Katherine McDonald is a post-doctoral researcher in the ‘Greek in Italy’ project run by James Clackson at Cambridge. The book reviewed here presents her research on language contact between Greek and South Oscan, for practical purposes defined as the Oscan written in Greek script. South Oscan is also the topic of her colleague Nicholas Zair’s work, who focuses on orthography and phonology.1
McDonald’s first chapter is a general introduction. This is followed by a theoretical chapter on bilingualism and language contact in written texts.
It is only in the third chapter that a more practical discussion of alphabets, orthography and epigraphy begins. The South Oscan alphabet is based on the Ionic alphabet of the Hellenistic period. Greek at the time did not have a sound [f], common in Italic and Etruscan. Etruscan used the symbol <8> for [f], which made its way into Central Oscan; South Oscan used <Θ> instead, but, according to McDonald, <8> was used later due to influence from Central Oscan. The symbol or its mirror image, also used for /f/, was derived from <8>. Stuart-Smith2 has convincingly shown that word-internally /f/ had always had a voiced allophone in Oscan, so that the use of the letter in this position is not a reliable dating criterion; previous scholars either assumed that /f/ was voiceless in all positions, or that medial voicing was a late phenomenon, so that inscriptions with for /f/ could be assumed to be late. McDonald is usually very cautious about dating inscriptions based on orthography. Thus, she points out that the diphthong /ɛi/ continued to be written <ει> even after the spelling <ηι> was introduced around 300 BC. However, when it comes to spellings for /f/, I feel that she should have been more cautious. Given how difficult it is to date Oscan in the Greek alphabet, it is not inconceivable that <Θ> and were purely orthographical corruptions of Etruscan <8>, which would make inscriptions with <8> older.
It is interesting that <ψ> and <ξ> were used in names, while their equivalents <πσ> and <κσ> appeared elsewhere. The letter <ζ> was used for the reflex of *dj-3 or for the intervocalic, voiced allophone of /s/.4 This suggests that the reflex of *dj- was also pronounced [z]. McDonald argues that where the letters <φ, χ, θ> are used for native stops, it is done mostly in order to create a Greek appearance. However, for [ŋg] the sequence <νγ> is used as in the Oscan national alphabet or in Latin, not <γγ> as in genuinely Greek texts.
The fourth chapter is about dedicatory inscriptions. On p. 115-21, McDonald discusses a monolingual Oscan inscription followed by ΝΗΗΗΔΠ, to be interpreted as the abbreviation for ‘coin’ (South Italian Greek νόμος, Latin nummus) and a Greek numeral notation for 315. On p. 130-1 we encounter an inscription that reads λευκιος [δ]εκκιο[ς]. The name is clearly Oscan, corresponding to Latin Lūcius Decius, and the gemination of /k/ in this phonological environment is a standard Oscan feature. And yet the man chose to Hellenize his name, using a Greek-style diphthong -ευ- instead of the Oscan /ou/5 as well as Greek nominative endings.
McDonald’s fifth chapter deals with curse tablets. This is perhaps the most complex, but also the most interesting part of the book. In curse tablets, it is very difficult to assess which Greek elements were the result of ‘natural’ language contact and which ones were deliberately chosen in order to obfuscate the text and to increase its magical powers. As is well known, the use of curse tablets in Italy began in the Greek-speaking communities of Sicily and spread north into other communities through cultural contact. McDonald notes that the ways in which such tablets were written changed as the practice moved northwards.
It is a remarkable statistic that half the names attested in South Oscan are from curse tablets. Direct translations of Greek formulae were common, as is to be expected. More unusual, and difficult to parallel in Greek, is the practice whereby writers included their names in the nominative (p. 165); this was avoided elsewhere in case the tablets were found by others.
Chapter 6, on legal language, shows that Oscan legal idiom was not simply calqued on corresponding Latin texts. There are even some syntactic differences: while both languages employed the future imperative for third-person commands, only Latin also used it in prohibitions, whereas Oscan had the perfect subjunctive, as seen in the Tabula Bantina.
Chapter 7 deals with the remaining texts, from official inscriptions to graffiti. Perhaps the most interesting discussion involves coin legends. The Lucanians were called Λευκανοί in Greek sources, with a dipthong -ευ- that corresponds etymologically to the diphthong /ou/ found in the Oscan form of the name. The ethnonym is related to Latin lūx‘light’. However, on some coins we find λυκιανων ‘of the Lucanians’, with an unusual spelling that is inappropriate for the diphthong. The coins have a wolf’s head on them, so there is clearly a pun on the ethnonym and Greek λύκος ‘wolf’.
The last chapter comprises McDonald’s conclusions, and this also brings me to the end of my review. Greek-Oscan language contact and bilingualism is an exciting topic, but as the book title itself says, the South Oscan corpus is fragmentary. Such a limited amount of text as we have it is sufficient for studying some aspects of the language, for instance spelling and its relation to phonology, but it is simply not enough to draw any major conclusions about the nature and extent of the language contact between Greek and Oscan. McDonald has a number of interesting case studies, and for these the book is worth reading. That said, there is not enough material to warrant the somewhat inflated theoretical apparatus that she presents us with. I can still recommend the book, but I would have preferred a shorter version, without the first two chapters, which contribute very little to the overall picture. The book is clearly structured and easily accessible, owing to many tables and maps as well as good indexes. It is beautifully produced, except that the photographs of inscriptions are often of very low quality, which severely curtails their usefulness. Outright errors are virtually non-existent.6
Notes
1. Nicholas Zair (2016), Oscan in the Greek Alphabet (Cambridge: CUP).
2. Jane Stuart-Smith (2004), Phonetics and Philology: Sound Change in Italic (Oxford: OUP).
3. As in ζωϝηι from *djowei, compare Latin DIOVEI‘to Jupiter’.
4. As in syncopated ϝενζηι from *wenesei, compare Latin rhotacized Venerī‘to Venus’.
5. Greek λευκός ‘shining, white’ and Oscan loúkis / Latin Lūcius are etymologically related, see below.
6. The one grammatical error that struck me was the translation of putíans as ‘may he be able’ (p. 147); the verb is of course in the third person plural.