On the status of Bìrọ̀m vowel harmony

Description

This is the R script that was used in running the acoustic analysis of the data presented in the paper

Steps to reproduce

Stimuli consisted of 55 CVCV morphophonemic words drawn from Roger Blench’s wordlist. Control for tone or consonant types was not done, but variations in words were captured. The ELAN (version 6.7) and Praat (version 6.2.23) software are used for annotation and extraction of acoustic measurements, respectively. Using scripts written by Riebold (2013), acoustic measurements from 1,650 tokens were automatically extracted and exported to RStudio. R (version 2023.12.1+402) and RStudio software (version 4.3.2) are later used for the descriptive statistical analysis of the acoustic measurements, while PhonR (version 1.0.7) is used to carry out statistical analysis between the mid vowels and for plotting the graphs for the vowel visualization. Tokens from speakers are anonymized, and the values of the acoustic parameters were z-score normalized to reduce the effect of inter-speaker variations. To identify the specific differences between the different ATR pairs of vowels, we did post hoc tests by repeating subsets of each relevant vowel pair; that is, between [e, ɛ] and [o, ɔ]. The null hypothesis in each case is that the tongue root does not affect the distribution of the acoustic parameters. Instead of focusing on any specific comparison in speakers' speech patterns, the study aims to describe the feature responsible for vowel harmony in Bìrọ̀m.

Institutions

• University of Nigeria

  Enugu State, Nsukka

Categories

Licence