A Guide to Analysis in
MaxEnt Optimality Theory
by Bruce
Hayes and Claire
Moore-Cantwell
Department of
Linguistics
UCLA
This is a book in progress that we are distributing in
draft form for comment. Right now you can download the table of contents
and the first two chapters.
Download
the text
Description
This document consists of the first two chapters of a
projected textbook for Maximum Entropy Optimality Theory, a.k.a. “MaxEnt”.
The text is directed to beginners (i.e. to linguists who are new to
MaxEnt) and is meant as a “how-to” book, mostly pedagogical, and only
occasionally polemical.
The book is rooted in our view that there are many
linguists who are in possession of gradient data (e.g. from observations
of free variation, corpora, or experiments) and wish to develop a
rigorous generative analysis of their findings — but lack an appropriate
formal framework for doing so. We feel that MaxEnt is a very useful such
framework, and hope to make it more accessible to researchers by making
it easy to learn, and by offering practical advice based on our own
research experience. We show how MaxEnt can be done effectively using
Excel spreadsheets, and offer example analyses along with spreadsheet
implementations.
Right now we are posting just Chapters 1 and 2, but hope to expand this
soon, posting versions (see Contents) of the remaining seven chapters.
Chapter 2 is the central tutorial chapter, and could be used, we think,
as a reading in a graduate course.
We would be grateful for feedback, including suggestions for citation
(bhayes@humnet.ucla.edu, moore.cantwell@ucla.edu).
Supplementary materials
This is mostly Excel spreadsheets on which the analyses
are calculated. You can use them as a study aid, or adapt them for your
own analyses.
Download
(zip file)
Back to Bruce Hayes's home page
Last revised 1 February 2025