Text recognition models
Technology
Writing is one of the human race's greatest achievements. It is thanks to writing that we have a record of events, ideas, and cultures, giving historians the tools they need to reconstruct detailed narratives, interpret the perspectives of people from different eras, and trace the development of societies over time.
But these historical texts come in many different forms, according to the writing materials that were available at the time. In this post, we'll explore some of the most common—and some not-so-common—writing materials that have shaped our history and explain how you can use Transkribus to transform these analogue forms into searchable digital text.
Clay tablets: The earliest preserved writing (from 3100 BCE)
The first written texts were thought to have been produced in ancient Mesopotamia around 3100 BCE. Here, scribes pioneered one of the earliest forms of writing on a surprisingly durable medium: clay. Using a reed stylus cut into a rectangular shape, they pressed cuneiform script into wet clay tablets. Once dried, these tablets became almost indestructible. Thanks to their incredible resilience, countless examples of these ancient texts survive today, offering invaluable insights into the economy, administration, and culture of the first civilisations.
A clay tablet from around 2000 BCE, featuring cuneiform writing in the Akkadian language. © The Letterform Archive via Wikimedia Commons
Papyrus: The paper of the pharaohs (from 3000 BCE)
While clay was durable, it wasn't exactly portable. This limitation likely spurred the ancient Egyptians to develop a lighter, more versatile alternative: papyrus. Made from the papyrus plant, a type of rush that flourishes along the Nile River, its creation was an ingenious process.
Strips were cut from the plant's stem and laid side-by-side to form a rectangle. A second layer of strips was then placed crosswise on top. Through a process of wetting, pressing (sometimes with adhesives), and hammering, the layers would bind together. After drying in the sun, the writing surface was polished smooth with a piece of ivory or a shell. To create longer documents, up to twenty of these rectangular sheets could be pasted together and rolled into a scroll: the book format of the ancient world.
The Edwin Smith Papyrus is the world's oldest surviving surgical document, and dates from around 1600 BCE. © Jeff Dahl via Wikimedia Commons
Bamboo books: Writing vertically (from 1500 BCE)
Meanwhile, in Asia, where the papyrus plant was not readily available, another plant became the medium of choice: bamboo. The vertical nature of early Chinese characters was perfectly suited to the long, thin strips of bamboo. For a single column of text, one bamboo strip was ideal. To create more extensive documents, scribes would simply thread the individual strips together with two lines of string. This practical design left a lasting legacy; one of the modern Chinese characters for a book (册) is a pictogram that beautifully evokes an image of these ancient bamboo strips threaded together.
Dating back to the Warring States period, these Chinese bamboo strips were used as a chart for calculations. © Tsinghua University via Wikimedia Commons
Palm-leaf manuscripts: A south Asian legacy (from 500 BCE)
In the Indian subcontinent and across Southeast Asia, the palm leaf became the dominant writing material, with its use dating back to at least the 5th century BCE. Scribes would use the dried and smoke-treated leaves of the Palmyra or talipot palm to record everything from religious texts to scientific treatises. This tradition of writing on palm leaves continued for centuries, only declining with the arrival of the printing press in the 19th century.
Transkribus is working with the Wikimedia Foundation to help preserve and transcribe a significant collection of Balinese palm-leaf manuscripts, making this unique cultural heritage accessible to a global audience.
Read more in "Preserving Cultural Heritage: Transkribus Integration with Wikimedia Projects"
Palm-leaf manscripts would be tied together to create small books, many of which have survived until today. © Tropenmuseum via Wikimedia Commons
Parchment and vellum: The rise of the book (from 200 BCE)
While leather had been used as a writing surface since about 2500 BCE, it was the development of parchment around 200 BCE that revolutionised the written word. This new technique treated animal hide so that both sides could be used for writing, making it a far more efficient material. From the 4th to the 15th centuries, parchment was the standard writing surface for medieval European scribes, and it is the material used in all the famous illuminated manuscripts produced in monastic scriptoriums.
For the most luxurious books, an even finer and softer version called vellum was used, crafted from the hides of young or even unborn calves, kids, and lambs. Crucially, parchment sheets could be easily folded and sewn together, leading to the creation of the codex—the book format we still use today. This was a significant advantage over the cumbersome papyrus scroll.
The State Archives of Zurich is one of the many archives around the world that uses Transkribus to digitise and transcribe parchment manuscripts.
Read more in "Transcribing Medieval Handwriting with HTR and Transkribus"
Parchment was used not just for written text, but for musical notation too, as this codex from 1297 shows. © Biblioteca Nazionale Centrale di Firenze via Wikimedia Commons
Paper: A revolution in writing (from 105 CE)
Paper as we know it first appeared in China in about 105 CE, but it would take nearly a thousand years for it to become the default writing material in Europe. The original process involved repeatedly soaking, pounding, washing, and boiling substances like rags and plant fibres. The resulting mush was strained in a mesh frame and left to dry, producing a sheet that was thinner, more flexible, and far more suited to mass production than either papyrus or parchment.
In the 19th century, the process was industrialised with the introduction of wood pulp, which was much easier to source than rags. However, this new paper was less robust, and sadly, many paper documents from even the most recent centuries have not survived in good condition.
The "Material Culture of Wills" project is using Transkribus to preserve the knowledge contained in 25,000 paper-based wills over a few hundred years.
Read more in How the "Material Culture of Wills" Project Transcribed 25,000 Wills with Transkribus
A 19th-century diagram showing the new manufacturing process of paper using wood pulp. © Robert Hunt via Wikimedia Commons
Bringing analogue history into the digital age with Transkribus
These diverse writing materials share one common trait: they are all analogue. Unlike digital files that can be instantly searched, shared, and analysed, the knowledge contained on clay, papyrus, or paper must first be digitised.
This is where Transkribus comes in. Our platform is designed to digitise and transcribe historical written materials, making them fully accessible for modern digital research. And the best part? Transkribus can work with any kind of writing material.
It doesn’t matter if your text is inscribed on a clay tablet, written on a palm leaf, or penned on modern paper. All you need is a clear digital image of the text. From there, Transkribus’s AI-powered technology can learn to read the script and produce an accurate, searchable transcription. The key is not the physical medium, but the quality of the image and the availability of a suitable transcription model for the specific script.
Ready to start your own transcription journey? For more information, visit our Help Center or check out the video below to get started.
