The Bengali Literary Collection of Rabindranath Tagore: Search and Study of Lexical Richness

Introduction

One of the most prolific writers in Bengali literature is Nobel laureate Rabindranath Tagore (May 7, 1861 - August 7, 1941). He had dominated both the Bengali and Indian philosophical and literary scene for decades. He was a social reformer, patriot and above all, a great humanitarian and philosopher. He had modernized Bengali art by changing its rigid classical forms. He was the ambassador of Indian culture to the rest of the world. For his eternal writing Gitanjali, he was awarded the Nobel Prize for Literature in the year of 1913, becoming the first Asian Nobel laureate. He is the only litterateur who penned anthems of two countries, Jana Gana Mana, the Indian national anthem and Amar Shonar Bangla, the Bangladeshi national anthem.

Different statistical techniques in stylistic analysis of literary texts have been studied for long time. An empirical law to estimate vocabulary size from collection size, which is known as Heaps’ law (Heaps, 1978), is now becomes a benchmark in the field of information retrieval, though it is not well-known in linguistics. Besides that, there are multinomial Bayesian approaches (Boender & Rinnooy Kan, 1987) and few essential but rarely followed procedures (Nation, 1993) to estimate the vocabulary size. Many lexical richness measures have also been studied and applied on English and other languages for years. Different measures of lexical richness were applied on the data from the works of three contemporary French singers (Ratkowsky & Hantrais, 1975), the volumes of the Travaux de Linguistique Quantitative (TLQ) series (Ratkowsky, 1988), which was initiated by Swiss publishing firm Slatkine in 1978, Biblical texts (Holmes, 1994) and sixteen works from eight English authors (Tweedie & Baayen, 1998) to study different lexical styles. The hidden connections in the medical literature have also been reported using lexical statistics (Lindsay & Gordon, 1999 May). The corpora of three playwrights, Euripides – a great tragedian of classical Athens, Aristophanes – a comic playwright of ancient Athens, and Terence – a playwright of the Roman Republic, was studied to compare the trends in vocabulary richness over time (Smith & Kelly, 2002). The number of different types in the first fifty thousand words in each text from the twelve texts of twelve different authors along with the effects of text-doubling and text-combining on measures of vocabulary richness have been reported (Hoover, 2003). Besides that, the studies of vocabulary richness have been done for child language and second language research to monitor changes in children and adults with vocabulary difficulties. Primarily type/token ratio was used to measure lexical diversity in child language research (Richards, 1987). After that, different advanced measures in child language and second language have been reported by many researchers (Bogaards & Laufer-Dvorkin, 2004; Haznedar & Gavruseva, 2008; Richards & Malvern, 2000 September). There is a large body of research works on information retrieval methods, including several commercial search engines for English speaking users. There are search engines for the literary works of Shakespeare.

There were no major works on statistical analysis as well as search engines for Bengali literary works. We have made an initial attempt on Rabindra Rachanabali collection to study vocabulary size and different lexical richness measures. Various measures of lexical richness have been computed for different genres of Rabindra Rachanabali collection and different chronological intervals. The statistical measures are also compared with the measures from another Bengali author Bankim Chandra Chattopadhyay. We also build a search engine, Anwesan, for Rabindra Rachanabali collection.