BMCR 2024.05.33

Latin loanwords in ancient Greek: a lexicon and analysis

, Latin loanwords in ancient Greek: a lexicon and analysis. Cambridge; New York: Cambridge University Press, 2023. Pp. 700. ISBN 9781108841009.

This text is one in a series of reviews of this work published by BMCR. The others are by Clifford Ando and Anthony Kaldellis; Éric Dieu; and Bruno Rochette. Links will be available at the bottom of the page as these are published.




Eleanor Dickey, whose achievements in Greek and Latin linguistics range from editing the Hermeneumata Dositheana to studying the two languages’ various forms of address, has in this hefty quarto volume, indispensable for anyone interested in or confronted by the subject, turned her attention to Latin words borrowed by Greek at all levels till (as is conventional) the end of the sixth Christian century. She begins by rehearsing the omissions of previous studies, in particular the by no means clear-cut distinction between loanwords and codeswitches; on p. 9 she sets out her criteria for recognizing a loanword, namely three good examples (textually certain, treated as Greek, earlier than ad 600, and independent of other instances), or two with other evidence such as survival into the seventh century or later (‘the Byzantine period’).

The meat of the book is the lexicon on pp. 20–502, in which each word is listed with meaning, origin, and references, but also labelled in the margin ‘Direct loan’ (in bold, with date), ‘with univerbation’, ‘with suffix’, ‘with influence’, ‘loan compound’, ‘deriv(ative) of loan’, ‘rare’ (meaning that there are good examples before 600 but too few to prove a loan even if, like ἀβολιτίων, ἀδνούμιον, and δροῦγγος—the second a malformation, the third a loanword even in Latin—the term is frequent in Byzantine usage), ‘foreign’ (said to mean that the word is so marked in all sources), ‘name’, ‘semantic extension’, ‘not Latin’ (even if borrowed by way of Latin, as Dickey allows that ἀββᾶ(ς), representing Aramaic אַבָּא , may have been), ‘not ancient’ (meaning that no example predates ad 600), or ‘non-existent’ (meaning that the reported attestations are insecure), the last three sometimes with a question-mark. Loans are dated to their first and last occurrence, which latter may be a particular century, ‘Byz.’, or ‘modern’.

The category of foreign words includes some not so marked at source, such as Hesychius’ gloss (p. 23) ἄβεις· ἔχεις, which is obviously not abis ἀπέρχει but habes in a substandard code that had transferred the verb to the fourth conjugation (see ThLL vi/3. 395. 51–62, esp. 52–3), and ἴσοξ, defined in lexica as a kind of fish (ποιός τις ἰχθύς), from a Celtic word for ‘salmon’ preserved e.g. in Welsh eog and well attested in Latin from the elder Pliny onwards as isox or more correctly esox. The lawyers’ ἀβστινατεύω is also designated ‘foreign’ (p. 25), but their ἀβστινατίων is called ‘rare’; there being no such Latin word as abstinatio, the distinction seems forced.

The lexicon is followed by discussions of particular questions, such as how Latin loanwords were accented in Greek. Dickey finds us knowing less than we thought: to be sure the Latin accent was retained in Greek if it fitted a normal pattern and overridden if it positively violated the rules, but otherwise in some words it remained, in others it was changed, and in many we do not know where it was placed—even if all speakers agreed, which seems improbable in the extreme; after all, in English loanwords from French such as chauffeur, some speakers put the stress on the final syllable, some retract it. Other loanwords were subjected to greater adaptation; the Aborigines most often became Ἀβοριγῖνες or Ἀβοριγῖνοι, though other forms are found.

For so long as the distinction between acute and circumflex was audible, the question arises whether Latin ŭ remained short as Greek ου; without positive evidence, Dickey is unwilling to believe a priori that it did (p. 538), and therefore prefers to accent σεκοῦνδος and δροῦγγος even while doubting the monstrous αὐρικαισωρίβους in a codeswitch. (This may seem excessively cautious until one has heard Scandinavians substituting [s] for [z] and Germans [ʃ] for [ʒ].) I take this opportunity of protesting against editors’ habit of printing ἄκτα despite the long vowel in ācta, Μάρκος for Mārcus, βράκαι for brācae, and λίβος for lībum as if it were the word for ‘secretion’; Dickey herself accents the dactylic diminutives βεργίον and βρᾱκίον proparoxytone against rule, though the former appears in Byzantine verse as βεργίν and the latter has survived into modern Greek as βρακί; by contrast, she gives the tribrach βίκιον ‘vetch’ the same accents as βῑκίον ‘bowl’. But, while we are on this topic, whence first came, in Plutarch’s codeswitched ueterem memoriam, the grotesque accentuation οὐετέρεμ, justifiable neither as Greek nor as Latin?

Another interesting question is why some Latin words were not integrated. Dickey discusses in detail Theophilus Antecessor’s choice from one place to the next of Greek or Latin script for Roman legal terms, relating it to established practice in earlier texts; she also contrasts his practice with Plutarch’s in the use of Latin inflections. As the chapter progresses, in fact, it becomes abundantly clear that there is no general answer to the question, only ‘because the author at that moment chose otherwise’. Theophilus in one passage writes in swift succession sententiam, sententia, and sententiai, with respectively Latin, ambiguous, and Greek inflection; but the same passage shows him writing, still in Latin script, opiniona, opinion, and (nominative) opiniones, the first two clearly Greek formations, the third ambiguous. To make things worse, in Latin words with Greek endings scribes sometimes change script within the word, without consistency; one recalls how German in the Alamodezeit used Roman script for its abundant Latin and French Fremdwörter but switched (with many blunders) to German script for their affixes.

Whereas one might have expected the greatest influx of loanwords to have followed the foundation of Latin-speaking Constantinople, Dickey shows that although the occurrences of such words greatly increased, the dating of their entry proves far more complex; no less so that of their departure, which she also considers, observing that presence in Modern Standard Greek is no proof of continuous survival.

Dickey finds some terms used only in particular parts of Egypt, such as κουιντανήσιος, ‘collector of the quintana tax’, confined to Berenice, and others well established in literary texts but absent from the papyri, such as βρᾶκαι; we should wish to know what occasion there was in Roman Egypt to speak of trousers, and how often the traditional term ἀναξυρίδες appears. That Acts contains more Latin terms than Luke is noted, and even supposed to cast some doubt on their shared authorship; however, three of these are place-names (two half-translated), two more are wind-names, χῶρος and εὐρακύλων, perhaps picked up at sea during the stormy voyage, and another the name of a terror group, σικάριοι (in Latin a common noun meaning ‘murderers’), also attested in Josephus and as סִיקָרין in texts relating to Jerusalem recorded in Jastrow’s dictionary of the Talmud; if these are to tell against common authorship, we need to be told how Luke would have expressed them had he had occasion to do so; would he have rendered Tres Tabernae Τρία Καπηλεῖα, as Zosimus was to do in a different age and place, or called the winds in question ἀργέστης and καικίας, even though like many other ancient wind-names these terms do not primarily denote direction?

Among the numerous questions Dickey raises for further research, some invite comparative studies with modern languages: when she asks, ‘Did Greek speakers ever pronounce Latin codeswitches or loanwords as in Latin, and if so in what circumstances? (p. 653), it is hard not to suppose that, exactly as with speakers of one modern language using words from another, it depended partly on their proficiency, partly on their audiences, but also on their desire to demonstrate cosmopolitanism or nationalism.

There are two appendices, one presenting the statistics represented by the graphs in the text, the other considering the definition of a word from the standpoint, not of linguistics (touching on such matters as univerbation and tmesis) or of socio-linguistics, as in a famous article by Anna Morpurgo Davies) but of lexicography, in particular whether or not in various lexica, including Dickey’s own, substantivized adjectives have separate entries, and whether or not proper names or their derivatives are included. The latter question leads into a discussion of capitalization, a purely modern and indeed language-specific problem even though in some European languages that do not capitalize (e.g.) months or derivatives of proper names a former practice of lower-casing such words in Greek and Latin has largely but not wholly given way to the English style.

On occasion Dickey seems excessively cautious: the question asked on p. 1, ‘was there in fact a distinctive variety of Greek spoken in Rome?’, invites the counter-question, ‘how could there not have been?’; (p. 568) the Latin reconstruction of a formulaic inscription does not need ‘perhaps’; (p. 593) nineteenth-century dialectal σαλγάμι ‘turnip’ can hardly not be a survival of σάλγαμον ‘pickling material’ in the diminutive (whereas modern κάρο ‘cart’ may for all we are told come not from κάρρον but from Italian carro).

Many readers may well find much of Dickey’s material unfamiliar; only those who work on such post-classical authors as Procopius will have encountered βάνδον, meaning a military standard, let alone recognize it as a ‘Roman’ term (which may indeed not mean ‘Latin’), and not even they may have encountered ὀνομάγγων (‘donkey-seller’) or συνθηκάριος (‘schemer’). On the other hand, there are occasions when the literary scholar will find the discussion wanting. Barbatus is not ‘only a name in antiquity’ (p. 69), being abundantly attested from Plautus onwards. If brute common sense did not suffice to establish that ου in reformed Boeotian spelling for unshifted υ could be short (whether in tinctured koine or authentic dialect), the poetry of Corinna should have removed the stumbling-block at p. 538 n. 1, e.g. in the anaclastic glyconics Ποτιδάων, τᾶν δὲ δουῖν and ἐν δόμως βάντας κρουφάδαν (PMG 654 iii 15, 20). Palatium is not merely the ‘imperial residence on the Palatine Hill’ (p. 662) but the hill itself, as abundantly attested in Cicero long before Augustus took it over.

Individual points invite comment:

  • P. 11, 14. Dickey records ‘beef’ as denoting only the meat and not the animal; that is so in modern standard English, but in earlier times one could well say ‘a herd of beeves’.
  • P. 18. Late Latin does indeed know ‘mosaic work’ as a sense for metallum. Bruneau 1988 rests on his observation at BCH 91 (1967), 443 n. 3.
  • P. 31. In denying (with a ‘probably’ that once again seems over-cautious) that τὸ αἴθριον was borrowed from atrium, Dickey dates its first appearance to the third century bc, but without further specification. In fact the earliest examples seem to come rather from second-century Alexandria (Callixenus fr. 1 ap. Athen. 5. 206 a on Ptolemy Philopator’s great ship, and LXX Ezekiel, where it sometimes represents מִפְתָּן, ‘threshold’, but in ch. 40 is used in passages the translator manifestly did not understand), in neither of which have we any reference to a Roman atrium.
  • P. 56. The first example of ἄσπρος, which would ultimately displace λευκός, is said to be Aelian, De historia animalium 1. 25, where according to the manuscripts hyaenas are born ἐν τοῖς καλουμένοις ἄσπροις χωρίοις. ‘In the so-called white places’ is puzzling; the sense ‘rough’ was restored in Jacobs’s λεπροῖς, but leaves καλουμένοις without point. That apart, Dickey, while accepting Psichari’s derivation from asper by way of the newly minted silver coin, objects to his posited asprum because ‘the neuter is usually asperum’; indeed in literary Latin it usually is, but the syncopated form, guaranteed by metre at Verg. Aen. 2. 379, Stat. Theb. 1. 622, and Pallad. De insit. 67, not only survives in French âpre and Italian aspro, but was likely enough to be favoured in the language that turned Catulus and Lentulus into Κάτλος and Λέντλος.
  • P. 87. Bonitarius, which Dickey marks with an asterisk and the Thesaurus Linguae Latinae does not recognize, is attested in Neo-Latin and well represented in modern languages.
  • P. 231 κουάσι = quasi. Dickey might have noted Theophilus’ quasicontracton at Institutes 3. 27, a univerbation with a great future ahead of it in the Latin Middle Ages and modernity but not found in the original, which always uses the phrase quasi ex contractu.
  • P. 255. Whence comes, unaccented and unexplained, Dickey’s mysterious λαργιλιατ?
  • P. 323–4, 575. Specific dated attestations of ὀγκία, οὐγγία, and the other forms would be welcome; and what is the authority for a long u in uncia?
  • P. 542. Although the error is irrelevant, Pseudo-Arcadius’ scansion of the first syllable in ἱδρώς as long warns us not to take ancient grammarians as gospel.
  • P. 548. ‘Λευκολλιανός has a Latin base, but the word as a whole may not have been borrowed from Latin since the Latin for those gardens was horti Luculliani.’ Since the Greek for Lucullus was Λεύκολλος just as the Greek for Lucius was Λεύκιος, it is easy to suppose that the adjective was adapted in borrowing; but the suffix ‑ιανός not having yet become current, the expected Greek derivative would have been Λευκόλλειος.
  • P. 552 n. 46. Having noted that loanwords retained Latin ns as νσ (for ‑ήσιος was not a derivative ‑ensis but a Greek suffix, as in Homeric Ἰθακήσιος), Dickey records two exceptions, of which the first is no exception at all, being μηνσώριος, in which the nasal was a folk-etymological distortion within Latin.
  • P. 562 n. 12. The Florentine Digest, with its abundant Greek, on which Mommsen’s edition was based, was written at most within a few decades of the promulgation.
  • P. 583. ‘[d]ictator is attested in Plautus (died 184 bc)’; more specifically in Pseudolus, performed in 191 bc, and in Trinummus, a few years afterwards, hence rather more than ‘only slightly’ before the appearance of δικτάτωρ in Polybius, who wrote two or three generations later (but still in the second century, not as on p. 553 the first). But, as at p. 577 Dickey seems to accept, we can hardly suppose that Q. Fabius Pictor had not used the word, even in FRHist 1 F15 = D.H. Ant. 7. 71. 2.
  • P. 589. ‘Many of these [sc. legal and medical terms] were no doubt unfamiliar to non-specialists, but within the relevant disciplines they may have been familiar as “adverb” is to modern Classicists.’ Not only to them.
  • P. 644. ‘[I]t is likely that Greek speakers processed adverbs in ως as inflectional forms of the corresponding adjectives, so that they could form them at will rather than learning them as individual forms.’ This, I take it, means that, regarding δικαίως as on a par with δικαίου, they could form σφαιροειδῶς from σφαιροειδής as unpremeditatedly as σφαιροειδοῦς, in which case for ‘likely’ read ‘screamingly obvious’; but we speakers of modern English, whose adjectives do not inflect, can form justly from just and spherically from spherical.
  • P. 659. ‘English “Rome” (from Latin Roma, perhaps via French)’; the name is already attested in Old English, undergoing the late-medieval narrowing of [o:] to [u:] and therefore pronounced [ru:m] well into the modern period before the Latin- and Romance-supported spelling pronunciation prevailed.

A review can hardly do justice to the wealth of linguistic and sociolinguistic content in Dickey’s book, from the manner of integration (or adaptation) of loanwords (pp. 503–32) to borrowed suffixes (pp. 543–56), from Latin words first attested in Greek (p. 583) to the semantic fields that yielded the richest crop of loans (pp. 625–42). This study will remain definitive for many years.