to this portion of the collection would be granted without his permission (as the Carcanet email archive is still officially a closed collection).
The correspondent chosen for the experiment was poet Elaine Feinstein. Feinstein had a longstanding relationship with Carcanet as her publisher, was a friend of Michael Schmidt, and passed away in 2019. Feinstein was deemed to be a good candidate because her records are no longer subject to data protection restrictions, and there is a significant number of email exchanged within the archive, covering a wide variety of topics, and including personal information shared with Schmidt. This enabled the project archivist to test the appraisal rubric developed by the Palladium project, and exposed a number of issues and challenges.
There are 1526 email identified as being exchanged with Feinstein in the collection dating from 2002-2019, and relating to: the editing, cover design and publicising of Feinstein’s books, to events and readings at which Feinstein gave readings and interviews, to reviews of and articles on Feinstein’s work, to the judging of poetry competitions, and to poetry, reviews and obituaries produced by Feinstein for PN Review.
As expected, interspersed within these email and in additional email, there is discussion between Feinstein and Schmidt of their personal lives: family, friends, other figures in the poetry world, medical concerns, and arrangements to meet socially. As this information is often interwoven within/included in professional email/conversations, these messages would also be prime candidates for testing the redaction tool we are developing as part of Palladium.
To identify the appraisal decisions, the follow standard ePADD labels were used:
- Do not transfer – applied to all email to be restricted
- Reviewed – applied to all email to confirm that they had been checked
To distinguish between this subset of email and the rest of the collection, the following bespoke labels were created:
- Do not transfer: Elaine Feinstein – to keep track of the restricted messages specific to this sub-collection, applied to restricted messages
- Elaine Feinstein: cleared for transfer to Delivery module – to assist in identifying the messages which are cleared for transfer, applied to reviewed messages to which access can be granted
After the email were each appraised individually, a process which took several weeks, 231 messages were identified as requiring restricted access, and labelled accordingly.
Based on the appraisal rubric, and best practice employed for appraising and Carcanet’s paper records, email containing information relating to the following subjects was restricted:
- Medical information relating to Michael Schmidt
- Information on/discussion of family members
- Personal information about individuals still living
- Contact information for individuals still living, including addresses, phone numbers, personal email addresses
- Private or group discussions of the judging panels of poetry competitions
Group email/circular email
When an email is sent to multiple participants, their email addresses are displayed in the header information of the email, sometimes in full, which presents a data protection issue in terms of granting access, even though the content of the email of often entirely innocuous. It is our intention to look into whether we might restrict this data, similar to the truncated email addresses provided in the Discovery module metadata.
Forwarded email/excerpts of email
Some email contains email forwarded from other correspondents, or excerpts of email pasted into the main body of the text. Sometime this will also include the email addresses of the correspondents, representing a similar data protection issue. This data is a candidate for potential redaction experiments. It was decided that, in terms of the text/body of the email message, unless it contained personal data, access would be granted in this case (via the Delivery module only), as this is the procedure I would have followed for the paper records.
Although we are now ready to explore transferring this sub-collection to the Delivery module, and the experiment may be deemed a success, the ongoing issue we face with email appraisal is that of volume. It is not going to be possible to devote several weeks to appraisal for researchers prior to granting access on a regular basis; the sheer number of email make this prohibitive.
Email, and digital records in general, require us to explore new appraisal techniques, and re-evaluate our approach to access. Whilst data protection restrictions are still in effect, the Palladium project has been exploring mediated access options, including collaborative appraisal and the large anonymised data sets that ePADD is able to produce.