Hostname: page-component-cd9895bd7-q99xh Total loading time: 0 Render date: 2024-12-27T06:32:34.707Z Has data issue: false hasContentIssue false
  • English
  • Français

Learning biases in opaque interactions

Published online by Cambridge University Press:  20 January 2020

Brandon Prickett
Brandon Prickett*
Affiliation:
University of Massachusetts Amherst
*
*E-mail: bprickett@umass.edu.
Get access Rights & Permissions [Opens in a new window]

Abstract

This study uses an artificial language learning experiment and computational modelling to test Kiparsky's claims about Maximal Utilisation and Transparency biases in phonological acquisition. A Maximal Utilisation bias would prefer phonological patterns in which all rules are maximally utilised, and a Transparency bias would prefer patterns that are not opaque. Results from the experiment suggest that these biases affect the learnability of specific parts of a language, with Maximal Utilisation affecting the acquisition of individual rules, and Transparency affecting the acquisition of rule orderings. Two models were used to simulate the experiment: an expectation-driven Harmonic Serialism learner and a sequence-to-sequence neural network. The results from these simulations show that both models’ learning is affected by these biases, suggesting that the biases emerge from the learning process rather than any explicit structure built into the model.

Type
Articles
Information
Phonology , Volume 36 , Issue 4 , November 2019 , pp. 627 - 653
Copyright
Copyright © Cambridge University Press 2020

Access options

Get access to the full version of this content by using one of the access options below. (Log in options will check for institutional or personal access. Content may require purchase if you do not have access.)

Footnotes

Thanks to Eric Baković, as well as the attendees of NECPhon 2017 and LabPhon 2018 for helpful discussion about the topics in this paper. The members of UMass's Sound Workshop and Phonology Reading Group, as well as the anonymous reviewers, also provided helpful insight. Most of all, I owe a thank you to Gaja Jarosz and Joe Pater for guidance, discussion and assistance throughout the research process. This study was supported by NSF grants #BCS-1650957 and #BCS-424077.

References

REFERENCES

Albright, Adam & Hayes, Bruce (2003). Rules vs. analogy in English past tenses: a computational/experimental study. Cognition 90. 119–161.Google ScholarPubMed
Baković, Eric (2011). Opacity and ordering. In Goldsmith, John, Riggle, Jason & Yu, Alan (eds.) The handbook of phonological theory. 2nd edn. Malden, Mass. & Oxford: Wiley-Blackwell. 40–67.CrossRefGoogle Scholar
Bates, Douglas, Maechler, Martin, Bolker, Ben & Walker, Steven (2015). lme4: linear mixed-effects models using ‘Eigen’ and S4. R package (version 1.1-12). cran.r-project.org/web/packages/lme4.Google Scholar
Brooks, K. Michael, Pajak, Bozena & Baković, Eric (2013). Learning biases for phonological interactions. Poster presented at the 2013 Annual Meeting on Phonology, University of Massachusetts Amherst. Available (August 2019) at http://idiom.ucsd.edu/~bakovic/bpb/BPB-Phonology2013.pdf.Google Scholar
Byrd, Richard H., Lu, Peihuang, Nocedal, Jorge & Zhu, Ciyou (1995). A limited memory algorithm for bound constrained optimization. SIAM Journal on Scientific Computing 16. 1190–1208.CrossRefGoogle Scholar
Chandlee, Jane & Jardine, Adam (2014). Learning phonological mappings by learning Strictly Local functions. In Kingston, John, Moore-Cantwell, Claire, Pater, Joe & Staubs, Robert (eds.) Proceedings of the 2013 Annual Meeting on Phonology. http://dx.doi.org/10.3765/amp.v1i1.13.Google Scholar
Corkery, Maria, Matusevych, Yevgen & Goldwater, Sharon (2019). Are we there yet? Encoder-decoder neural networks as cognitive models of English past tense inflection. ArXiv Preprint. http://arxiv.org/pdf/1906.01280.pdf.Google Scholar
Ettlinger, Marc (2008). Input-driven opacity. PhD thesis, University of California, Berkeley.Google Scholar
Giulianelli, Mario, Harding, Jack, Mohnert, Florian, Hupkes, Dieuwke & Zuidema, Willem (2018). Under the hood: using diagnostic classifiers to investigate and improve how language models track agreement information. ArXiv Preprint. http://arxiv.org/pdf/1808.08079.pdf.Google Scholar
Jarosz, Gaja (2014). Serial markedness reduction. In Kingston, John, Moore-Cantwell, Claire, Pater, Joe & Staubs, Robert (eds.) Proceedings of the 2013 Annual Meeting on Phonology. http://dx.doi.org/10.3765/amp.v1i1.40.Google Scholar
Jarosz, Gaja (2015). Expectation driven learning of phonology. Ms, University of Massachusetts Amherst.Google Scholar
Jarosz, Gaja (2016a). Learning opaque and transparent interactions in Harmonic Serialism. In Hansson, Gunnar Ólafur, Farris-Trimble, Ashley, McMullin, Kevin & Pulleyblank, Douglas (eds.) Proceedings of the 2015 Annual Meeting on Phonology. http://dx.doi.org/10.3765/amp.v3i0.3671.Google Scholar
Jarosz, Gaja (2016b). Refining UG: connecting phonological theory and learning. Paper presented at the 47th Meeting of the North East Linguistic Society, University of Massachusetts Amherst. Slides available (August 2019) at https://blogs.umass.edu/jarosz/files/2016/10/NELS2016_final_red.pdf.Google Scholar
Kenstowicz, Michael & Kisseberth, Charles (1971). Unmarked bleeding orders. Studies in the Linguistic Sciences 1:1. 8–28.Google Scholar
Kim, Yun Jung (2012). Do learners prefer transparent rule ordering? An artificial language learning study. CLS 48:1. 375–386.Google Scholar
Kiparsky, Paul (1968). Linguistic universals and linguistic change. In Bach, Emmon & Harms, Robert T. (eds.) Universals in linguistic theory. New York: Holt, Rinehart & Winston. 170–202.Google Scholar
Kiparsky, Paul (1971). Historical linguistics. In Dingwall, William Orr (ed.) A survey of linguistic science. College Park: University of Maryland Linguistics Program. 576–642.Google Scholar
Kiparsky, Paul (2000). Opacity and cyclicity. The Linguistic Review 17. 351–365.CrossRefGoogle Scholar
Kirov, Christo (2017). Recurrent neural networks as a strong domain-general baseline for morpho-phonological learning. Poster presented at the 91st Annual Meeting of the Linguistic Society of America, Austin. Available (August 2019) at https://ckirov.github.io/papers/lsa2017.pdf.Google Scholar
Kirov, Christo & Cotterell, Ryan (2018). Recurrent neural networks in linguistic theory: revisiting Pinker & Prince (1988) and the past tense debate. Transactions of the Association for Computational Linguistics 6. 651–666.CrossRefGoogle Scholar
Kurtz, Kenneth J. (2007). The divergent autoencoder (DIVA) model of category learning. Psychonomic Bulletin and Review 14. 560–576.CrossRefGoogle Scholar
Li, Jiwei, Chen, Xinlei, Hovy, Eduard & Jurafsky, Dan (2015). Visualizing and understanding neural models in NLP. ArXiv Preprint. http://arxiv.org/pdf/1506.01066.pdf.Google Scholar
Luce, R. Duncan (1959). Individual choice behavior: a theoretical analysis. New York: Wiley.Google Scholar
McCarthy, John J. (1999). Sympathy and phonological opacity. Phonology 16. 331–399.CrossRefGoogle Scholar
McCarthy, John J. (2007). Hidden generalizations: phonological opacity in Optimality Theory. Sheffield & Bristol, Conn.: Equinox.Google Scholar
McCoy, R. Thomas, Frank, Robert & Linzen, Tal (2018). Revisiting the poverty of the stimulus: hierarchical generalization without a hierarchical bias in recurrent neural networks. ArXiv Preprint. http://arxiv.org/pdf/1802.09091.pdf.Google Scholar
Moreton, Elliott & Pater, Joe (2012a). Structure and substance in artificial-phonology learning. Part 1: Structure. Language and Linguistics Compass 6. 686–701.CrossRefGoogle Scholar
Moreton, Elliott & Pater, Joe (2012b). Structure and substance in artificial-phonology learning. Part 2: Substance. Language and Linguistics Compass 6. 702–718.CrossRefGoogle Scholar
Moreton, Elliott & Pertsova, Katya (2016). Implicit and explicit processes in phonotactic learning. In Scott, Jennifer & Waughtal, Deb (eds.) Proceedings of the 40th Annual Boston University Conference on Language Development. Somerville, Mass.: Cascadilla. 277–290.Google Scholar
Nazarov, Aleksei & Pater, Joe (2017). Learning opacity in Stratal Maximum Entropy Grammar. Phonology 34. 299–324.CrossRefGoogle Scholar
Pullum, Geoffrey K. (1976). The Duke of York gambit. JL 12. 83–102.CrossRefGoogle Scholar
Pycha, Anne, Nowak, Pawel, Shin, Eurie & Shosted, Ryan (2003). Phonological rule-learning and its implications for a theory of vowel harmony. WCCFL 22. 423–435.Google Scholar
R Core Team (2016). R: a language and environment for statistical computing. Vienna: R Foundation for Statistical Computing. http://www.r-project.org.Google Scholar
Rahman, Fariz (2016). seq2seq: sequence to sequence learning with Keras. Available (August 2019) at https://github.com/farizrahman4u/seq2seq.Google Scholar
Rasin, Ezer, Berger, Iddo, Lan, Nur & Katzir, Roni (2017). Rule-based learning of phonological optionality and opacity. Paper presented at the 48th Meeting of the North East Linguistic Society, University of Iceland. Available (August 2019) at http://www.mit.edu/~rasin/files/abstracts/nels2017.pdf.Google Scholar
Rossum, Guido van (1995). Python tutorial. Amsterdam: Centrum voor Wiskunde en Informatica.Google Scholar
Sanders, Nathan (2003). Opacity and sound change in the Polish lexicon. PhD dissertation, University of California, Santa Cruz.Google Scholar
Sutskever, Ilya, Vinyals, Oriol & Le, Quoc V. (2014). Sequence to sequence learning with neural networks. Advances in Neural Information Processing Systems. 3104–3112. Available (August 2019) at https://arxiv.org/abs/1409.3215v3.Google Scholar
Tieleman, Tijmen & Hinton, Geoffrey (2012). Lecture 6.5-rmsprop: divide the gradient by a running average of its recent magnitude. COURSERA: Neural Networks for Machine Learning 4. 26–31.Google Scholar
Wilson, Colin (2006). Learning phonology with substantive bias: an experimental and computational study of velar palatalization. Cognitive Science 30. 945–982.Google ScholarPubMed
Zuraw, Kie (2003). Probability in language change. In Bod, Rens, Hay, Jennifer & Jannedy, Stefanie (eds.) Probabilistic linguistics. Cambridge, Mass.: MIT Press. 139–176.Google Scholar
Supplementary material: File

Prickett supplementary material

Prickett supplementary material 1

Download Prickett supplementary material(File)
File 155.4 KB
Supplementary material: File

Prickett supplementary material

Prickett supplementary material 2

Download Prickett supplementary material(File)
File 6.2 KB
Supplementary material: File

Prickett supplementary material

Prickett supplementary material 3

Download Prickett supplementary material(File)
File 29.4 KB