IT WAS A MUGGY Kolkata afternoon in 2017 when the first lines of code for Shabdakalpa were carefully entered into a lab computer at Jadavpur University in West Bengal. In the School of Cultural Texts and Records (SCTR), established at the university in 2003, Sukanta Chaudhuri and his team huddled over scruffy printouts of 19th-century Bengali-language newspapers and digital scans of medieval manuscripts. In retrospect, the moment seems almost prophetic: by linking centuries of Bengali texts, the team planted the seeds of an ambitious and enduring project.
With Chaudhuri, the founding director of SCTR and the head of the Shabdakalpa project, at the helm, and supported by scholars like Subha Chakraborty Dasgupta and Abhijit Gupta, the team embarked on an audacious journey to map every Bengali word through time, in a comprehensive digital database.
“With almost mad optimism, some colleagues and I set about it when we had very little money,” Chaudhuri told me. “We didn’t have money even to employ a single project assistant. We would pay the workers per job done – process one file, and we give you so many rupees.” Yet, he stressed that a historical dictionary, by its very nature, is a large-scale undertaking: “You cannot have a historical dictionary on a small scale. It is necessarily a large-scale project.”
Shabdakalpa, which loosely translates to “imagination”, is unlike any previous Bengali lexicon. Conceived as a “historical dictionary of the Bengali language,” it aims to compile a corpus of every Bengali text from the earliest literature to the present day. In practice, this requires custom-built software to parse, date and record each occurrence of every word. The result is an electronic archive that traces a word’s first appearance as well as its changing meanings and forms across the centuries. Shabdakalpa is not merely a word list or translation tool, but a comprehensive record of the Bengali language’s history.