Tehmina Goskar, Research Curator, talks about the museum’s dairy-related objects as part of an ongoing major review of our farming collections.

A critical part of our Farmers will like it project is to review the collections we already have and get to know them again. As well as good museum practice to monitor the condition of objects and make sure they are where our database say they are, it is a lovely way to get to know the stories behind the objects anew.

How we review our collection

Over the last couple of months we have been focused on the Dairy section of the museum, and also expanding our search to find other objects and photographs related to dairy work in Cornwall in our collections. This work has been carried out by me and Mary, Museum Assistant (Collections). The museum doesn’t have much working space for reviewing objects outside our fabulous costume store so we do this work in the gallery during visiting hours. I like to do this when it is not too busy because it’s a chance for us to talk to visitors about museum work that is often hidden away.

The Dairy in the museum is in the central part of our building, called the Meat Market. The Meat Market extension was opened to the public in 1983 and not an awful lot has changed since then when it comes to the display and interpretation of our farming collections. We now have new interpretation boards to encourage visitors to ‘look up and around’ at the many rare agricultural artefacts that can get missed during a visit. Here was also an opportunity to bring in some of our many farming photographs into the displays and we hope to do more of this.

The Meat Market can be a challenging gallery to work in as the floor slopes (the old market followed the contours of the old market floor plan which was cut into Penrose Hill and essentially just had walls and roofs added to make the market building). We set up a trestle table with a covering of Tyvek – an inert and waterproof synthetic fabric which has many museum uses to protect objects. Our object database, Modes, was set up on a laptop so we could quickly and easily compare objects, locations and their condition. Working systematically area by area we created a register of items, their labels and photographed them for our catalogue.

Beetles on holiday

As the majority of our museum objects are on open display they can be vulnerable to changes in the environment and pests such as furniture beetle. Our agricultural objects, such as baskets, the hafts of spades, forks, hoes and other hand tools, are mostly made of wood and wicker are particularly vulnerable to insect damage. Sadly many came to the museum as living hotels for these tiny, hungry beetles. This problem has the potential to get very serious because of climate change as the insects breed during mild wet seasons which are getting longer. As part of this project we will be investing funds in conserving these precious artefacts and keeping the pests at bay. We will be writing more about this soon.

Making new friends

Collections review work is slow. You have to be methodical and deliberate. This can be challenging when you are working in a public environment but, like the saying ‘measure twice, cut once’ it pays to do it slowly between two people. It sounds strange but working at the pace you would have a nice lunch with a friend is about right for doing museum collections work properly. Always do it in pairs.

In Cornwall we sometimes call lunch croust and both Mary’s and my favourite object in the Dairy section is the beautiful wicker croust basket, the kind used for carrying a pasty (not like the modern kind but sometimes just containing lard, dripping, ham, potatoes, leeks or even nothing – called a ‘windy pasty’). Bread, butter, cream and cheese or anything tasty available might be other options, especially if you were yourself a dairy worker.

While one of us physically checked an object for a marked number and examined its condition, the other checked the record on the Modes database and annotated our spreadsheet. We end each review with a photograph which will then be added to the object’s record on our collections database.

Cream, butter, eggs, milk

Dairy is important to consider when reviewing our farming collections as we are not just concerned with tools of the field, but also what happened to the goods produced. The museum’s origins are also in dairy commence, starting out in the disused butter and egg market of Helston.

Our review showed holdings that far exceeding our expectations and researchers and visitors can expect to learn a lot more about the making of clotted cream, butter making, eggs, and of course milk. Specialised objects in the collection include barrel butter churns, including the Royal Prize Victory Churn from the 1920s, from Higher Lanner Farm near Porthleven; the “Ventilator” patent egg carrier; egg preserving pale that used bacteria-resistant isinglass; hessian sack for milk equivalent (milk powder); cream separators; and a range of butter stamps and pats, both decorative and practical; milkmaids’ yokes; milking stools; and objects relating to calving including a calving pulley and blocks, used to help deliver calves and a number of calf muzzles and spiked collars which were used to stop calves feeding from their mothers and depleting their milk stocks. This is no longer in practice.

ChatGPT talking to our dairy catalogue

The state of our documentation of dairy objects was mixed. Several objects had no corresponding record on the database but there were older index card records for them. Some objects such as the butter churns had misplaced labels and numbers making verification difficult. Some display labels had more information than our database records. This kind of ‘documentation creep’ happens in most museums. This is why collections reviews like this are critical and central to museum work in general as we can quickly identify problems and remedy them, while also learning new things.

We don’t find our collections database very user-friendly and so I am looking at new ways to interrogate our collections records. As part of this project I commissioned a short piece of research from Tom Goskar, archaeologist and computing specialist. The experiment was based on 129 records that happen to contain the word ‘dairy’. This included objects documented in the Dairy section of the museum, dairy photographs, and non-dairy objects that happen to be situated in this location. As AI and ChatGPT have reached the headlines, flooding us with scare stories and those of wonder, I thought it would be interesting to see what this technology made of museum catalogues.

Using Open AI’s ChatGPT 4 Large Language Model (LLM) this experiment was able to standardise inconsistent data such as the style of dates and the spelling of common words (not altering dialect or other historically significant spellings or expressions), it could identify specific aspects of the records that talk about craft and making, it could single out just the objects related to butter, it could count the number of objects located in the Meat Market and in other locations, e.g. the photographic archive, it could single out toys and dolls whose records were mixed into general dairy records because they might be a doll of a milkmaid, or a corn dolly, and it could converse with us about the detail of an object’s provenance, pick out ones described as being from a particular period, such as the 18th century, and from a housekeeping point of view, it could list objects that did not have a recent documented location.

While these are prosaic enquiries that could be answered more conventionally, the speed and accuracy of this method surprised us.

However using ChatGPT and other AI responsibly requires expertise, knowledge and experience. You have to know how to ask questions and you have to be able to interrogate answers when you might suspect it is making them up. We will be continuing to work with experts like Tom to see what they can do to help us better manage the information and data we hold on the collections.