Lingua Franca and its Grammar Footprint: Introducing an Index for Quantifying Grammatical Diversity in Written and Spoken Language^*

ABSTRACT

The paper introduces a method to quantify grammar diversity based on part-of-speech tagging (POS) and presents a POS diversity index which can be used for analysing written and spoken communication independently of a text’s length. For this purpose, a series of transcripts of the radio programme Lingua Franca, broadcast by the Australian Broadcast Corporation, are assessed. The radio presenter’s grammatical footprint is compared with the POS index values of the invited guests, and the hypothesis that a significant difference in the grammar structures used by the different speakers is tested for validity using a standard ANOVA calculation. Significantly different POS diversity is found in a comparison between groups of speakers. The index provides for comparative studies of grammar structures used by individual speakers or authors.

Notes

¹ A list of grammar classes included in the C5 tagset is displayed on the Internet at: http://ucrel.lancs.ac.uk/claws5tags.html