In lieu of an abstract, here is a brief excerpt of the content:

Reviewed by:
  • Doing optimality theory
  • Eulàlia Bonet
Doing optimality theory: Applying theory to data. By John J. McCarthy. Oxford: Blackwell, 2008. Pp. xxii, 310. ISBN 9781405151351. $74.30 (Hb).

This book is a textbook, but not precisely the standard type of textbook. Its topic is optimality theory (OT), to which McCarthy already devoted a textbook in 2002. While the earlier book was a description of the OT framework, the new book focuses on the issues that arise in doing and presenting research within OT. The main justification for the book is offered in the following observation: ‘Many of OT’s basic premises are very different from other linguistic theories. This means that OT requires new and often unfamiliar ways of developing analyses, arguing for them, and even writing them up’ (x). Given the number of topics covered and the way they are dealt with in the book, including the exercises provided and questions posed at the end of many sections, the potential reader ranges from the graduate (even undergraduate) student to the experienced researcher. One gets the impression that in this book M is trying to suggest solutions to many of the problems, even ethical problems, that have come up in his own teaching and research, as well as in discussions with colleagues and in reading papers by others. This makes the book very unusual but at the same time very useful. Even people working in other frameworks or even outside linguistics altogether will find in it helpful advice. The intricacies of OT are presented gradually and very clearly. Though many different languages are used by way of illustration, Yawelmani appears as a leitmotiv throughout the text, with very simple facts being presented at the beginning (very first example), and other aspects being introduced through exercises and again in the text in later sections (to almost the last one). This is an effective strategy because the reader can concentrate on the theoretical points being illustrated rather than investing time in trying to grasp the complexities of different languages. Though most of the issues discussed are illustrated with examples from phonology, other examples are drawn from syntax. The book reads very smoothly and is written in an informal, almost novel-like, style. For example, what in other books is often called the ‘Preface’ here has the title ‘Read this first!’.

Ch. 1, ‘An introduction to optimality theory’, is meant for people who are not familiar with OT but will still be of interest to those who are. It addresses issues like the universality of constraints, the reasons for their violable character, and the worries of researchers who are concerned by the large number of candidates that Gen generates.

Ch. 2, ‘How to construct an analysis’, is by far the longest chapter. Although it starts with basic issues like how to find a topic for a term paper, and mentions panic situations that can arise in the process of constructing an analysis, it discusses many of the problems that even a sophisticated researcher can encounter, from appropriate ways of formulating generalizations and constraints to finding ranking arguments and relevant candidates. Especially interesting is the discussion around richness of the base, and it is significant that lexicon optimization is relegated to a mere footnote (n. 16). M also shows how detecting ranking inconsistencies or finding the appropriate constraint ranking can be done with useful methods, like the recursive constraint demotion algorithm (RCD) and elementary ranking condition (ERC) entailment, which are explained very clearly in this chapter. Computer software that can help at this stage of an analysis, like OTSoft (Hayes et al. 2003), is also described.

Ch. 3, ‘How to write up an analysis’, is the sort of chapter that one would never find in a linguistics textbook, but it is certainly useful. In most OT work the ultimate goal is to come up with the set of constraints and the constraint ranking adequate for a given set of data, a result that can be presented in a single tableau. For this reason, it is not always easy for a researcher to see how his/her analysis should be conveyed to the reader. M uses what he has shown about Yawelmani...


Additional Information

Print ISSN
pp. 942-944
Launched on MUSE
Open Access
Back To Top

This website uses cookies to ensure you get the best experience on our website. Without cookies your experience may not be seamless.