1Corpus linguistics, a field defined by Sinclair [1996] as “a collection of language data selected and organized according to explicit linguistic criteria to serve as a sample of language,” has become a cornerstone of contemporary linguistic research. Emerging in the late 1950s with the creation of the Survey of English Usage [1959], corpus linguistics was conceptualized two decades later, leading to distinctions between corpus-based studies – those verifying existing linguistic theories – and corpus-driven research, which generates new insights directly from empirical data (Sinclair [2004]; Tognini-Bonelli [2001]). This duality has proven essential for understanding lexical patterns and collocations, a core focus of this book, which explores corpus-based analysis in second language (L2) acquisition and pedagogy.
2The book under review, part of the Corpus Linguistics series, benefits from an advisory board that includes leading scholars such as Paul Baker (Lancaster University) and Jesse Egbert (Northern Arizona University). Authored by Pawel Szudarski1 (University of Nottingham), an expert in applied linguistics and corpus-driven vocabulary acquisition, it is divided into five sections spanning 92 pages, with 14 pages of references (pp. 76-89) and appendices reproducing the corpora used throughout the research.
3The introduction (pp. 1-3) outlines the scope, aims, and research questions of the book, emphasizing its relevance to both linguistic research and practical language teaching. Chapter 2 (pp. 3-24) begins by examining various definitions related to the complex concept of collocations, contextualizing them alongside related terms such as co-occurrence, formulaicity, phraseological patterns, idioms, binomials, and lexical bundles. The author defines collocations as “a type of formulaic language” (p. 4) identified using both phraseological and corpus-based criteria. Section 2.4 focuses on the role of corpora in collocation research, exploring key corpus-related measures for identifying word relationships and describing how collocational words interact through both textual and psycholinguistic dimensions.
4Section 3 (pp. 24-42) reviews the theoretical approaches and research methodologies used in corpus-based collocation studies, covering quantitative, qualitative, and mixed-method analyses. Szudarski highlights the complementary nature of these methods, noting that quantitative approaches provide essential statistical data, while qualitative methods allow for a more nuanced and contextualized interpretation of the findings (p. 25). He also skillfully addresses specific methodological elements related to collocation measures, such as frequency-based ranking, collocational span, and proximity of units (compactness), before detailing more complex statistical measures like t-score, Mutual Information (MI), LogDice, and Delta P. The author demonstrates the practical application of these measures through tools such as SketchEngine (Kilgariff) and GraphCall (Brezina), which facilitate grounded collocation analysis using the aforementioned statistical models.
5Apart from the tools mentioned, it exists yet another powerful software tool for collocations exploration. Hyperbase, developed by the Université Côte d’Azur (France), is a tool, which Vanni & Mittmann [2016] qualify as essential for studying collocations through both first and second-degree relations. The software’s adaptability lies in its ability to visually map co-occurrence networks, enabling users to uncover hidden lexical structures and their relations. Its “Thème” feature identifies overused lexicon around key expressions, while its adjustable specificity threshold (default set at 28.50) allows for varying levels of collocational density and complexity. By enabling users to modify the threshold, Hyperbase supports diverse analytical needs, making it a powerful interactive tool for identifying meaningful patterns in large-scale collocation networks.
6Section 4 (pp. 42-68) shifts focus to the application of corpus-based research in L2 collocation learning and teaching, addressing its intersection with second language acquisition (SLA) and language pedagogy. Szudarski aptly discusses usage-based approaches, emphasizing how learner corpora contribute to understanding lexical development, interlanguage accuracy, and sophistication. The section highlights key ways in which corpora assess learner production and contrast it with native-like collocational performance. While the section provides valuable examples of corpus-based studies, Szudarski devotes only a brief discussion (2 pages, subsection 4.3.3, pp. 65-66) to corpus-based teaching resources defined as “tools that draw on affordances and functionalities of computer-assisted language learning (CALL)”. Given the book’s stated aim of promoting corpus-based tools for pedagogical purposes, this limited exploration is somewhat disappointing. Although the author mentions ColloCaid, a tool for aiding L2 learners in real-time writing, its placement under “technology-enhanced L2 research” blurs distinctions between corpus-informed teaching and general technological tools.
7A key distinction made by Szudarski concerns the indirect and direct use of corpora in language teaching. Indirect corpus use involves the development of teaching materials, such as vocabulary lists, based on corpus findings. This approach focuses on structured, classroom-external instruction. Direct corpus use, or data-driven learning (DDL), involves engaging students directly with corpus data during lessons through hands-on activities such as analyzing concordance lines to identify collocational patterns. Szudarski rightfully acknowledges that the successful implementation of DDL depends heavily on teachers’ corpus literacy and readiness to adopt corpus-informed methods. As current research (Chambers [2019]; Koester [2022]) indicates, bridging the gap between corpus research and practical teaching requires targeted teacher training and accessible pedagogical resources.
8Section 5 offers an in-depth discussion of the book’s conclusions and future directions for corpus-based collocation research. Szudarski brings to the fore the importance of balancing group-based trends with individual learner needs, as noted by Bell [2009], to enhance lexical acquisition. He highlights the need for replication studies (Gries [2012]) to validate previous findings and recommends mixed-method approaches that combine quantitative corpus data with qualitative classroom interaction analyses for a more holistic understanding. The chapter also calls for expanding corpus research beyond English to include other languages and cultural contexts (Gabryś-Barker [2022]). Importantly, the author stresses ongoing collaboration between researchers and practitioners to adapt corpus-based research for real-world pedagogical use. One of the recent examples of this effort is a PhD thesis by Nicol-Bakaldina [2023]2 aimed to examine how corpus-driven insights can support teachers in designing content-specific language instruction across disciplines.
9Szudarski’s book is a comprehensive resource that successfully integrates theoretical frameworks with practical applications for collocation research. Through detailed explanations of corpus measures, statistical tools, and methodological practices, it provides a strong foundation for researchers and educators. However, the somewhat limited discussion on corpus-based classroom resources leaves room for further exploration. Despite this minor drawback, the book remains an essential reference for those seeking to apply corpus-based methods in L2 vocabulary teaching and beyond. Its emphasis on both empirical rigor and pedagogical potential highlights its significance for linguists, language instructors, and SLA researchers.