Machine Learning Models of Universal Grammar Parameter Dependencies

Dimitar Lubomirov Kazakov; Guido Cordoni; Andrea Ceolin; Monica-Alexandrina Irimia; Shin-Sook Kim; Dimitrios Michelioudakis; Nina Radkevich; Cristina  Guardiano; Giuseppe Longobardi

doi:10.26615/978-954-452-040-3_005

Machine Learning Models of Universal Grammar Parameter Dependencies

Dimitar Lubomirov Kazakov, Guido Cordoni, Andrea Ceolin, Monica-Alexandrina Irimia, Shin-Sook Kim, Dimitrios Michelioudakis, Nina Radkevich, Cristina Guardiano, Giuseppe Longobardi

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

Abstract

The use of parameters in the description of natural language syntax has to balance between the need to discriminate among (sometimes subtly different) languages, which can be seen as a cross-linguistic version of Chomsky’s (1964) descriptive adequacy, and the complexity of the acquisition task that a large number of parameters would imply, which is a problem for explanatory adequacy. Here we present a novel approach in which a machine learning algorithm is used to find dependencies in a table of parameters. The result is a dependency graph in which some of the parameters can be fully predicted from others. These empirical findings can be then subjected to linguistic analysis, which may either refute them by providing typological counter-examples of languages not included in the original dataset, dismiss them on theoretical grounds, or uphold them as tentative empirical laws worth of further study.

Original language	English
Title of host publication	Proceedings of The Knowledge Resources for the Socio-Economic Sciences and Humanities Workshop
Pages	31-37
Number of pages	7
DOIs	https://doi.org/10.26615/978-954-452-040-3_005
Publication status	Published - 1 Sept 2017

Bibliographical note

This is an author-produced version of the published paper. Uploaded with permission of the publisher/copyright holder. Further copying may not be permitted; contact the publisher for details

Access to Document

10.26615/978-954-452-040-3_005

learningdependencies-CRC-final
This is an author-produced version of the published paper. Uploaded with permission of the publisher/copyright holder. Further copying may not be permitted; contact the publisher for details
Accepted author manuscript, 368 KBLicence: Unspecified

Cite this

@inproceedings{26c732b4933749f582d679e9fbde8ca9,

title = "Machine Learning Models of Universal Grammar Parameter Dependencies",

abstract = "The use of parameters in the description of natural language syntax has to balance between the need to discriminate among (sometimes subtly different) languages, which can be seen as a cross-linguistic version of Chomsky{\textquoteright}s (1964) descriptive adequacy, and the complexity of the acquisition task that a large number of parameters would imply, which is a problem for explanatory adequacy. Here we present a novel approach in which a machine learning algorithm is used to find dependencies in a table of parameters. The result is a dependency graph in which some of the parameters can be fully predicted from others. These empirical findings can be then subjected to linguistic analysis, which may either refute them by providing typological counter-examples of languages not included in the original dataset, dismiss them on theoretical grounds, or uphold them as tentative empirical laws worth of further study.",

author = "Kazakov, {Dimitar Lubomirov} and Guido Cordoni and Andrea Ceolin and Monica-Alexandrina Irimia and Shin-Sook Kim and Dimitrios Michelioudakis and Nina Radkevich and Cristina Guardiano and Giuseppe Longobardi",

note = "This is an author-produced version of the published paper. Uploaded with permission of the publisher/copyright holder. Further copying may not be permitted; contact the publisher for details",

year = "2017",

month = sep,

day = "1",

doi = "10.26615/978-954-452-040-3_005",

language = "English",

isbn = "9789544520403",

pages = "31--37",

booktitle = "Proceedings of The Knowledge Resources for the Socio-Economic Sciences and Humanities Workshop",

}

TY - GEN

T1 - Machine Learning Models of Universal Grammar Parameter Dependencies

AU - Kazakov, Dimitar Lubomirov

AU - Cordoni, Guido

AU - Ceolin, Andrea

AU - Irimia, Monica-Alexandrina

AU - Kim, Shin-Sook

AU - Michelioudakis, Dimitrios

AU - Radkevich, Nina

AU - Guardiano, Cristina

AU - Longobardi, Giuseppe

N1 - This is an author-produced version of the published paper. Uploaded with permission of the publisher/copyright holder. Further copying may not be permitted; contact the publisher for details

PY - 2017/9/1

Y1 - 2017/9/1

N2 - The use of parameters in the description of natural language syntax has to balance between the need to discriminate among (sometimes subtly different) languages, which can be seen as a cross-linguistic version of Chomsky’s (1964) descriptive adequacy, and the complexity of the acquisition task that a large number of parameters would imply, which is a problem for explanatory adequacy. Here we present a novel approach in which a machine learning algorithm is used to find dependencies in a table of parameters. The result is a dependency graph in which some of the parameters can be fully predicted from others. These empirical findings can be then subjected to linguistic analysis, which may either refute them by providing typological counter-examples of languages not included in the original dataset, dismiss them on theoretical grounds, or uphold them as tentative empirical laws worth of further study.

AB - The use of parameters in the description of natural language syntax has to balance between the need to discriminate among (sometimes subtly different) languages, which can be seen as a cross-linguistic version of Chomsky’s (1964) descriptive adequacy, and the complexity of the acquisition task that a large number of parameters would imply, which is a problem for explanatory adequacy. Here we present a novel approach in which a machine learning algorithm is used to find dependencies in a table of parameters. The result is a dependency graph in which some of the parameters can be fully predicted from others. These empirical findings can be then subjected to linguistic analysis, which may either refute them by providing typological counter-examples of languages not included in the original dataset, dismiss them on theoretical grounds, or uphold them as tentative empirical laws worth of further study.

U2 - 10.26615/978-954-452-040-3_005

DO - 10.26615/978-954-452-040-3_005

M3 - Conference contribution

SN - 9789544520403

SP - 31

EP - 37

BT - Proceedings of The Knowledge Resources for the Socio-Economic Sciences and Humanities Workshop

ER -