Scott and Tribble's belief that drawing the attention of corpus users to texts is "the best starting point for a corpus-informed language pedagogy" (p. x) is the inspiration for this book. The authors intend to motivate the reader to see corpus data within the linguistic and social contexts that produced it. The book is divided into two sections. Part I (Chapters 1 to 5), written by Mike Scott, is resource and theory-based; and Part II (Chapters 6 to 10), written by Christopher Tribble, is methodologically oriented.
In Part I the reader is given help conceptualizing how "theory derives from pattern; pattern reflects theory" (Sinclair, 2004, p. 10) at both lexical and grammatical levels. In Chapter 1 Scott introduces corpus-based methods in studying corpora of texts or authentic language, and he emphasizes relationships between text patterns, four aspects used to frame a corpus analysis (Language, Text, Mind, and Culture), and contextual scope in language use. He subsequently devotes the rest of his section of the book to operationalizations. Starting with generating a word-list from a text or collection of texts (Chapter 2), a study can then, depending on its purpose, systematically produce different formats of a word-list: e.g., single-word or cluster (or lexical bundle in Biber et al., 1999), tagged or untagged (a corpus with word functions or without), different ordering schemes (e.g., alphabetical or frequency). Then, the notion of concordances or co-occurrences of a word is introduced (Chapter 3) to explain how a study concentrating on the linkages and associations between words in texts can be conducted. Following this, "Keyness" is clearly explained through demonstrations of a method for identifying key words in a text (Chapter 4) and in multiple texts (Chapter 5).
Part I has strong, comprehensible explanations of corpus-based theories. Scott's explanations, combined with a sizable number of visuals, including graphs, tables, figures, plots and computer screen shots, are so clear that the reader can easily conceptualize analytic frameworks in which corpus analysis tools (wordlists, concordances, key words) can be used to describe text patterns. Scott's writing style is effective. Beginning each of his chapters with a key quotation of a well-known corpus linguist helps the reader conceptualize the theory before reading the whole chapter. Ending each chapter by tying the conclusion back to the main idea of the quotation reminds the reader of the chapter's focus. Additionally, Scott effectively uses global metaphors to help the reader understand the theories. For example, he explains that within a school class person X and person Y can be "commonly found 20 or more positions apart, yet still within the same group. They collocate within the same text but keep away from each other" (p. 36).
In Part II the methodological studies illustrate how students language teachers and researchers can use corpus-based approaches to develop pedagogically valuable insights into language in use. Tribble starts his section by demonstrating the power of key-word analysis through an analysis of key words in small collections of spoken and written texts (Chapter 6), followed by his examining key words at a discourse level to spot the different moves of a word in a corpus (Chapter 7). Additionally, a cluster analysis is demonstrated as a useful investigation of contrast between published and unpublished writing (Chapter 8). The analysis can help build an understanding of an important aspect of how published texts are formed and the extent to which unpublished writing might be similar to or different from the published production. Key word analysis also helps identify themes in a corpus (Chapter 9). Furthermore, key word lists of some functional words (e.g., titles and pronouns) in combination with key words' collocations can be used to investigate the issue of gender balance and better explain the environments in which the gendered terms occur. Since a corpus-based analysis is claimed to be a systematic approach to explore distributional features, the book also shows how to employ corpus tools and computer software tools to analyze a single, unreliably patterned text (Chapter 10).
These demonstrative studies are very well organized and executed. Step-by-step procedures are extremely helpful and practical for apprentice corpus-analysts. With a good combination of detailed description and revisited explanation of theoretical frameworks, the studies become convincing and valuable for the reader. Similar to those in the first five chapters, the clear explanations in these chapters are supplemented with appropriate visual formats. As such, they can be good models of corpus-based analysis, models that teachers and learners can apply in their own investigations. The emphasis on the contributions of the corpus-based findings provides convincing evidence for pedagogic applications. For example, the findings from analyzing both three-word and four-word clusters in published texts can help apprentice writers learn more about writing styles in academic writing (Chapter 8).
It is a very good idea to provide studies which represent a variety of lexical and grammatical foci, text types, and discipline areas. The illustrations of how to analyze lexico-grammatical features in a corpus (e.g., that as a collocate of of in Chapter 6, key words related to who, what and where, and the key words' collocates in Chapter 9) successfully show an effective application of theories and frameworks to the methodology. The reader will additionally learn how to use different types of texts in corpus-based analyses (written texts in Chapters 6-10 and spoken texts in Chapter 6). The range of demonstrations will also help readers from different disciplines understand and appreciate an analysis of texts in their own fields (i.e., humanities and law and technology in Chapter 5, business in Chapter 7, literature MA dissertations in Chapter 8, news language in Chapter 9, short stories in Chapter 10, and general published/expert writing from disciplines in Chapters 6 and 8). I am certain that the book will help novice corpus researchers and teachers to either replicate the analyses or conduct their own.
The book is limited, both in corpus software and variety of English. The reader should keep in mind that only Wordsmith Tools software is used in the demonstrative analyses, and that British English corpora are highlighted in this book. Accordingly, readers should not limit their investigations to these models. Instead they should apply the corpus-based outcomes and insights obtained from the book to other available software and to their own Englishes.
Without question, Textual Patterns provides insights into theoretical and pedagogical implications of corpus linguistics for both apprentice and professional language educators. However, whether applying this data-rich sub-discipline of language study to chosen texts will be meaningful depends on the research or pedagogical purposes of each user. A corpus-based analysis is challenging because although corpus tools can expedite quantitative aspects of research, language-pattern interpretation remains in the hands of the study's conductor.