Using LSTMs to Assess the Obligatoriness of Phonological Distinctive Features for Phonotactic Learning

To ascertain the importance of phonetic information in the form of phonological distinctive features for the purpose of segment-level phonotactic acquisition, we compare the performance of two recurrent neural network models of phonotactic learning: one that has access to distinctive features at the start of the learning process, and one that does not. Though the predictions of both models are significantly correlated with human judgments of non-words, the feature-naive model significantly outperforms the feature-aware one in terms of probability assigned to a held-out test set of English words, suggesting that distinctive features are not obligatory for learning phonotactic patterns at the segment level.

[1]  D. Kemmerer,et al.  Phonotactics and Syllable Stress: Implications for the Processing of Spoken Nonsense Words , 1997, Language and speech.

[2]  Geoffrey E. Hinton,et al.  On the importance of initialization and momentum in deep learning , 2013, ICML.

[3]  David B Pisoni,et al.  Perception of Wordlikeness: Effects of Segment Probability and Length on the Processing of Nonwords. , 2000, Journal of memory and language.

[4]  George N. Clements,et al.  The Role of Features in Phonological Inventories , 2005 .

[5]  Jeff Mielke,et al.  The Emergence of Distinctive Features , 2008 .

[6]  Timothy J. O'Donnell,et al.  A Generative Model of Phonotactics , 2017, TACL.

[7]  Alexander M. Rush,et al.  Character-Aware Neural Language Models , 2015, AAAI.

[8]  Richard N Aslin,et al.  Young children's sensitivity to probabilistic phonotactics in the developing lexicon. , 2004, Journal of experimental child psychology.

[9]  Ewan Dunbar,et al.  Quantitative methods for comparing featural representations , 2015, ICPhS.

[10]  Bruce Hayes,et al.  Explaining sonority projection effects* , 2011, Phonology.

[11]  Robert R. Sokal,et al.  A statistical method for evaluating systematic relationships , 1958 .

[12]  John R. Anderson,et al.  Human memory: An adaptive perspective. , 1989 .

[13]  Noam Chomsky,et al.  Some controversial questions in phonological theory , 1965, Journal of Linguistics.

[14]  P. Jusczyk,et al.  Infants' sensitivity to phonotactic patterns in the native language. , 1994 .

[15]  Amélie Bernard.  Novel phonotactic learning: Tracking syllable-position and co-occurrence constraints. , 2017, Journal of memory and language.

[16]  Matthew Goldrick,et al.  Phonological features and phonotactic constraints in speech production , 2004 .

[17]  Jürgen Schmidhuber,et al.  Long Short-Term Memory , 1997, Neural Computation.

[18]  Luca Antiga,et al.  Automatic differentiation in PyTorch , 2017 .

[19]  Jeff Mielke.  A phonetically based metric of sound similarity , 2012 .

[20]  Hermann Ney,et al.  From Feedforward to Recurrent LSTM Neural Networks for Language Modeling , 2015, IEEE/ACM Transactions on Audio, Speech, and Language Processing.

[21]  J. Werker,et al.  Cross-language speech perception: Evidence for perceptual reorganization during the first year of life , 1984 .

[22]  Peter Graff,et al.  Communicative Efficiency in the Lexicon , 2014 .

[23]  Bruce Hayes,et al.  A Maximum Entropy Model of Phonotactics and Phonotactic Learning , 2008, Linguistic Inquiry.

[24]  B. Elan Dresher,et al.  The arch not the stones: Universal feature theory without universal features , 2015 .

[25]  Adam Albright,et al.  Feature-based generalisation as a source of gradient acceptability* , 2009, Phonology.

[26]  Charu C. Aggarwal,et al.  On the Surprising Behavior of Distance Metrics in High Dimensional Spaces , 2001, ICDT.

[27]  H. Storkel,et al.  Differentiating phonotactic probability and neighborhood density in adult word learning. , 2006, Journal of speech, language, and hearing research : JSLHR.