The 18th volume in the Studies in Corpus Linguistics series, this work is a welcome contribution to the fields of corpus linguistics and language teaching. The volume, based on Römer's dissertation, thoroughly describes the results of two related, corpus-driven studies that provide a detailed description of the progressive in natural use in addition to its use in so-called school English. The first study is a comparison of two spoken corpora of native-speaker British English (both conversation), whereas the second is an analysis of school English, a corpus of conversational English appearing in two English as a foreign language textbook series used in Germany. The beginning chapters (2 and 3) provide the theoretical rationale, followed by the methods and results (4–6), and the final chapters (7 and 8) conclude the volume with pedagogical implications and general conclusions.
Chapters 2 and 3 introduce the reader to the difference between corpus-based and corpus-driven studies. The study described here is corpus-driven (following the work of Sinclair; e.g., Sinclair, 1991), “meaning that the analysis starts from the corpus evidence and may lead to new insights into the language” (p. 3). Römer's approach is influenced by Mindt's (2000, 2002) model of a didactic grammar—a grammar based on a detailed corpus-driven investigation of the language that seeks to solve pedagogical problems of selection, progression, and presentation. Chapter 3 describes progressives in theoretical studies, including the work of Comrie (1976), as well as more empirical and corpus-based studies.
Chapter 4 begins with a description of the methodology before moving on to the results of the first study. The two corpora used for this part of the analysis include the spoken register of the British National Corpus (10,365,464 words) and the spontaneous speech register of the Bank of English (20,078,901 words). Römer chose to include two corpora to increase the generalizability of her results, and she discusses and displays the findings separately. To keep the study manageable, only the 100 most frequent verbs were analyzed, leading to an impressive number of concordance lines (over 10,000) that were analyzed for form and function. The data were first coded for progressive forms: past, present, present perfect, and past perfect. Additionally, other collocational features were coded, such as forms of to be, subjects, objects, prepositions, negation, and adverbials, among others. Next, the data were coded for functional characteristics and two central functional features were discovered: continuity and repeatedness. Seven other functions appeared in the data: general validity, politeness or softening, emphasis or attitude, shock or disbelief, gradual change and development, old and new habits, and framing. A further analysis combined the two corpora to investigate lexical-grammatical patterns, and the results are displayed by verb type in several useful and thorough tables.
Chapter 5 describes the methods and results for the second study. The corpus used in this study was built by the author and includes only those parts of the textbooks (from two widely used textbook series in Germany) that represent spoken language, equaling roughly 100,000 words. The same 100 verbs were used in this analysis, and the concordance lines were coded for all of the features found in the first study. Chapter 6 compares the results of the two studies, focusing on how conversational the school corpus was, and nicely demonstrates the mismatch between the results of the two studies. Here, the inadequacy of pedagogical descriptions is made quite evident.
Chapter 7 focuses primarily on pedagogical implications. Based on the results of the comparisons between the two corpora, Römer suggests making only moderate changes to the presentation of the progressive in textbooks until more research is available. Examples of such changes include using authentic examples to increase fluency and to expose learners to the messiness of natural language, teaching the typical to improve communicative competence, presenting the progressive in natural contexts, focusing on frequent functions, and shifting emphasis to lexis. The final chapter looks ahead to the future and calls for more descriptive studies of progressives in usage and for studies conducted from a SLA perspective.
Overall, this volume has many strengths. In addition to the impressive analysis of the data and the detailed presentation of results, this work succeeds in demonstrating how evidence from corpus linguistic research can be used to inform and change language teaching. However, as Römer mentions, we cannot be too quick to make drastic changes without further investigations from multiple points of view.
