Bengaluru: At least 20 per cent of all content in previously studied Kannada inscriptions in the state potentially contains errors, experts have said, a discovery that has already set right some anomalies.
After four years of working to digitise inscriptions in the state, experts from the Bengaluru Inscriptions 3D Digital Conservation Project, Mythic Society, noticed discrepancies between the actual content of inscriptions and the existing literature on them.
Startled by this finding, experts embarked on a project to identify the most common anomalies and combed through nearly 30,000 characters in total. They found over 3,700 characters were commonly misread or misunderstood, which could have altered the meaning of some inscriptions.
Commonly misread characters
Some of the most commonly misread characters are: Ra, Da, Ya, Ka, Va, and Ga. For example, Ka was misread as Ra and vice versa or Sa was misread as Na, if the text on the stone was not clear.
Over 4,700 corrections were made, which provided more clarity on 31 places, 28 people and 40 dates, enabling the team to learn precise timelines regarding the construction of heritage buildings or the dates of events.
Udaya Kumar P L, founder of the conservation project, shared an example. The date of the Katanayakanapura inscription (near Hemmigepura, beyond NICE Road) was rectified from 1346 CE to 1336 CE. This discovery corrects the previous assertion made in B Lewis Rice’s 'Epigraphia Carnatica', which had claimed the last Hoysala inscription in the region was dated 1346 CE.
Another striking example: The place name that was earlier read as ‘Saramu’ has been corrected to ‘Sarakki’ in the Begur Nagatara hero stone inscription (900 CE). Sarakki is now a popular locality in south Bengaluru, and this correction shows that the place’s name is at least 1,124 years old.
"However, it is important to note that not all of the corrections have equally significant impacts; a majority of them are orthography corrections that don't substantially alter the meaning of the earlier readings,” Udaya Kumar told DH.
The errors in comprehension could have crept in due to the stone’s texture, erosion, the handwriting of the person who etched the inscription, and the time period in which each inscription was created, he postulated.
Looking ahead
“Based on the data we identified, data science students from the NMKRV College for Women, Jayanagar, used existing data science tools to create a programme that could quickly identify these errors in over 200 Kannada inscriptions,” said Udaya Kumar.
Karnataka is estimated to have around 30,000 inscriptions and nearly 50% of them are yet to be discovered and studied.
Therefore, this effort, which combines historical knowledge with modern data tools, will also help future discoveries by epigraphists and historians, who can learn which letters are most likely to be misunderstood and pay close attention to them.
“This will also help us develop a dictionary of Kannada characters and their changes over the years. So in the future, more historians can refer to it while studying inscriptions for more accurate readings,” Udaya Kumar added.