The development of scientific writing is an important aspect of English language studies and has drawn considerable attention over the past decades, with much of the emerging research in the field relying on corpus linguistics approaches. The ongoing project of the Coruña Corpus of English Scientific Writing (CC) is an important contribution to this enterprise, and the volume reviewed here is the central product of the Corpus of English Philosophy Texts (CEPhiT), which is the second part of the CC. CEPhiT was compiled for the description of philosophical texts written in English in the eighteenth and nineteenth centuries. This volume neatly discusses the compilation of the corpus and presents empirical research on the philosophical discourse and the sociolinguistic influence of the period, documenting the idea that ‘every scientific field is likely to have its own writing traditions and restrictions in terms of conventions’ (p. 1).

Academic writing in philosophy typically involves narratives containing ‘fictional tales’, ‘twin-Earth fantasies’ and ‘imaginary conversations’ (Bloor Reference Bloor, Ventola and Mauranen1996), which problematises issues of intersubjective positioning and sociolinguistic situations of philosophical writing. Exploring the historical changes of the discourse conventions in philosophy, the volume comprises nine chapters, which are organised into two different parts. Chapters 1 to 4 give an account of the general context under which the texts were produced, the theoretical and practical decisions on compilation as well as the essential characteristics that make CEPhiT different. Chapters 5 to 9 report empirical studies which explore various language features of CEPhiT texts. Thus, as a whole, the book contributes considerably to our knowledge of both the construction of a diachronic corpus of academic writing and the development of philosophical language.

Isabel Moskowich opens the volume by spelling out the compilation issues of CEPhiT in chapter 1, ‘Philosophers and scientists from the Modern Age: Compiling the Corpus of English Philosophy Texts’. These are governed by two basic principles, ‘balance and representativeness’ (p. 5). In terms of the time-span represented, the corpus includes the philosophical texts published between 1700 and 1900, which record ‘the movement away from philosophy as being enshrined in the words of unquestionable authorities to the opening up to new approaches’ (p. 3). The scholastic tradition underwent a radical transformation, confronted with the emerging new approaches of inductive reasoning and observational empiricism. These social and epistemological changes brought about the need for a new language in the transmission of philosophical knowledge, and it is this language development that is captured in the compilation and analysis of CEPhiT. In doing so, the corpus includes both informative genres, such as treatise and essay, and instructive ones, such as textbook and lecture, and also presents a balanced distribution of the gender and geographical information of the authors.

In chapter 2, ‘Genre categorisation in CEPhiT’, Begoña Crespo introduces the categorisation of text samples in CEPhiT, considering genre as a social and epistemological artefact. The compilation follows four main criteria of categorisation: (1) the author's choice and an explicit determination of what he or she was writing, as stated in various possible kinds of prefatory material; (2) the textual organisation to which the sample belongs, that is, its rhetorical construction, the purpose of writing the text, the target audience, the medium and the level of technicality; (3) the author's academic background and position in society; and (4) the addressee (p. 26). Accordingly, six genres are included: treatise, essay, lecture, article, textbook and dialogue. In the end, Crespo notes that ‘as society evolves and literacy increases, the formats to convey science also change so as to reflect the requirements of each particular moment in time’ (p. 39).

Gonzalo Camiña and Inés Lareo discuss the editorial policy of CEPhiT in chapter 3, ‘Editorial policy in the Corpus of English Philosophy Texts: Criteria, conventions, encoding and other marks’, by introducing the criteria, conventions and encoding issues involved. The aim of constructing the corpus is to achieve a balance between ‘showing the text in the way it originally was’ and ‘offering researchers the possibility of manipulating the data in the texts in an open, flexible and productive way’ (p. 45). Despite the difficulty caused by the antiquity of ancient texts and typewriting, they add and unify headers, fonts, titles, paragraphs and lines. XML is used to encode CEPhiT texts because of its ease of use, flexibility and cross-platform operability. Additionally, some editorial marks are also inserted in particular scenarios such as [unclear], [table] and [missing pages] to guarantee the exploitation of the corpus at its full potential.

After the compilation of CEPhiT, the implementation and application of corpus annotation and index are highly crucial to researchers, and these are elaborated by Andrew Hardie in chapter 4 (‘Infrastructure for analysis of the CEPhiT corpus: Implementation and application of corpus annotation and indexing’). For the advantages of facilitating ‘frequency counts’, ‘concordances’ and ‘the consistency of analysis’ while ‘avoiding the duplication of labour’ (p. 62), four types of corpus annotation are applied: spelling regularisation, part-of-speech tagging, semantic tagging and lemmatisation. To solve the multi-layered annotation of CEPhiT, Open Corpus Workbench (CWB) is used, but in the meantime the CWB-indexed corpus requires a large amount of technical know-how for effective use. However, CWB corpora can be accessed via CQPweb, which is a web-based graphical user interface system. Finally, Hardie comments that ‘by indexing the CEPhiT corpus within CQPweb, access to the multiple layers of annotation has been made much more straightforward and user-friendly’ (p. 74).

Marina Dossena reports the first empirical study on CEPhiT in chapter 5 (‘On the shoulders of giants: An overview on the discussion of science and philosophy in late Modern times’), by examining the main linguistic strategies employed by late Modern authors in CEPhiT and the Corpus of Modern Scottish Writing (CMSW) to present their source texts and express their attitude to them. As Dossena has found, previous scholars and literature are seldom referred to without certain qualification, and this rhetorically opens discursive space for the interpretation of the source proposition. For example, prior authors are praised when their findings are underlined, while they are criticised, albeit indirectly, when their opinions are incorporated into a counterargument. She further shows how the different patterns of co-occurrence of reporting verbs, adjectives, adverbs and conjunctions guide readers’ interpretation and elicit agreement. For instance, boosting adverbs such as certainly and clearly are much more frequent than hedges such as possibly and seemingly in both CEPhiT and CSMW, which ‘stresses the certainty, truth and validity of the prediction’ (p. 89) and persuasively engages readers in the alignment with the argument.

In chapter 6 ‘Abstractness as diachronic variation in CEPhiT: Biber's Dimension 5 applied’, Leida María Monaco investigates the diachronic change of abstractness degree in CEPhiT by drawing on Biber's Multi-Dimensional Analysis, in an attempt to detect variations across two centuries (1700-1898) and across some of the genres represented in the corpus. She finds that although the most frequent abstractness marker is agentless passives, particularly in the nineteenth-century essays, the high scores for past participle clauses and conjuncts contribute most to a high dimension score in all subsections of the corpus. The relative abundance of past participle clauses in CEPhiT indicates ‘a greater presence of syntactic compression in late Modern English philosophical writing, compared with present-day academic prose’ (p. 118). She further concludes that while CEPhiT is rather heterogeneous as far as genres and topics are concerned, presenting various degrees of specialisation, late Modern philosophy sees a general increase in abstractness across the two centuries, and essays appear markedly more abstract than treatises and lectures.

Elena Seoane focuses on the diachronic evolution of authorial presence and identity expression in late Modern English philosophical writing as presented in CEPhiT in chapter 7 (‘Authorial presence in late Modern English philosophical writing: Evidence from CEPhiT’), exploring the frequency and functions of impersonal passives, other impersonal-subject constructions and first and second person pronouns. She identifies a statistically significant increase in the use of self-mention across time together with a surprising shift in use from I to we, although all texts are single-authored. The analysis of impersonal passives and other impersonal-subject constructions as strategies of author avoidance shows that these are less frequent than self-mention and that the trend grows throughout the period. Intradisciplinary variation is also observed when comparing two texts whose authors have different academic backgrounds. Thus, as she comments, ‘the diverse experiences and biographies’ of disciplinary community members, ‘alongside their diverse goals and methods when engaging in writing’, also shape their discourse (p. 139), and this calls for future research to investigate the degree of interdisciplinary variation.

Francisco Alonso-Almeida & Inés Lareo take a closeup look at a single lexical item in chapter 8, ‘The status of seem in the nineteenth-century Corpus of English Philosophy Texts (CEPhiT)’, examining the meaning and functions of seem as used in the nineteenth-century texts of CEPhiT. Basically, seem in CEPhiT serves as both a copular verb and an evidential one. On the copular use, seem reflects a comparison and can be replaced by look like (as in the effect of impulse seems analogous to the motion of the stone in its particular direction), while it expresses a lesser degree of propositional commitment when implying the authorial intention to indicate the way in which information is acquired (cf. some degree of pleasure seems to accompany all our sensations). Additionally, the authors also identify some contexts in which epistemic meaning prevails and seem is used as a hedging device to indicate lack of commitment, by analysing seem to, it seems to, subjective it seems to, and it seems as if/that. Thus Alonso-Almeida & Lareo clearly demonstrate the epistemic and evidential meanings of seem as two distinct concepts serving different functions in philosophical writing.

In the final chapter, ‘Explaining the use of if. . . then. . . structures in CEPhiT’, Luis Puente Castelo discusses the evolution in the use and functions of if. . . then. . . structures in the CEPhiT eighteenth- and nineteenth-century philosophy texts. His study shows that if. . . then. . . structures do not exhibit any new nuance of meaning compared with simple conditional if clauses, but their use is favoured in particular syntactic environments, specifically ‘in ambiguous, long or complex contexts’ (p. 175). More significantly, the diachronic distribution of the structures presented in the study indicates two very different perspectives on their use. The old perspective, influenced by scholasticism, sees if. . . then. . . structures occur freely and wholly at the discretion of the author while the new view considers their use as driven by the syntactic context. Castelo concludes that the diachronic change of these two perspectives is likely to ‘chart a gradual process of specialization, one which runs in parallel with the demise of scholasticism’ (p. 179).

The book comes with a copy of CEPhiT on CD-ROM, which is certainly a bonus to its readers and is useful for scholars in need of a small reference corpus of academic writing. One problem with the volume is the lack of a consistent acronym for the Corpus of English Philosophy Texts throughout the volume, because almost all chapters mention its full name before using its acronym, CEPhiT. This is a minor quibble, which by no means overshadows all the strengths and benefits the volume holds for those who are interested in English scientific language and historical linguistics.

All in all, this volume collects contributions from project members and external collaborators which give a comprehensive discussion on the compilation and empirical research of CEPhiT; it is thus undoubtedly a useful reference work for both empirical and theoretical studies of diachronic change in scientific language. For one thing, the volume provides a detailed introduction to the editorial criteria, encoding practice, corpus annotation and query methods which concern almost every aspect of compiling a corpus. For another, by presenting well-designed studies on various linguistic and rhetorical features, the volume also wisely points to possibilities for further research on how writers craft social interaction with their readers. One of the fruitful directions is the ways writers explicitly recognise and engage readers into the discourse rather than manage their own performance of self (see Hyland & Jiang Reference Hyland and Jiang2016). Thus these potential research avenues provide a valuable addition to the growing number of similar diachronic projects (e.g. Taavitsainen & Pahta Reference Taavitsainen and Pahta2010; Biber & Gray Reference Biber and Gray2016).