Tag Archives: Fossils

Considering labels

The Cockburn Geological Museum at the Grant Institute holds an extensive collection of over 130,000 geological specimens that reflect the whole spectrum of earth science materials, including minerals, rocks and fossils. Most of these specimens have labels – some have multiple labels, some of these labels are loose paper in the bottom of specimen boxes, while others are glued directly on to the rock or mineral. Some information is written on with red or blue paint. Some specimens have all of the above – some don’t have any labels at all.

There are several specimens at the Cockburn that are clearly marked ‘Sir C Lyell’ – in what looks to be his own handwriting – a good indication that they were originally part of his own collection.

Now that the University of Edinburgh has acquired Lyell’s 294 Notebooks, for the first time, in a long time, both the specimens and the documentary records, can be brought together to share the same space. The notebooks offer the chance to enrich our knowledge of the specimens, adding valuable context and insight into when and where they were collected, and what they were potentially used for.

Gillian McKay, Curator of the University of Edinburgh's Cockburn Museum

 

 

 

 

 

Using our now well-developed Lyell ‘next level’ palaeography skills, we feel ready to explore the links between specimens and the written information – but to get us started, we brought in the label expert!

Kate Bowell studies the Cockburn Museum's Lyell specimens.

 

Postgraduate researcher Kate Bowell is exploring the stories the National Museum of Scotland has told in their collection of 20,000 exhibition labels and how these stories have changed over time (See Kate’s blog here https://blog.nms.ac.uk/2021/12/14/a-history-of-exhibition-labels-and-the-stories-they-tell/ ). Her experience in studying the stories behind labels means she is the perfect person to help us start formulating a plan.

 

 

 

 

We were also pleased to have undergraduate student Will Adams join us. Currently in 4th year Archaeology at the University of Edinburgh, Will’s interested in archives and how they relate to archaeological collections – he is also on the quest to find a dissertation topic.

 

 

 

 

 

 

Could we join forces to help each other out? What followed was a joyous 3 hour discussion – exploring the history of labels, the history of collections, why people collect, how people use labels, personal collection administration, split and movement of collections, the rise and purpose of museums – and how museums subsequently label items, both for use and for public enjoyment.

Lyell’s administration throughout his collection – his page numbering, indexing and the labelling of his specimens – show that he actively used them as a resource for his work. No actual catalogue exists – and so we have to start slowly working out how he kept his collection in order, and how he used specimens to aid his understanding. Now that the collections are together, it should be possible to start to see how it all linked up – and there is huge potential to learn much more about the specimens.

For example, one of the Cockburn’s specimens, and part of Lyell’s original collection is this amazing Agate, labelled in Lyell’s own handwriting:

We recognise Lyell’s distinctive ‘e’ – and the place name Mount Horne points us to British Columbia[1]. The specimen’s original owner is noted by Lyell as the Honourable C.A. Murray. In many ways similar to Lyell, Charles Augustus Murray was an author and diplomat,. He attended Oxford University, and spent several years travelling across Europe and America from 1835 and 1838, describing his experiences in popular books on his return [2].

We know Lyell visited British Columbia several times; the collection includes both Charles and Mary’s certificates recording their passing behind Niagara Falls to Termination Rock dated the 7th June 1842; and a card representing Columbia College dated 1853. As we are able to identify critical information – names of people, places, mineral types – on the labels – these can be cross-referenced to text in the notebooks, allowing us to focus in on the history of the specimens. Creating this framework of knowledge allows us to develop our hypothesis about how travel, collaboration, and collecting (or trading) specimens fed into the larger ideas of the time relating to “how the earth systems worked”.

CA Murray's specimen relating to Mount Horne, part of the Lyell Collection at the Cockburn Museum

Will’s presence also helped us see how he can add archaeological detail to the specimens. Lyell’s interests where wide ranging, and his exploration of the history of man resulted in him collecting neolithic objects ranging from tools to beads. Of course, we cannot be experts in everything, and with the collection of specimens being held by the Grant Institute, they have been categorised very much as geological specimens. Will’s contribution proved how collaboration with people who can view the objects with an “archaeological eye” adds significant detail to the objects. Our meeting provided him with the perfect opportunity to dive in and begin to think about a project combining his interest in archives and collections. Inspired, Will has booked into the CRC Reading Room to start looking at the collection in more detail, and is talking to his dissertation advisor to firm up a plan.

The benefits in bringing both the collections and experts together are tangible. Collaborative work will really enhance the Lyell collection – indeed, our afternoon spent considering label gave us a practical insight into how he himself worked and used the collection.

[1] https://www.mindat.org/feature-6081205.html

[2] Charles Murray (author and diplomat) – Wikipedia

In Lyell’s Own Words

This month we hear from Lyell Project Archivist Elise Ramsay and Project Volunteer Erin McRae. Elise and Erin each reflect on their recent progress transcribing the Sir Charles Lyell notebooks using ground-breaking AI and machine learning, and their work together to develop this incredible AI tool for further use with the Lyell collections. 

Elise Ramsay, Project Archivist, holding an open scientific notebook and smiling

Elise Ramsay
Lyell Project Archivist

For me, the written word is the most captivating and characterful element of the Sir Charles Lyell collections. When reading Lyell’s own words on the page in graphite and ink, I can tell when he is writing from a desk, or in the field. In decoding his idiosyncrasies, I have come to understand a bit of the man himself. Understanding Lyell’s handwriting is the key to opening up this internationally significant collection. But it is also the first barrier. Lyell’s handwriting is of his time; often liberally abbreviated, topic specific, and faded. Complete transcription of the collection is paramount to accessibility, and recently, we have made some exciting progress towards this goal.

In early March 2021, the Charles Lyell Project team took part in hosting the EDITOR Transcription virtual workshop. In preparation for the workshop, two digitised notebooks from the Lyell collections (MSVII and Notebook No 4) were selected to be trialled with the Transkribus platform. Over 8 weeks, EDITOR project interns Evie Salter and Nicky Monroe transcribed these notebooks word for word. With this data, an algorithmic model of Lyell’s handwriting was created, effectively teaching Transkribus to recognise Lyell’s words on the page, and to decipher them automatically. This innovative work by the EDITOR Team, has revolutionised our systems and methods of cataloguing. Already we can see this balance of machine learning and human input has introduced new efficiency (and enjoyment!) to the task of transcription.

To build on this momentum, we were delighted to offer a remote volunteer opportunity aimed at trialling the newly created Transkribus model and testing the many features Transkribus offers. In this capacity, Erin McRae joined us in March, contributing to key cataloguing efforts and scoping the features of Transkribus for further use with the collections. Erin is a recent graduate from the MSc in History programme at the University of Edinburgh and holds an MA in Archives and Records Management from University College Dublin in Ireland. In only two months, Erin has produced tremendous material, and we are indebted to her. Here, Erin reflects on her first impressions of the Sir Charles Lyell collections and using Transkribus:

Profile Picture, Erin McRae, Volunteer

Erin MacRae
Lyell Project Volunteer

When I think of Sir Charles Lyell, I see a man in constant motion and possessing a thirst for knowledge that knew no boundaries. I can picture him observing the volatile Mount Etna, or immersed in the identification of mollusc species, or exploring geologic formations and petrified fossils millions of years old. I imagine him pausing to scribble down his observations in notebooks in his own inimitable style (a combination of English, French, Italian and Latin), so he wouldn’t miss any detail.

The detail of the collection is of untold value to researchers and presents interesting challenges as we describe the collection. In addressing these challenges, the Transkribus platform is an invaluable tool.  

Transkribus  is “a comprehensive platform for the digitisation, AI-powered text recognition, transcription and searching of historical documents – from any place, any time, and in any language.”1 Using the algorithmic handwriting model developed on the EDITOR project, we were able to upload more raw material from the Lyell collections to the Transkribus platform. In my recent work with Sir Charles Lyell’s notebooks, I found that Transkribus was able to decipher Latin species names with which I was unfamiliar. This saved me a significant amount of time and gave me the ability to transcribe much faster.  An example of this occurred when Transkribus identified “Fissurellagraeca”.2 A species of mollusc, this name has since been replaced by the accepted name “Diodora graeca3 . It is remarkable that it was correctly interpreted by the software in the first place. 

A screenshot of the Transkribus platform. On the left is a digitised image of a page of Charles Lyell's notebook 65. The handwriting is in ink, and an untidy scrawl. On the right is typed words, corresponding to each line in the image. The words are a word for word transcription.

An example of transcription output from the Transkribus platform.
From Sir Charles Lyell Notebook, No. 65
(Ref: Coll-203/A1/65) – (with apologies for the poor quality image).

 

 

 

 

 

 

 

 

 

The transcriptions that Transkribus produces require minor to moderate spellcheck amendments, primarily where vowels are mistaken. There were some instances of errors in phrases, names, and once a whole line of text. In this case I transcribed this line myself which I had done previously with indexes in two other notebooks. These issues are minor and they  do not detract from the immense amount of time I saved  using Transkribus compared to transcribing without the aid of the algorithmic model. In particular, we were all struck by the accuracy of the model in recognising and deciphering antiquated species names. This was invaluable and changes the role of the transcriber.

The overall benefit of the Transkribus software is that it is helping us to develop a much more comprehensive approach to describing and interpreting the Sir Charles Lyell Collections. To a much greater degree than previously possible, we can document and unlock the life and travels of this principal figure in the evolution of the discipline of geology.  

Elise Ramsay, Lyell Project Archivist
Erin McRae, Lyell Project Volunteer 

Sources and further information:
1.
Transkribus.” Read Coop. Accessed April 19, 2021.
2. “Fissurella graeca (Linnaeus, 1758).WORMS: World Register of Marine Species. Accessed April 19, 2021.
3. Ibid.

You can learn more about our revelatory transcription work on the Sir Charles Lyell Collections, part of the EDITOR project, on YouTube:
Editor Transcription Workshop: Day 1/Session 3 – Video 3 of 10 – YouTube
Editor Transcription Workshop: Day 2 /Session 3 – Video 6 of 10 – YouTube