The findings of previous chapters are pulled together, and new directions for research suggested.

The distribution of loanwords between papyri, inscriptions, and literature is investigated: more appear in literature than in other sources, but papyri have the highest density of Latinisms. Local and regional loanwords existed, and these can be seen not only in papyri and inscriptions, but also in literature, which preserves traces of loanwords specific to the city of Rome. Special attention is paid to the New Testament (especially Acts of the Apostles), Atticising writers (especially Athenaeus and Lucian), the Edict of Diocletian, Roman historians, medical writers (especially Galen), Hesychius, the antiquarian John Lydus, and texts on Roman law (especially Theophilus Antecessor, the Scholia Sinaitica, and Modestinus).

This chapter consists of a lexicon of all Latin loanwords in Greek, as well as many of the codeswitches and words that have been claimed to be Latin borrowings but do not meet our criteria for loanwords (because they are very rare, unintegrated, marked as foreign, not necessarily ancient, not necessarily derived from Latin, semantic extensions, superseded readings, etc.). Evidence for (or against) considering the word a Latin loanword is provided, with references to further discussions (both elsewhere in this book and in other scholarship).

The scope of the investigation is clarified: ‘ancient Greek’ refers to extant texts in the Greek language written before AD 600, ‘Latin’ excludes other languages of ancient Italy, and ‘loanword’ includes derivatives of borrowings but excludes codeswitches and semantic extensions. The criteria for distinguishing loanwords from codeswitches are investigated, with a focus on frequency, integration, and not being marked by Greek speakers as foreign.

Previous claims of suffix borrowing are investigated, particularly for -arius, -aria, -arium, -ianus, -atum, -atus, -ensis, -tor, -ator, -atio, -ura, -inus, and -ella. Some of these were borrowed, others werenot, and in a few cases Greek speakers had not borrowed a suffix but believed that they had done so. Not all Latinate suffixes found in Byzantine and modern Greek go back to antiquity.

The main questions about Latin loanwords in Greek are raised, with the help of a passage from Athenaeus, and the evidence that will be used to answer them is explained.

What is a word? Are names (of people, places, gods, buildings, etc.) words? In antiquity spelling was not standardised, and gender, suffixation, and inflectional categories could also be variable: what kind of divergences indicate that a variant form found in an ancient source should be considered a distinct word from other related forms? Although such questions cannot be definitively answered, the approach taken in this book is clarified and justified by detailed comparison with other lexica.

Loanwords are divided into cultural borrowings and core borrowings, then categorised into semantic fields to allow typological comparisons. Fewer borrowings come from Roman political and military power (i.e. fall into semantic fields connected to law, government, and the army) than was previously thought. An analysis by parts of speech shows that nouns predominate but adjectives and verbs were also borrowed. Two loanwords related to identity, ‘Roman’ and ‘Christian’, are given more detailed consideration in the context of the imperial and late antique world.

Although it is sometimes argued that Latin loanwords were simply accented like Greek words, the reality was more complicated. We have evidence from the grammarian Herodian (via the epitome of pseudo-Arcadius) that Latin words sometimes had different accents from Greek words with the same terminations. There is also an accented papyrus showing a Latinate accent on a loanword. But Herodian also tells us that some Latin words did change their accents in Greek; there is no hard-and-fast rule for predicting which these were, and all available evidence must be examined on a case-by-case basis.

Some Greek texts contain words in Latin script and/or with Latin endings. Latin script occurs mainly in Roman law texts; these are investigated with particular attention to Theophilus Antecessor and the Scholia Sinaitica. Evidence for script mixture within individual words is considered. Words in Latin script and/or with Latin endings are more likely to be codeswitches than loanwords, but some loanwords appear with one or both these features. Multi-word phrases retain their original script and inflections much more often than individual words; these are mostly codeswitches, but some phrases may be loanwords.

Latin loanwords (and codeswitches) were normally written in the Greek alphabet and took Greek endings. Their spellings started out as approximate transcriptions of the Latin pronunciation (not transliterations of the Latin spelling), but over time the Greek spellings could either remain fixed as the Latin pronunciation changed or be updated to reflect such changes. Most loanwords joined a Greek declensional class that closely resembled their Latin declension, but some changed declension or gender when borrowed. Some borrowings (including all verbs) acquired Greek suffixes as part of the borrowing process. Some loanwords were created by univerbating Latin phrases, making Latin-Latin compounds, or making Greek-Latin compounds with the Latin element taken directly from Latin. Derivatives could also be formed from previously-borrowed loanwords using any of the usual Greek derivation and compoundingprocesses.

Apart from after /u/, /w/ and /kw/, where the raising of /ɔ/ was retarded until the first century BC, which is discussed below (Chapter 8), there are few cases of <o> for <u> arising from these contexts in the corpora. Even where we do find <o>, a confounding factor in identifying old-fashioned spelling of /u/ of these types is the lowering of /u/ to [o] which eventually led in most Romance varieties to the merger of /u/ and /ɔː/. According to Adams (2013: 63–70), this can be dated to between the third and fifth centuries AD, and did not take place at all in Africa. This requires him to identify a number of forms which show <o> for /u/ as containing old-fashioned spelling (or having other explanations) in the Claudius Tiberianus letters (Adams 1977: 9–11, 52–3; 2013: 63–4).

The Latin alphabet inherited from its Etruscan model a superfluity of signs to represent the phoneme /k/: <c>, <k> and <q>. It also inherited, to some extent, the convention in early Etruscan inscriptions whereby <k> was used in front of <a>, <q> before, and <c> before <e> and, although consistent usage of this pattern is found rarely even in the oldest Latin inscriptions (Hartmann 2005: 424–5; Wallace 2011: 11; Sarullo 2021). Over time, <c> was preferred for /k/ in all positions, while the digraph <qu> was used to represent the phoneme /kw/. Nonetheless, both <k> before <a> and <q> before (with the value /k/) lived on as optional spellings into the imperial period.