BMCR 2024.05.15

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 Éric Dieu, Leofranc Holford-Strevens, and Bruno Rochette. Links will be available at the bottom of the page as these are published.




Latin Loanwords is an extraordinary achievement. In it, Eleanor Dickey provides a lexicon of Latin loanwords into Greek as found in texts up to the year 600. She includes not only those that she has detected; she also aggregates earlier lexica. But even Loanwords’ lexicon is no mere compilation. Dickey assesses entries proposed in earlier scholarship with a critical eye; some are deemed “Not Latin” or, on occasion, “Not Latin?”. Each entry is provided with analytic tags; citations from literature, ostraka, papyri, and epigraphy; and references to relevant secondary literature. The lexicon is preceded by a chapter on method and followed by chapters of linguistic and historical analysis. The volume is accompanied by figures—many tracing chronological patterns of borrowings; the weighting of borrowings relative to the amount of surviving Greek; survival, accumulation and loss of loanwords; and so forth. (For the benefit of readers, the Contents and list of Figures are reproduced below.) The result is a piece of scholarship that will affect research across many fields. BMCR will therefore publish four reviews: this one, as well others by Éric Dieu, Leofranc Holford-Strevens, and Bruno Rochette.

Although the heart of Dickey’s achievement is contained in the lexicon and the remarkable and precise labor it distills, it draws on work in many fields of scholarship. This includes many lexica and glossaries: most famously, of course, Mason’s Greek Terms for Roman Institutions (1974) and Cervenka-Ehrenstrasser’s Lexica,[1] but much earlier work, too, including Cameron’s “Latin words in the Greek inscriptions of Asia Minor” (1931), reaching back to the 19th century, in scholarship by Lafoscade and Psichari. Some of these draw on specific texts, corpora, or topics: Magie’s splendid De Romanorum iuris publicis sacrique vocabulis sollemnibus in graecum sermonem conversis (1905); Dubuisson’s Le latin du Polybe (1985); and Avotins’ works on the language of the Code and Novels of Justinian, are examples along these lines. In these domains, work has been relatively continuous for the past century and a half.

Dickey also draws on analytic and interpretive work in fields that strike us as relatively neglected and everywhere her work suggests the value of renewed attention to subfields in both linguistic and cultural study. For example, the culture and practice of translation, both within specific institutional contexts but also beyond, deserve substantial and specific study. Weber’s catalog of Greek translations of Latin literary texts (K.F. Weber, De Latine scriptis quae Graeci veteres in linguam suam transtulerunt [4 volumes, 1835-1852)], for example, has been selectively updated for specific centuries and contexts (e.g., V. Reichmann’s Römische Literatur in griechischer Übersetzung (1943), which included a glossary, and E. Fisher, “Greek translations of Latin literature in the fourth century A.D.,” Yale Classical Studies 27 [1982]: 173-215), but the discipline desperately needs not only a catalog with testimonia but a debate on method. When do traces of the reception of Suetonius in the eastern Roman empire, for example, cease to be evidence of the reading of his text and become evidence of its translation in Greek? The excitement aroused by the production of epigraphic corpora led to numerous relevant collations of evidence, not least concerning translators: W. I. Snellman, De interpretibus Romanorum deque linguae Latinae cum aliis nationibus commercio, part 1, Enarratio, but esp. part 2, Testimonia veterum (Leipzig: Dieterich, 1914-19), is a case in point. Dickey herself draws attention, via reference to practice at the TLL, to how much we know about Latin via “Latin” words that are attested only in Greek (583). Someone should make a lexicon.


An early view (CA)

Dickey’s analytic and quantitative work cautions us against thoughtlessly regarding periods of well-attested interest in Latin in the Greek east as the obvious contexts for borrowings and the crafting of loanwords. Even the initial phase of Roman diplomatic contact as attested in documents did not coincide, she helpfully reminds us, with some initial moment when Greeks needed to talk to Romans or describe and understand Roman institutions (576). She therefore discounts the notion that the existence of multiple terms in Greek for a Roman concept—the mere fact of non-standardization in the age of Polybius, for example—is evidence of a coinage or that we find ourselves in an initial phase of contact or, more precisely, an initial moment of need to render a Roman concept in Greek. (Indeed, cautions about the world of spoken language—and therefore about the status of textual evidence—are pervasive, precise and cogent in the book.) But Dickey does agree with others that “the language of Roman decrees” was “influential in Greek” (579). The question of who produced the Greek translations of those decrees therefore matters, and the question does not seem as settled to us as it does to Dickey, who insists that Roman documents were “given Greek translations in Rome” before inscription in Greek cities (578). A few Roman documents survive in multiple Greek copies, and in any number of instances, the copies differ among each other in ways that suggest, if not urge, their origin in different translations (examples include the lex de provinciis praetoriis [early] and the Sacrae Litterae of Severus and Caracalla [late], on which see L. Robert, BCH 1978: 435-6).

The question of standardization raises a number of interesting points. Far the most frequent term for the Roman senate—and very regular in translations of Roman documents—is σύγκλητος. But it is by no means the only term: Polybius, Diodorus, Dionysius, Plutarch and Cassius Dio all occasionally use συνέδριον. What is striking is that βουλή does appear, including in the Greek translation of the Res Gestae of Augustus (1.2), but it was clearly felt inapposite. There was a non-correspondence between the Roman senate and Greek institutional operations, which is why Plutarch (Romulus 13.2) felt the need to explain that σενᾶτος was what the Romans called the σύστημα of the patricians. Now σύστημα is a word employed by Greeks to explain many forms of collective (including Roman priestly colleges), but not, so far as I can tell, the βουλή of a polis. In the same chapter (Romulus 13.4), Plutarch says that foreigners (οἱ μὲν ἔξωθεν) down to the present call senators ἄνδρας ἡγεμόνας—in fact, an extraordinary range of terms is attested—but the Romans (αὐτοὶ δὲ Ῥωμαῖοι) call them πατέρας συνγγεγραμμένους. These examples are cautions about the ability of loanwords to help us map cultural dissonance. In this case, even when Plutarch provides the “Roman” term and marks it as “their” usage, he translated it.[2] Plutarch was, of course, perfectly capable of transliterating Roman terms: the ῥαβδοῦχοι who preceded Romulus in public are called (by the Romans) λικτώρεις—but Plutarch’s flagging of the term as the Roman one causes Dickey to tag it as “foreign,” which is one of great affordances of her work (her practiced is explained on pp. 14, 22).

Of course, the study of loanwords has a hallowed place in the study of language contact precisely because they can be catalogued. Dickey herself closes her conclusion with a paragraph of questions. (The final chapter is in fact called, “Overall conclusions and remaining questions.”) Among these are, “How much was the native Greek vocabulary affected by semantic extensions due to Latin influence?” Here there is an irony, produced by the preference of many Greeks “for expressing Roman concepts in pre-existing Greek terms” (577), as follows. The extraordinary stress on Greek language produced by its confrontation with an alien thoughtworld—a confrontation that coincided with the need to adapt “pre-existing” Greek terminology to radically shifting material and institutional realities—is evidenced precisely in instances excluded from a lexicon of loanwords. The use of διαστολή in I Maccabees 8:7; the reference to Roman institutional operations via νομοθεσία in SEG 50, 1211 from Pergamum—to say nothing of the rapid transformations in the range of διοίκησις[3] and πολιτεία[4]—: the exploitation by historians of the riches that Dickey has made available will require placing her data in this larger world.


A late view (AK)

Latin Loanwords is an outstanding work of classical philology. It is addressed primarily to scholars of language contact in antiquity and, by extension, touches on issues of ancient cultural exchange and imperialism. Its significance and utility, however, extend far beyond that, into the Byzantine period and the makeup of modern Greek as well. It fills a critical gap in the lexicography of medieval Greek, which operated on several registers, the “highest” of which aimed to reproduce Attic Greek grammar and diction and to avoid using any of the vernacular terms that defined spoken Greek at the time. Spoken Greek, which later came to be called “Romaic” after the Roman identity of its speakers, sported an impressive number of Latin loanwords, whose origins are systematically and thoroughly explored here for the first time. As a result, Latin Loanwords will become an invaluable research tool for students of medieval Greek.

The following are some stray observations, additions, corrections, and ideas for future research, offered by someone who works primarily on those later Greek texts. The corrections should in no way detract from the admiration in which this book should rightly be held by all whose research areas it illuminates so brilliantly.

The distinction between loanwords and codeswitching is fundamental to the analysis in Latin Loanwords and is used, for example, to break down its tabulation of statistical data (2-3). Dickey defines these terms carefully and in a nuanced way (8-16) because they are not straightforward. I will illustrate this with a word that is missing from the lexicon, Ὀλλάρια (i.e., “market for cookware”), from the sixth-century historian Agathias.[5] It is clear from his account that the term was part of the standard vocabulary of Constantinopolitans (and so a loanword), but Agathias himself, writing in a purist idiom, flags it as a foreign word and so codeswitches (“as a Latin might call it, but a Greek-speaker would call it Χυτροπώλεια”). Thus, it is possible that loanwords do not need to meet all the criteria laid out by Dickey (9), provided we have a single ancient author who gives us all the information we need, in his own way.

Dickey also excludes texts in Coptic, Aramaic, or other languages, while acknowledging in a note (5 n. 4) that their Latin terms probably came via Greek, rather than directly via Latin, even though that it cannot be proven. However, in some cases it verges on certainty. A number of terms that are listed by Dickey as “not ancient”—they have only middle Byzantine attestation—are attested in Arabic administrative documents soon after the conquest, indicating that they were probably in use already during the ancient period.[6] Arabic would not have picked up Latinate neologisms from post-seventh-century Greek, and it almost certainly did not pick up terms from Latin directly, to use them, moreover, in the same way that they would be used later in Constantinople! We can thereby move some terms confidently out of the “not ancient” category, including σακελλάριος for a finance official (406); κοῦρσον for a raid or attack (240); and σιγίλλιον for an official sealed document (421).

Likewise, Dickey marks κάππα (“coat” or “cape”) as “not ancient?” (180), but it is attested without codeswitching in Maurikios’s Strategikon from ca. 600 AD.[7] I owe this reference to Andreas Kyropoulos, who posted a blog entry with other words that might be included in the lexicon, such as σκρόφα (female pig) and φαλκίδιον (a cutting instrument).[8] In the comment section there, a philologist who goes by Ιήτης noted some additional terms. We flag them here because blog posting in modern Greek will otherwise go unnoticed (I [AK] was already aware of two of them, but Ιήτης posted them first): κοξός (in Romanos Melodos, Hymn 60) from coxō/coxōnem > vulg. Lat. *coxus; and ὁλόβηρος (from Prokopios’ Secret History 21.8, about a type of silk product) from ὁλός + vērus. A curious case not discussed by Dickey is the hippodrome chant τοῦ βίγκας, i.e., tu vincas, which is attested reliably for the reign of Justinian in the Paschal Chronicle).[9] This is a Latin phrase that made its way into Greek and continued to be used in middle Byzantine times. It is best seen as a loanword rather than assuming that the crowd was self-consciously codeswitching into Latin, just for this expression.

There is one area where Dickey treats this fascinating intersection of Greek and Roman culture from a retrograde position. In her analysis of identity-terms at the end, she essentializes Greek identity from a western European standpoint. So regarding the Ῥωμαῖοι of the east, she says: “At that point ‘Roman’ came to mean specifically what we would call ‘Greek’, both in ethnic and linguistic terms, and that meaning persisted throughout the Byzantine period” (647). “We,” of course, means western Europeans. Before the nineteenth century, we misidentified the east Romans as “Greeks.” Since the later nineteenth century, “we” have misidentified them as “Byzantines.” The game is now up, and “we” are turning to calling them Romans. Similarly the name Roman, which, Dickey says, “eventually came to mean ‘Greek’” (653). What this should say is that “Greek” became the western exonym for east Romans. Unfortunately, this error has made it into the entry itself (400). If Romaios simply “meant” Greek, then centuries of western writers would not have used the latter term specifically to avoid using the former and we would never have had the terms “Byzantium” and “the Byzantines.”

I conclude with two broader interpretative questions that point to possible future research. Dickey recognizes that some Greek words changed their meaning to accommodate the senses of similar Latin words, though she includes only those that sounded the same; for example, θρίαμβος accommodated the meaning of triumphus (145) and ἐπὶ βαλανείων was a calque from the Latin a balineis (140). “Words of entirely native etymology can sometimes be influenced by foreign words so as to become in some sense lexical borrowings” (16). She mentions also the case of “semantic extensions (also called semantic loans), in which an existing word extends its meaning under the influence of a word in another language that already matches some of its meanings” (ibid.). These are not usually considered loanwords (17), so they are not included in the lexicon or analysis (nor would it have been feasible to do so). But it remains a fascinating and largely unexplored frontier of research because it points to a deeper influence of Latin over Greek, going beyond lexicography. I have argued, for example, that this happened to the word πολιτεία, whose semantic range mutated so that it essentially became a translation of Latin res publica, with senses lacking from the ancient Greek usage of the word.[10] We will likely find that the same is true about many other medieval Greek words: they were Greek in every respect but the origin of their meaning.

Finally, another frontier for further research is the precise contexts of transmission for specific words, especially how it happened that a native Greek for a common thing was displaced by a Latin one. Such explanation is not the goal of Latin Loanwords, of course, but it is an area of research that it opens up. We are not talking here about legal terms used in legal contexts and court titles used at the court, but rather of shifts in common terms, including words for “house” and for various household items such as “ovens” and “beds.” The presence of the Roman army may well have been behind the adoption of these loanwords. Dickey treats the army and trade as separate vectors of transmission (e.g., 640), but it is likely that the vocabulary used by the largest buyer of bulk goods (or via requisition, if not on the market) would impact the vocabulary of common goods. Latin words (e.g., λουκάνικον for sausage) for items purchased by the army in bulk could well have displaced Greek words in this way. An original argument of Latin Loanwords is that borrowing may have started earlier than is usually recognized (571), but many of the earliest borrowings are names that refer to Rome, Romans, and their distinctive institutions. During the empire, and especially the later empire, many Greek works for common items were replaced in bulk with Latin words, and this is a development that calls for socio-economic explanation. This is one among several avenues of exciting future research that is opened by this marvelous “lexicon and analysis.” It is a resource that will continue to pay dividends for decades to come.



List of figures. p xi

Acknowledgements. pp xii-xiv

  1. Introduction. pp 1-4
  2. The parameters of this study. pp 5-19
  3. Lexicon. pp 20-502
  4. How were Latin words integrated into Greek? Spelling and inflection. pp 503-532
  5. How were Latin loanwords accented in Greek? pp 533-542
  6. Which Latin suffixes were borrowed into Greek? pp 543-556
  7. Why were some Latin words not integrated? pp 557-570
  8. When were loanwords used? pp 571-594
  9. Where were loanwords used? pp 595-622
  10. Which words were borrowed? pp 623-650
  11. Overall conclusions and remaining questions. pp. 651-654

Abbreviations. pp 665-680

References. pp 681-703

Index locorum. pp 704-706

Index of Latin words. pp 707-732



  1. Compounds of loanwords for measurement
  2. New borrowing and derivatives versus accumulated loanwords
  3. Appearance of new borrowings compared to the amount of surviving Greek literature and documents
  4. Comparison of all the Latin words used in books 1-5 of Polybius with those in Greek translations of decrees of the Roman senate from the second century BC
  5. Estimated total occurrences (tokens) of Latin words (both loanwords and codeswitches) in Greek documentary papyri per century, according to Dickey (2003: 256)
  6. Probable codeswitches by century
  7. Survival, accumulation, and loss of loanwords
  8. The ultimate fate of ancient loanwords
  9. New loanwords by text type, absolute numbers per century
  10. New loanwords by text type, relative to amount of surviving evidence
  11. Direct borrowing classified according to the categories of Haspelmath and Tadmor (2009: 22-34)
  12. Overall semantic distribution of Latin loanwords (borrowings and derivatives)
  13. Direct borrowings by century and part of speech
  14. Derivatives of borrowings by century and part of speech



[1] I.M. Cervenka-Ehrenstrasser, Lexicon der lateinischen Lehnwörter in den grieschsprachigen dokumentarischen Texten Ägyptens: volume 1 (Α) = 1996; volume 2 (Β-Δ) = 2000.

[2] Among the most curious early translations of Roman terms is surely ξενοκρίτας for recuperatores in the lex de provinciis praetoriis (Roman Statutes no. 12, Cnidos copy, col. IV, l. 35).

[3] Note, e.g., SEG 17, 526, an honorific instruction in which a lawyer is thanked for his great benefactions to τὴν Περγαμηνὴν διοίκησιν.

[4] Note the rapidity with which the public law regulation of a province is described as its πολιτεία: RDGE 43, ll. 9-10, from 115/14 BCE; Strabo 14.1.38.

[5] Agathias, Histories, 2.20.5; ed. R. Keydell, Agathiae Myrinaei historiarum libri quinque (Berlin 1967 = Corpus fontium historiae byzantinae v. 2) 67.

[6] See, for examples, P. M. Sijpesteijn, Shaping a Muslim State: The World of a Mid-Eighth-Century Egyptian Official *(Oxford 2013) 69.

[7] Maurikios, Strategikon 1.3.2, 3.1.8, 12B.11.9; ed. G. T. Dennis, Mauricii Strategicon (Vienna 1981 = Corpus fontium historiae byzantinae v. 17) 86, 146, 432 (respectively).

[8] Το λεξικό λατινικών δανείων της Eleanor Dickey #3, Smerdaleos (blog), Aug. 10, 2023.

[9] Chronicon Paschale, ed. L. Dindorf (Bonn 1832), v. 1, 623.

[10] A. Kaldellis, The Byzantine Republic: People and Power in New Rome (Cambridge, MA 2015) ch. 1.