Merging 3 centuries of metadata - experiences from an astronomical scientific letter collection
Abstract
The 1800s scientific letter collection after Norwegian geophysicist and astronomer Christopher Hansteen (1784-1873) was from the beginning considered to be of great value due to its range and the position of Hansteen as the first astronomer and leading scientist. In the years after Hansteen's death, the collection containing approximately 1500 letters by 420 persons was organized, indexed, and packaged with skill and care, and it survived movings and different types of exploration. In 2005 a project digitized and cataloged approximately half of the letters before funding went out, and the preliminary database was left in the shadows. In 2018 the support for the database ceased, and we transferred the data experimentally to the ExLibris Alma library catalog. Then finally in 2024 - new funding was allocated to finalize the digitization and cataloging, was aligned with the decision to use the "Alvin Platform for digital collections and digitized cultural heritage" for the University of Oslo library special collections. In this presentation, we will explore issues and explain decisions that influence and enable the use and dissemination of these types of fully digitized scientific letter collections. One main observation is that a "one stop solution" does not exist, but if properly documented and preparation for subsequent development is cared for, next century will be grateful.
Full text
Merging 3 centuries of metadata - experiences from an astronomical scientific letter collection Line Nybakk Akerholt University of Oslo Library November 6th 2025
Background The Universe aligns 2024: •Funding to secure and organize the Observatory archives –Observatory project •Pilot period with Swedish Alvin - platform for cultural heritage decided to become permanent and in operation •Permanent working-group for special collections with head archivist as lead established •An already 2005 semi-digitized part of the archives needed attention – Hansteen’s letter collection –earlier attempts on migrations were not suitable •New functional working space for handling special collection and digitization Illustrative photo from Digitisation workstation at University of Oslo Library Photo by: Line Nybakk Akerholt
Background: Observatory project •The Observatory archives 1834-1934 moved along with the Observatory Library collections into the new Institute of theoretical astrophysics in 1934, and back to the Observatory appr 2015. •Today: University of Oslo - Museum of Cultural History –responsible for collections, University of Oslo Library for the printed collections. •The old archives was kept as one unit, but only the documents related to Hansteen had a written finding aid. Over the years the different heaps were flipped around in the storage room https://www.alvin-portal.org/alvin/view.jsf?pid=alvin-record:632115 Illustrative photo from Digitisation workstation at University of Oslo Library Photo by: Line Nybakk Akerholt
What we discovered of the old archival list/finding aid •A good structure to begin with, but some additions has been made •Not alphabetic within «character» •Å is whithin A (Å was not formally back in Norwegian before 1917) –and also some difficulties with umlaut vs not umlaut, and i vs j •Very high quality considering which authoritative sources available, but also advantage of being close in historic time •Essential for the series of projects connected to this collection •Some authors very difficult to verify –clues in this finding aid misleading Illustrative snapshot from the “Catalogue of Scientific letters to Prof.Hansteen “ https://www.alvin-portal.org/alvin/view.jsf?pid=alvin-record:529264
The Principle of provenance •The very first line in the finding aid is dedicated to Abel, N.H. AND with a glued-on note –the history of the letters visiting “sister´s daughter” in Kongsberg, and with references to enclosed copy of a request of return to the University library’s collection. •1999 –the University and the National library splits and older collections are from then on under 2 different ministries (Norwegian and “abroad” were the criteria) •At one point the scientific letter collection was suggested handed over to the National library, but without consulting the institutional unit responsible for the collections at the Observatory. Done in “good faith”, but this project would have been impossible – in retrospect - a very good learning point Illustration of the Oslo Observatory by Christian Heinrich Grosch. Digital copy retrieved from https://commons.wikimedia.org/wiki/File:Observatoriet.jpg
2005– computers do everything easier? In retrospect – the “positive sides” : •very nice work with metadata, what we found was wrong, was probably due to lack of sources or time consulting them •Scanning quality ok, but done with a smaller scanner Sorry, no image available from this process
2005– computers do everything easier? In retrospect “netagtive sides”: •When funding went out, no one followed up or documented properly why –all the nice work with 707 letters was nearly unavailable for 15 years •Database was ok, but had no front-end – not discoverable outside project •Clean backsides not scanned –but how do we know they were empty in 2024? •The knowledge gained by those working with the project in 2005 was not available as such in 2024 •Correcting some errors –creating new: Airy →Airz, Daubree →Daubru (dead bridge in Norwegian) Screenshot of rescued metadata and file, but with new error createdlater migrated to Alma
Rescue attempts 2018 •“We don’t handle archives” •“We don’t handle special collections” •→But we still did –quietly –because we experienced more frequent questions –special collections are “in between” and worth rescuing •About 10 ”special collections” just within the Science library that were identified (how many more are there?) Screenshot of spreadsheet with first attempt to get overview over our ”special collections”
1 GitHub conversion cleanup of the Letter collection •1 unique colleague with just the right competencies was able to start the resque process from the Oracle database •GitHub as repository for metadata and temporary front-end with JSON •A simple but very useful table-front-end was set up to serve search purposes for the collection •Matching with assumed(!) person authorities in our library catalogue •A clean-up of IDs and filenames was nessessary •All very well documented @ https://github.com/uio-library/hansteen-brev Screenshot of temporarily ”table front-end”
Summary of experiences «Today, the simplest definition of metadata, “data about data”, is widely known and accepted worldwide. As simple as it sounds, the metadata definition is far more complex than just these three words. The challenge lies in the diversity of definitions, formats, standards, and contexts, resulting in a lack of clarity regarding the true nature of the metadata» Banerjee, Sanghmitra (2025). Metadata: a scientific jargon. Retrieved from https://www.openscience.eu/article/infrastructure/metadata-scientific-jargon Illustrative photo of index-card drawer. Photo by Line Akerholt
The different LAM metadata concepts •What happens when they meet in one small project? •Recognizing the different needs of different types of collections are necessary –as well is respecting the different competencies –an ongoing learning project, but easy within group (although we still talk above eachother’s heads) •Difficult when the greater library is involved due to all other (larger, more important (?)) stakeholders
General systems will never be perfect for the single special collection •All hopes, ideas and promises –do not trust that a single system can cover all user expectations •Prepare to use some add ons! •For use and dissemination of complex collections; a reading guide / user guide should be considered •embedded in as file or a readme with version control! •Write some sort of publication (report, article..) do not trust webpages or systems to be around for more than a decade! Illustrative photo of index-card drawer. Photo by Line Akerholt
Suggestions for better future practice •Document everything (did we?) →especially why som roads were not chosen •Each collection has it´s own peculiarities – the core of what lies in «special collections»; don´t throw that away •Plan for «lowest common multiplum» to secure most important metadata and at least some reachable goals and success early on. •Always a new and better method around the corner, but easier to adapt if these smaller goals are completed – and discoverable. •Libraries, archives, museums, IT, usergroups! – the metadata scope is not uniform – each has its own special norms and demands •Most important is. • DO the REGISTRATION!