Posts by Collection

portfolio

publications

The systematic stretching and contracting of ideophonic phonology in Pastaza Quichua

Published in International Journal of American Linguistics, 2016

This paper analyzes systematic differences between sounds used in ideophones and sounds used in the non-ideophonic or “prosaic” lexicon of the Pastaza Quichua language of Amazonian Ecuador. We compare a digitized corpus of vocabulary items with a list of ideophones identified from field observations. We find that if a sound, syllable structure, or stress pattern is distributionally restricted in Pastaza Quichua, it is likely to be normalized and expanded within ideophones. The overall system is also stretched among ideophones by the addition of new sounds to the obstruents. These expansions are complemented by an overall contraction among sonorant sounds within ideophones. Many of the sounds and structures that expand from the prosaic into the ideophonic system are found to be statistically significant in their differences between the ideophonic and prosaic systems. We conclude that ideophonic sounds and structures add greater complexity to obstruent sounds while diminishing the use of sonorants.

Recommended citation: Nuckolls, Janis B., Elizabeth Nielsen, Joseph A. Stanley, and Roseanna Hopper. "The systematic stretching and contracting of ideophonic phonology in Pastaza Quichua." International Journal of American Linguistics 82, no. 1 (2016): 95-116. https://www.journals.uchicago.edu/doi/abs/10.1086/684425

Avoidance of Dissimilar Labial Onsets: The Case of Subanon

Published in Oceanic Linguistics, 2016

Early work on the statistical structure of the “Proto-Austronesian” (PAn) morph revealed a distinct pattern of avoidance between the onsets of adjacent syllables if these were nonidentical labials (*pVb, *pVm, *bVp, *bVm, *mVp, *mVb). Descriptions of many attested languages that have retained a reflex of PAn *-um- ‘actor voice’ show that this dispreference extends to affixed words, in which case the avoidance triggers a pattern of allomorphy that has been called “pseudo-nasal substitution.” While this pattern is very prominent for word bases that contain an initial labial stop and a reflex of the infix *-um-, it is much less prominent for word bases that contain an initial labial stop and a reflex of the prefix *ma- ‘stative verb’. Variations in the way this family-wide dispreference is realized in particular languages are surveyed briefly, and then it is shown that Subanon of the southern Philippines extends the usual avoidance pattern by adding to pseudo-nasal substitution a metathesis of mo- (< *ma-) so as to convert even the reflex of *ma-p or *ma-b to an innocuous medial homorganic nasal-stop sequence.

Recommended citation: Blust, Robert, and Elizabeth Nielsen. "Avoidance of Dissimilar Labial Onsets: The Case of Subanon." Oceanic Linguistics 55, no. 2 (2016): 620-33. Accessed August 25, 2020. http://www.jstor.org/stable/26408430. https://www.jstor.org/stable/26408430

Modeling adnominal possession in multilingual grammar engineering

Published in Proceedings of the 25th International Conference on Head-Driven Phrase Structure Grammar, 2018

In this paper we describe insights gained from building an extension to the LinGO Grammar Matrix customization system to cover adnominal possessive phrases. We show how the wide range of such constructions attested in the world’s languages can be handled with the typical major phrase types used in HPSG and discuss the value of feature bundling in the multilingual grammar engineering context.

Recommended citation: Nielsen, Elizabeth, and Emily M. Bender. "Modeling adnominal possession in multilingual grammar engineering." In Proceedings of the 25th International Conference on Head-Driven Phrase Structure Grammar, University of Tokyo, pp. 140-153. 2018. https://web.stanford.edu/group/cslipublications/cslipublications/HPSG/2018/hpsg2018-nielsen-bender.pdf

The role of context in neural pitch accent detection in English

Published in Proceedings of EMNLP, 2020

Prosody is a rich information source in natural language, serving as a marker for phenomena such as contrast. In order to make this information available to downstream tasks, we need a way to detect prosodic events in speech. We propose a new model for pitch accent detection, inspired by the work of Stehwien et al. (2018), who presented a CNN-based model for this task. Our model makes greater use of context by using full utterances as input and adding an LSTM layer. We find that these innovations lead to an improvement from 87.5% to 88.7% accuracy on pitch accent detection on American English speech in the Boston University Radio News Corpus, a state-of-the-art result. We also find that a simple baseline that just predicts a pitch accent on every content word yields 82.2% accuracy, and we suggest that this is the appropriate baseline for this task. Finally, we conduct ablation tests that show pitch is the most important acoustic feature for this task and this corpus.

Recommended citation: Nielsen, Elizabeth, Mark Steedman, and Sharon Goldwater. "The role of context in neural pitch accent detection in English." Proceedings of EMNLP (2020). https://arxiv.org/abs/2004.14846

talks

teaching