Quantitative Methods in Linguistics


Keith Johnson
Department of Linguistics
University of California, Berkeley
December, 2006

Introduction

Increasingly, linguists handle quantitative data in their research. Phoneticians, sociolinguistics, psycholinguists and computational linguists deal in numbers and have for decades. Now also, phonologists, syntacticians and historical linguists are finding linguistic research to involve quantitative methods. Consequently, mastery of quantitative methods is becoming a vital component of linguistic training.

This book has two foci. First, we will introduce and discuss general strategies and methods of quantitative analysis as they apply in several subdisciplines in linguistics. Second, the book provides detailed instruction in practical aspects of handling quantitative linguistic data, using a particular statistical package (R) to discover patterns in quantitative data and to test linguistic hypotheses. After two introductory chapters, the book is divided into chapters by subdiscipline of linguistics, though the methods presented in any one chapter are likely to be relevant in almost any other subdiscipline. So, though t-test is presented in the phonetics chapter, for example, you would certainly want to make use of t-tests in psycholinguistics.


Table of Contents

Front matter - acknowledgements, design of the book, table of contents.

1. Fundamentals of quantitative analysis -- Observations, distributions, central tendency, variability.
2. Patterns and tests -- Probability density functions, statistical inference/hypothesis testing, correlation.
3. Phonetics -- two-sample and paired t-tests, multiple regression, principle components analysis.
4. Psycholinguistics -- Analysis of variance, between groups and within groups factors, repeated measures.
5. Sociolinguistics -- Chi-squared, logistic regression.
6. Historical Linguistics -- Lexicostatistics, language similarity, clustering, and multidimensional scaling.
7. Syntax -- Magnitude estimation, linear mixed effects models, mixed effects logistic regression.

References

Appendix 1: Getting started with R
Appendix 2: Data sets and scripts used in examples and exercises.


Useful links

  • Download the R statistical package from the R Project homepage - follow the link CRAN to the download mirror sites.
  • While you are at the R Project site, be sure to look at the manuals page as a starting point for documentation.
  • I found "Notes on the use of R for psychology experiments and questionnaires" by Jonathan Baron and Yuelin Li to be particularly useful.
  • Get Isidore Dyen's IndoEuropean language similarity matrix and read about the Dyen, Kruskal & Black lexicostatistical work at their project page.
  • Find out more about the use of magnitude estimation in syntactic research at the web_exp project at the University of Edinburgh.
  • The National Institute for Science and Technology has a useful e-Handbook of Statistical Methods.

    A note about Software

    One thing that you should be concerned with in using a book that devotes space to learning how to use a particular software package is that software changes at a relatively rapid pace.

    In this book, I chose to focus on a software package (called "R") that is developed under the GNU license agreement. This means that the software is maintained and developed by a user community and is distributed not for profit (students can get it on their home computers at no charge). It is serious software. Originally developed at AT&T Bell Labs, it is used extensively in medical research, engineering, and science. This is significant because GNU software (like Unix, Java, C, Perl, etc.) is more stable than commercially available software - revisions of the software come out because the user community needs changes, not because the company needs cash. There are also a number of electronic discussion lists and manuals covering various specific techniques using R. You'll find these resources at the R project web page.


    Biographical note

    I have a PhD in Linguistics from Ohio State University (1988) and was a post-doctoral fellow at Indiana University in Psychology, and at UCLA in Linguistics. I was on the research staff in the medical school at the University of Alabama, Birmingham for a year before moving back to Ohio State to teach linguistics (1994-2004). In 2005 I joined the Department of Linguistics at UC Berkeley, where I currently direct the Phonology Lab.

    I have edited two volumes of research papers in phonetics and phonology (Johnson & Mullennix, 1997 and Hume & Johnson, 2001 - both published with Academic Press). Early in my career I was an editor for the 3rd edition of Language Files, an introductory linguistics text published by the Department of Linguistics at Ohio State (Bissantz & Johnson, 1985), and I have always had an interest in what makes for a good textbook. My Acoustic and Auditory Phonetics (2002) seems to have been a success thanks to input from many students and fellow teachers, and I was pleased with how that book helped open up acoustic phonetics to phonologists who might not have otherwise taught acoustics in their courses. I hope that a book on quantitative methods may help to move Lingusitics as a discipline toward greater explicitness and rigor.