Until recently, the grammatical study of the Greek language has mostly been confined to the ancient and the present-day stages of its history; with a few exceptions, the medieval period has been left out. As the authors point out, ‘the gap in our systematic grammatical knowledge of Greek extends roughly from Late Antiquity until the creation of the Modern Greek state, i.e. has a span of around 1,500 years, a gap unparalleled in the historical study of any language’ (p. xvi). This gap has at last been triumphantly filled by the CGMEMG.

The Grammar is characterized by ample generosity: the generosity of the authors in providing a wealth of detailed analysis and a treasure-house of examples (presented in approximately chronological order within each section), even from texts that date from before the beginning of the appointed period 1100–1700 and from shortly after its end; and the generosity of the publisher in providing so much space for all this information. The authors’ names appear conveniently in alphabetic order, with Holton as director of the project and Horrocks as co-director at the top of the list.

The aims of the Grammar are set out as follows at the end of the General Introduction:

Volume 1 consists of the General Introduction, a sixty-page bibliography, and Part I of the main text (Phonology, 237 pages), while Volume 2 consists of Part II (Nominal morphology), which runs to more than 1000 pages. Each numbered Part of the Grammar is given its own separate introduction.

Volume 3 comprises Part III (Verb morphology), its 600 pages suggesting that, in Medieval Greek, nominal morphology is even more complex and diverse than verb morphology. The volume is divided into morphology proper and morphosyntax; the former covers synthetic (single-word) forms, while the latter analyses periphrastic forms, i.e. chiefly the future and perfect tenses. Fifty pages are devoted to an account of the multifarious ways of expressing future time and conditionality. The Grammar records that, in the absence of synthetic future-tense forms in Medieval Greek, there were as many as twenty-two different ways of expressing the future.

Lastly, Volume 4 consists of Part IV (almost 200 pages of Syntax) plus more than fifty pages of indices. At first sight it may seem disappointing that less than 10% of the whole Grammar is devoted to syntax – syntax has usually been an unjustly neglected sibling of morphology in traditional grammars of Greek – but plenty of syntax is covered in the copious sections (almost 200 pages) devoted to morphosyntax in Part III.

Begun in 2004, the CGMEMG has taken fifteen years to come to fruition. This is hardly surprising, considering that very little had already been published on the grammar of texts covering a period of six hundred years. Older studies were based on often unreliable editions and lacked the benefit of texts that have been published more recently; these include especially non-literary documents. The Grammar is based on an electronic corpus of three million words (xxv). In an effort to obviate editorial intervention the authors have consulted, where possible, manuscripts and diplomatic editions.

The period covered by the Grammar corresponds exactly to Kriaras’ (not yet completed) dictionary of medieval vernacular Greek. The authors point out that theirs is not a grammar of ‘Byzantine Greek’, nor does it cover the Early Medieval period (p. xix). Their periodization is based on linguistic rather than historical criteria.

As regards the place of Medieval Greek in the historical development of the language, the authors view Medieval Greek as an entity in itself, not as ‘a stage in the evolution of Greek in general’ (xvi). They rightly avoid ‘a teleological evolutionary view in the direction of Modern Greek’ (xxvii). Instead, although the CGMEMG is ‘a text-based descriptive grammar’ (xx), the authors set out to describe ‘[t]he linguistic form which (we believe) represents the innate grammatical system of a native speaker of Greek living in this period’ (xviii). Some scholars, myself included, have shown that this is at least partially possible in the study of the grammatical rules and communicative purposes behind the placement of the weak object pronoun before or after the verb, contrary to a formerly widespread belief that in Medieval Greek ‘anything goes’. However, the authors stress that, while theirs is ‘not a historical grammar of Greek’, ‘the perspective of the Grammar is […] not synchronic but diachronic’ (xxviii).

The authors divide their source texts into ‘literary’ and ‘non-literary’, and the former into prose and (the great majority) verse (xx-xxi). Almost all the published literary texts of the period, amounting to 360 texts of vastly varying lengths, have been taken into account. Of the total body of 1200 non-literary publications, half are notarial documents from Venetian-occupied areas: Crete, the Heptanese and the Cyclades (xxvi). The illustrative examples of usage illustrate the geographical and chronological distribution of each phenomenon, as well as registers and types of text. Non-literary texts are accompanied by their date and geographical provenance where these are known.

The authors avoid using the term ‘dialect’, since there was at the time no standard from of which a given form might have been a variation; instead they talk about ‘geographical areas’ (xxx). Due to the unfeasibility of quantitative analysis, the authors characterize the relative frequency of each grammatical phenomenon by using terms such as ‘regular, free, restricted, rare’ (xxviii). They also use the label ‘general’, but they avoid committing themselves as to whether or not there can be said to have been a spoken or written Greek Koine at any stage or in any geographical area during the period.

Whereas the printed version of Kriaras’ dictionary, from a certain volume onwards, (and its electronic version from the outset) has imposed the present-day monotonic system on the headwords and the quoted examples of usage (a policy about which I once took issue with the author himself), the authors of the Grammar are content, with some systematic exceptions, to quote their examples from edited texts using the orthography adopted by the editor. Still, I would have preferred it if they had distinguished between τσ [ts] and τζ [dz], since the use of the digraph τζ for both sounds, though justified on historical grounds, does not indicate how it was pronounced.

Constantinople is not among the regions specified on p. xix as being a major geographical source of examples, nor does it figure in the list of locations under ‘regional variation’ in the Index, though it does appear in the title of one subsection on p. 1949. One might have welcomed an explanation for the almost complete absence of the Βασιλεύουσα from the Grammar: does it imply that there is little regionally specific linguistic evidence to be found in texts from Constantinople, or that texts from Constantinople were composed in a non-regional variety – or neither? Although some eighteenth-century examples are included in the Grammar, from a linguistic point of view that century largely remains a terra incognita in the history of Greek. Should we think of it as being the final stage of ‘Early Modern Greek’, as Tasos Kaplanis has suggested,Footnote ¹ or (as the authors seem to imply) as the first stage of ‘Modern Greek’? On this score, it is interesting that the authors of the Grammar appear to have discovered no instance of the epistemic use of θέλει + personal verb. By contrast, eighteenth-century writers from Constantinople, who frequently use personal θέλω + infinitive for the future (θέλουν έρθει ‘they will come’), besides θενα and θα + personal verb, reserve impersonal θέλει + personal verb for epistemic use (θέλει ήρθαν ‘they must have come’). The term ‘epistemic’ is absent from the index of the Grammar, as are ‘probability’ and ‘possibility’. It would be good to know what constructions were available for the expression of probability during the period covered by the Grammar.

In a brief review it is impossible to do justice to such a monumental intellectual undertaking as the CGMEMG. Suffice it to say that technical language is always elucidated, and the volumes are impeccably edited; the number of typographical errors is infinitesimal given the length and complexity of the text.