Computing prosody to detect the Arud meter in Punjabi Ghazal

Abstract

Ghazal is a very popular poetic form of Punjabi poetry. Every verse of a Ghazal follows the same rhythmical pattern. Punjabi Ghazal is written in Hindi meter and Arud meter. In this research, we deal with Arud meter. Arud is the science of versification being followed in Arabic, Turkish, Persian, Urdu, Hindi and Punjabi as well as some other languages of South Asia. It is a complex set of rules and has a steep learning curve for a novice. In this work, we devised an algorithm to detect the Arud meter in the Punjabi verse and developed a web-based application as well. This web application will not only facilitate the professional poets but also help the students to analyse the poetry in the context of prosody rules. Computing prosody of any poetry depends on recitation instead of written transcription. In the first phase of orthography, we analyse the text phonetically and phonologically to transform it according to the recitation by adding (as in gemination), removing (as in weightless nasalization and aspiration), modifying (as in tonal sounds) and grafting (to assimilate the sounds of neighbouring words) the letters. In the second phase of verse scansion, the verse is passed through a pipeline process of syllabification and step by step weight assignments, figuring out short, flexible and long syllables, ending up in one or more rhythmical patterns. All of these rhythmical patterns are compared one by one to standard and most famous 37 Arud meters. The Cartesian product type matching between verse patterns and Arud meters is solved using nested iterations, regular expressions and finite state automata. The meter that matches with verse pattern is declared as the Arud meter of the verse. This automatic process is more efficient than the manual process and yields satisfactory results.