TY - GEN
T1 - Exploring the relationship between learnability and linguistic universals
AU - Rafferty, Anna N.
AU - Griffiths, Thomas L.
AU - Ettlinger, Marc
N1 - Funding Information:
Acknowledgements. This work was supported by an NSF Graduate Research Fellowship to ANR, grant number BCS-0704034 from the NSF to TLG, and grant number T32 NS047987 from the NIH to ME.
Publisher Copyright:
© 2011 Association for Computational Linguistics
PY - 2011
Y1 - 2011
N2 - Greater learnability has been offered as an explanation as to why certain properties appear in human languages more frequently than others. Languages with greater learnability are more likely to be accurately transmitted from one generation of learners to the next. We explore whether such a learnability bias is sufficient to result in a property becoming prevalent across languages by formalizing language transmission using a linear model. We then examine the outcome of repeated transmission of languages using a mathematical analysis, a computer simulation, and an experiment with human participants, and show several ways in which greater learnability may not result in a property becoming prevalent. Both the ways in which transmission failures occur and the relative number of languages with and without a property can affect whether the relationship between learnability and prevalence holds. Our results show that simply finding a learnability bias is not sufficient to explain why a particular property is a linguistic universal, or even frequent among human languages.
AB - Greater learnability has been offered as an explanation as to why certain properties appear in human languages more frequently than others. Languages with greater learnability are more likely to be accurately transmitted from one generation of learners to the next. We explore whether such a learnability bias is sufficient to result in a property becoming prevalent across languages by formalizing language transmission using a linear model. We then examine the outcome of repeated transmission of languages using a mathematical analysis, a computer simulation, and an experiment with human participants, and show several ways in which greater learnability may not result in a property becoming prevalent. Both the ways in which transmission failures occur and the relative number of languages with and without a property can affect whether the relationship between learnability and prevalence holds. Our results show that simply finding a learnability bias is not sufficient to explain why a particular property is a linguistic universal, or even frequent among human languages.
UR - http://www.scopus.com/inward/record.url?scp=85121636852&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85121636852&partnerID=8YFLogxK
M3 - Conference contribution
AN - SCOPUS:85121636852
T3 - Proceedings of the Annual Meeting of the Association for Computational Linguistics
SP - 49
EP - 57
BT - Workshop on Cognitive Modeling and Computational Linguistics, CMCL 2011 at the 49th Annual Meeting of the Association for Computational Linguistics
A2 - Keller, Frank
A2 - Reitter, David
PB - Association for Computational Linguistics (ACL)
T2 - 2nd Workshop on Cognitive Modeling and Computational Linguistics, CMCL 2011 at the 49th Annual Meeting of the Association for Computational Linguistics: Human Language Technologies, ACL-HLT 2011
Y2 - 23 June 2011
ER -