The book under review aims to map the regional language variation in written American English. The study reported in the book is innovative and stands out from previous studies in dialectology in terms of its research data and its research methods. To be specific, it focuses on written English with corpus-based and statistical approaches, while most previous studies are devoted to spoken or casual English with methods such as questionnaire surveys and interviews (e.g. Kurath et al., Reference Kurath, Hansen, Bloch and Bloch1939; Carver, Reference Carver1987; Labov, Ash & Boberg, Reference Labov, Ash and Boberg2006). In addition, its findings, i.e. the newly identified American dialect regions, may challenge generally accepted theories in dialectology. Thus, the book may be of significance to studies of dialects and language variation.

The volume is comprised of seven chapters, including an introduction to American dialectology (Chapter 1), methods for data collection (Chapter 2) and data analyses (Chapters 3 to 5) in the study, results explanation (Chapter 6), and conclusion (Chapter 7).

Chapter 1 introduces dialectology, the traditional methods to examine dialects, and the reason for the present study. It first reviews previous studies in American dialectology, noting that they largely used methods such as questionnaire surveys and interviews. Also, those studies mostly examined lexical, or phonetic and phonological variations in spoken vernacular from various parts of the USA. Based on the findings, the author identified research gaps in the area, such as ‘the nature of the regional grammatical variation’ across America, ‘the nature of American dialect regions’, and ‘the development of American dialects’ (p. 13). At the end of this chapter, the author briefly introduces the significance, methods, and the structure of the book.

Chapter 2 focuses on the methodology and the data used in the study, such as corpus design, corpus compilation, and corpus dimensions. The author first outlines the advantages of the corpus-based approach over questionnaire surveys and interviews, such as its efficiency, reliability, generalizability, and the possibility to elicit abstract grammatical constructions in quantity from natural language. Then, the author introduces the criteria and methods to be used in the study for data selection. He also clarifies how the corpus data was collected, cleaned, sorted, and processed. Finally, the author introduces the corpus dimensions, including issues such as the geographic distribution, the year of publication, the size of corpus and sub-corpora, and the numbers of words, texts, and authors. The detailed description of the corpus offers the possibility for other researchers to replicate and expand the study.

Chapters 3 to 5 report on the statistical methods used to analyse the data, and to identify the individual and common patterns of the regional variations. Chapter 3 centres on the selection of alternation variables in grammatical variation and the plotting for each variable. It first introduces the alternation variable and its significance in dialect studies. The selection of alternation variables for the study and their measurement are discussed with maps for each alternation variable plotted and discussed in the order of word classes. As the author admits, the maps presented in this chapter could not clearly show or identify the regional patterns of the dialects. Thus, it seems unnecessary to devote so much space (approximately 50 pages, nearly one quarter of the body of the book) to discussing the regional pattern for each variable with the aid of maps. This is particularly true when the validity of such discussions may be challenged.

To address the possible limitations of the last chapter, the author uses the spatial analysis in Chapter 4 to identify the significant regional patterns of grammatical variation. The spatial analysis and its distinct features are first introduced, then two terms for spatial analysis, spatial stationary and spatial non-stationary, are distinguished. Statistical methods for the spatial analysis such as the local spatial autocorrelation analysis (a statistical method to identify significant spatial clustering patterns) and the Getis-Ord G_i statistic (a statistical method for the local spatial autocorrelation analysis used to plot the local spatial autocorrelation maps) are also discussed. The local spatial autocorrelation maps were plotted based on the spatial analysis. One obvious strength of these maps is that they show relatively significant variations, which can be used to recognize the regional pattern of each variation much more easily. Of course, as the author argues, what the local spatial autocorrelation map identifies is the individual regional pattern of each variation. That is, the spatial analysis still cannot compare different features of grammatical variation with each other and identify the significant regional patterns among all patterns.

To address the limitations of the spatial analysis discussed in the previous chapter, the author employs more sophisticated statistical methods for the analysis of the data in Chapter 5. Firstly, a factor analysis is used to identify the most important common regional patterns. Then, a fuzzy cluster analysis is conducted to identify major regions of American dialects. In contrast with the arguably sharp division between dialect regions in previous studies, the borders of dialect regions identified by the fuzzy cluster analysis are obscured when a membership weight is provided for each cluster, which accordingly allows for gradual changes between dialect regions.

Chapter 6 explains the results of the study from linguistic, social, and cultural perspectives. Firstly, the author argues that formality is an important linguistic reason for grammatical variation, while density of communication is a main social and cultural reason for dialects diverging. In addition, the author compares the results of the study with previous surveys and concludes that dialect regions change over time. He also predicts American dialect regions in future based on cultural tendencies and contemporary communication patterns. Interestingly, the author claims that the five dialect regions identified in the study correspond to the five main cultural centres in the USA. That is, social and cultural reasons play a significant part in the shaping of dialect regions (e.g. see Finegan & Biber, Reference Finegan, Biber, Biber and Finegan1994, Reference Finegan, Biber, Eckert and Rickford2001).

Chapter 7, the final chapter, summarises the methods, findings, contributions, and significance of the study. It also discusses the limitations of the study, and suggests that future studies on grammatical variation be expected in other registers and across different social groups.

The volume contributes to research in dialectology from several perspectives. First and foremost, it is the first study in regional linguistic variation or American dialectology that uses a corpus-based approach as its research method (p. 13). With the set of statistical analyses upon the corpus data, the findings of the study seem more reliable, comparable, and replicable. As the literature review in Chapter 1 finds, data analyses in previous dialect studies were largely based on the judgement of researchers. Hence, most of them are not easy to replicate nor to compare with each other. However, the more robust methods employed in the present study not only generate more objective and accurate results, but also serve as a model for researchers in the area to investigate regional variations in other areas or to conduct diachronic research in future.

In addition, it is also the first study to examine regional patterns of grammatical variation in standard written American English. The findings of the study expand the theory of American dialect regions with a five-way division in standard written English. Meanwhile, it challenges the assumption that dialect variations mostly exist in spoken English and that the nature of linguistic variation is stable across the country. More interestingly, it also finds that the dialect regions, to a large extent, correspond to the main cultural or administrative centres rather than reflecting historical settlements, a point widely discussed and accepted in previous studies.

Other possible contributions of the study include identifying a set of grammatical alternation variables for the examination of regional grammatical variation and plotting the map of American dialect regions. Such accomplishments may help facilitate research in related disciplines, such as communicative studies, geography, and sociology.

The book is not without its limitations. The most significant concern is that the study does not clearly identify the relationship between the social background of the corpus-data writers and their linguistic use. As the findings of the study reveal, the division of dialect regions specifically pertains to the background of the speakers or writers of the dialects. In other words, factors such as social status, class, and ethnicity of the speakers and writers may, to a large extent, exert influence on their linguistic use. Accordingly, inaccessibility to the background information limits the study from more in-depth examination of the roles that background factors may play in the regional variation of the dialects as well as the linguistic use of the dialect speakers/writers.

Secondly, the corpus data could be more balanced in terms of location and time. The corpus data is not always evenly distributed across the country, i.e. some areas, such as cities in western Nebraska, are less well represented than others. In addition, as explained in chapter 2, the data used for the study ranges from 2000 to 2013. However, 79% of the data was selected from newspapers published between 2005 and 2009. This uneven distribution may slightly blur the comparisons between some areas by date of publication. For example, some cities in the Northeast contain earlier publications while the others include texts published later.

Thirdly, the detailed process of identification of the grammatical alternation variables may be of interest to many linguistic researchers. The author, in Chapter 3, notes that the grammatical alternation variables were identified from two resources: a review of previous literature and an analysis of the corpus data. However, more detailed information on how the grammatical alternation variables were identified would be extremely helpful.

Finally, research on regional variation across time, registers, or between social groups is also expected in the future, since issues such as time, writing style, and social background of the language users, is equally likely to influence regional language use.

To summarise, the book is invaluable for its innovations in the use of corpus data and various robust statistical methods in dialectology as well as its contribution to the theory of American dialects. It will be of great interest to researchers in the areas of dialectology, linguistics, and geography, among others.