Loanwords in classical languages have been investigated for a long time almost exclusively in one direction, i.e. from ancient Greek towards Latin. The prestige that Greek culture exerted in the Roman world made the influence of Greek on Latin almost inevitable. This accounts for the huge number of detailed contributions devoted to Greek loanwords present in Latin. Less known is that, parallel to this active investigation of Greek loanwords in Latin, there was also a no less active investigation of Latin loanwords in Greek during the nineteenth and twentieth centuries. Needless to say, such contributions, despite the relevance they could eventually have, especially for late antiquity, always gave the impression of being occasional. The situation has now radically changed thanks to D.'s volume on Latin loanwords in ancient Greek. The present volume, splendidly printed, is the necessary and long-awaited attempt to provide scholars with a systematic picture of Latin loanwords in ancient Greek.

D.'s contribution is an admirable and impressive piece of work, in which a supposedly minor topic finally receives the adequate and systematic treatment that it surely deserves. The lexicon collects and systematises the very rich material available from the beginning of Latin documentation until the sixth century ce, conventionally established as the end of ancient Greek.

After Chapter 1, which consists of a short introduction to this enterprise, Chapter 2 illustrates the parameters by means of which the study has been organised. Special attention is paid to the distinction between loanwords and codeswitches on the basis of notions such as frequency and integration.

The third chapter comprises a long list (pp. 23–502), alphabetically ordered, of Latin loanwords that entered the Greek lexicon. This long section is preceded by a key to the labels characterising each entry, i.e. if the words are genuine or not, the different spellings in which they occur, the quantity of the Latin words as well as the types of word formation and the frequency of these words. Each entry provides the following pieces of information: (1) whether the loanword is direct or indirect; (2) the date when it occurs for the first time; (3) the form or forms in which the loanword is attested; (4) the type of documents in which it occurs; and (5) an essential bibliography concerning the loanword under scrutiny. In the second part of the Lexicon D. provides a detailed analysis of the loanwords from a semantic point of view.

Chapters 4–12 represent a wonderful episode of the history of the relations between Greek and Latin. Chapters 4–7 describe the way in which Latin was integrated (or not integrated and why: Chapter 7) into Greek phonetically and morphologically, including an interesting chapter on the survival of loanwords into modern Greek. Chapters 8 and 9 thoroughly illustrate when and where, i.e. in what genres and what authors, the loanwords were used, whereas Chapter 10 describes which words have been borrowed, with a final short discussion about identity expressed through the terms ‘Roman’ and ‘Christian’. The short Chapter 11 has the interesting title of ‘Overall conclusions and remaining questions’. Chapter 12 comprises two appendices, on how to count the numbers of the figures and what counts for a word. At the end of this monumental work there are the abbreviations, an impressive bibliography, an index locorum and an index of Latin words.

D. explicitly aims the volume at Classicists. Obviously, the choice of aiming the volume at Classicists is legitimate even though it has the consequence, of which D. is aware, that on several occasions linguists who investigate the history of Latin and Greek would like to know more and have at their disposal supplementary information about numerous entries. For reasons of space, I mention one example that illustrates the advantages the volume could provide to linguists if it were aimed at their needs.

From the entry devoted to the loanword repudium readers are informed that it is a direct loan, i.e. taken directly from Latin and documented from the fourth century ce to the Byzantine period. The documents that witness this loan are papyri that have been found in different places and go back to different ages: the Papyrus Lipsiensis (fourth century ce) and two passages occurring in two different papyri from Oxyrhynchus (the first going back to the fourth-fifth centuries ce and the second to the sixth century ce). However, another crucial piece of information is given without the attention it deserves: repudium is documented as both ῥεπούδιον and ῥιπούδιον. At the beginning of the introduction D. clearly and correctly warns readers about the difficulties in ascertaining the different diastratic, diaphasic and diatopic values that have to be ascribed to the different forms occurring in the various types of texts. However, it should also be clear that the form ῥιπούδιον cannot be treated simply as a more or less relevant variant of the expected form ῥεπούδιον. Unfortunately, readers who wish to know in which of the three papyri the form ῥιπούδιον occurs have no other possibility but to check themselves the sources quoted by D. to ascertain where it occurs. Since such a form shows the same pretonic vowel as in Italian from the origin, i.e. ripudio (‘repudiation’), questions arise that are important for linguists. Should the occurrence of the form reminiscent of Italian be regarded as a pure coincidence? Or is it to be ascribed to a frequent tendency, manifesting itself in some areas of the Romania, of pretonic [e] to become [i]? This issue is particularly thorny. Especially during the late period of the Empire, the Roman army gathered a huge number not only of foot-soldiers, but also of sailors in the military navy, coming from different corners of the territory subdued by Rome, especially from the East. The conclusion is that the place where the various documents have been unearthed was only negligibly influential on the written texts, surely less relevant than the mother languages of their authors.

In the case of the variation between ῥεπούδιον and ῥιπούδιον there is the possibility that a tessera of a linguistic mosaic can be recovered where /e/ changed to /i/ in Romance, as in the case of πρασίδιον, for instance, instead of the reasonably expected πραισίδιον as borrowing of praesidium (‘garrison’), or we could simply face a mistake of the scribe. Nonetheless, for the sake of completeness, the possibility should be mentioned that the form πρασίδιον mirrors a pronunciation in which the diphthong [ae̯] had become [a]. It should be kept in mind that such a development is frequent in late and Vulgar Latin but not uncommon even in some Indo-European languages. Should we imagine that the scribe could have inadvertently reproduced a phonological feature of his Germanic mother tongue? Obviously, it would be unimaginable to present all the possible hypotheses that could account for a particular form attested in the documents, but some additional information could have helped readers, especially linguists, to evaluate better some of the forms, their sociolinguistic status and consequently their usage. Unfortunately, the problem remains that in most cases we do not know what the authors’ mother tongue was. The conclusion is that looking at these phenomena with the eyes of a linguist could have suggested the supply of additional pieces of information for each entry where possible.

The quantity of Latin loanwords in Greek is almost unimaginable, and D. has done a wonderful job, even though it is possible that a few indirect residues of Latin loanwords not yet identified could be included in a second, desirable edition of this work. All scholars, both Classicists and linguists, in particular those dealing with Latin and Greek, should be deeply grateful to D., who undertook and brought to its end such a huge and necessary enterprise. The critical remarks mentioned in this review underline how much has been done and how much still remains to be done within the field of Latin loanwords in ancient Greek.