The paper sets to investigate the amount of variation between the North Sami speakers living in two different majority language contexts: Finnish, spoken in Finland, and Norwegian Bokmål, spoken in Norway. We hypothesize that the majority language is a significant factor in recognizing variation of the North Sami language. Although North Sami is the biggest of the nine currently spoken Sami languages and it has become a lingua franca among the Sami speakers, there are clear differences in the pronunciation of the North Sami spoken in Finland and Norway, so that the difference can be used to recognize which majority language region the speaker comes from. Using a corpus of spoken North Sami collected in locations in Finland and Norway, we experimented in classifying the speech samples into categories based on the two majority languages. We used the i-vector methodology to model both intra- and between-dialect variations, and achieved the average recognition of about 17.31% EER for classifying the Sami speech samples. The results support our hypothesis that the variation is due to the majority language, i.e. Finnish or Norwegian, spoken in the given context, rather than individual variation.
Cite as: Jokinen, K., Trong, T.N., Hautamäki, V. (2016) Variation in Spoken North Sami Language. Proc. Interspeech 2016, 3299-3303, doi: 10.21437/Interspeech.2016-1438
@inproceedings{jokinen16c_interspeech,
author={Kristiina Jokinen and Trung Ngo Trong and Ville Hautamäki},
title={{Variation in Spoken North Sami Language}},
year=2016,
booktitle={Proc. Interspeech 2016},
pages={3299--3303},
doi={10.21437/Interspeech.2016-1438}
}