Conclusions

Chapter 10 draws the main conclusions together and offers some reflections on the old debate about the value of the Latin grammarians on the Latin accent. The book has largely set aside the question whether Latin grammarians tell us the truth about accents, but it has asked whether Latin grammarians’ statements, when taken on their own terms, are actually intended to pertain to the audible sound of Latin. This book has argued that it depends: some of their statements are so intended, and some are not. To finish with, the discussion is expanded to show how Latin grammarians treated prosody as they treated other aspects of the Latin language too, and how they proceeded as people usually do proceed when they learn techniques for talking about a language from people who use them to talk about a different language.

The Latin Circumflex

To some scholars, the Latin grammarians’ distinction between acute and circumflex accents for Latin has to be taken seriously. To others, it is a particularly clear example of absurdity inspired by Greek grammarians. Chapter 8 considers the evidence for this doctrine and shows that it has some intriguing characteristics brought out, in part, by grammarians’ discussions of accent mistakes and of the accentuation of Greek words used in Latin. The chapter goes on to argue that at an early stage of the Latin tradition on prosody, learned Romans made sense of the Greek distinction between acute and circumflex accents in such a way that they trained themselves to think they could hear a corresponding distinction in Latin.

que, ue, ne, ce

Latin grammarians held that the enclitics que, ue, ne, and ce caused the accent of the preceding word to fall on that word’s final syllable, regardless of the quantity of that syllable. Some modern scholars have regarded the grammarians as somewhat inconsistent on this point, or have noted that relevant discussions occur only in late works. But differences of opinion have focused especially on whether to consider the grammarians’ view worth serious attention in the first place. Chapter 6 considers Latin grammarians’ discussions first and foremost on their own terms: what do grammarians actually say about que, ue, ne, and ce, and what do they mean by it? The chapter also returns to the question whether the grammarians are telling us something serious about Latin, and if so for what period of the language.

‘For the Sake of a Distinction’?

Chapter 9 considers three prominent manifestations of the idea that a word can have an unusual accent for the sake of making a distinction: the accent ascribed to the final syllable of the preposition and/or adverb pone, the one ascribed to the final syllable of ergo ‘for the sake of’, and the one ascribed to the final syllable of circum in Vergil’s maria omnia circum (Aeneid I. 32). Although slightly different conclusions emerge in each case, all three case studies reveal that to some extent a scholar or teacher might indeed pronounce an accent ‘for the sake of a distinction’. In no case, however, is an accent simply invented for the sole purpose of creating a distinction.

Latin Vowel Length

Chapter 7 introduces the concept of vowel length in Latin grammatical thought, and ways in which the grammatical tradition responded to a changing linguistic reality. By the late antique period, the classical Latin contrast between long vowels and short vowels has been largely or entirely lost. Yet grammarians continue to consider vowels long or short, in the traditional way, and to appeal to these quantities for various purposes. For example, traditional vowel quantities are deployed to help predict which conjugation a verb belongs to. In this capacity, traditional vowel length has become an abstract entity or technical device of the descriptive system. But words for ‘long’ and ‘short’ still have transparent meanings, and can also be used for vowels that are literally long or short in late antique pronunciation. In late antique synchronic analyses there is a complex interplay between abstract ‘length’ and ‘length’ as literal duration.

Introduction

Chapter 1 introduces the relationship between Greek and Latin discussions of prosody, beginning with exchanges of ideas that took place between figures such as Cicero, Varro, and the elder Tyrannio in the Rome of the first century BC. With reference to Quintilian’s work from the 90s AD, Chapter 1 then shows that the Greek and Latin traditions remained in close contact for a time. The chapter goes on to discuss how Latin authors from Quintilian onwards recognize both similarities and differences between Greek and Latin accentuation; a comparison is offered between a late antique introduction to Latin accents (from Donatus’ Ars maior) and a Greek text ultimately related in content. Chapter 1 concludes with a statement of the aims and structure of the book.

Latin Proclitics I

Chapter 4 considers the late antique evidence for a widespread doctrine of the Latin grammatical tradition: that prepositions, relative pronoun forms, and certain other words have an acute accent on their final syllables, in apparent violation of the usual principles of Latin accentuation. The doctrine belongs to a way of talking about the accentual behaviour of proclitic words: words that are normally pronounced without an accent and form a prosodic unit with what follows. An abstract acute accent is assigned to the final syllable so that this can undergo a rule ‘lulling’ an acute on a final syllable into a grave (non-accent) in connected speech. The lulling rule is borrowed from descriptions of Greek, but we see various efforts to adjust its details so as to avoid results that are not intended for Latin. We also see other ways of saying that proclitic words are normally pronounced without an accent.

Latin Proclitics II

While Chapter 4 considered the late antique evidence for Latin grammarians’ thought on proclitic words, Chapter 5 considers the much scantier evidence dating from the first three centuries AD, working backwards in time from comments in a glossary preserved on a third-century AD papyrus (P.Sorb. inv. 2069) to comments in Aulus Gellius, Velius Longus, Quintilian, and a fragment of Remmius Palaemon. The chapter argues that these texts repeatedly present us with a picture compatible with what we find in late antique grammatical texts, and sheds new light on the notoriously difficult discussion in Quintilian.

Ancient Greek Theory of Prosody

Chapter 3 takes a look at Greek grammatical theory on prosody, and especially the area that was of most interest to the Latin grammarians: how a word’s accent can be affected by its context in connected speech. The Greek grammatical tradition provided a way of thinking about accents in terms of two levels of description: an abstract level at which each word had exactly one accent (its ‘natural’ or ‘own’ accent), and a concrete level at which each word had the correct accentuation for its context. The two levels were connected by a system of rules. The ‘natural’ accent was thus an abstract entity: it existed on the abstract level of the descriptive system, as a starting point for applying rules. The chapter introduces features of this descriptive system that will help to shed light on Latin discussions of the Latin accent.

Some History of Scholarship

Modern discussion of the Latin accent can be said to have begun in earnest with the publication of Weil and Benloew’s Théorie générale de l’accentuation latine in 1855. Responses to this work divided scholars strongly into two opposing camps—or rather, they strengthened and extended a pre-existing division into two camps that had originally concerned only the relationship (if any) between Latin metrical forms and the position of the Latin word accent. On closer inspection the two camps turn out to be rather loose alliances, but when the focus is on the Latin accent itself they rally around opposing answers to a central question: did Latin have a pitch accent or a stress accent? Chapter 2 sketches the beginnings of this battle and the main turns it has taken, and then argues that it is a mistake to see ‘pitch or stress accent’ as the crucial question, or even as a meaningful one. Even attempts to offer intermediate views mostly put a misconceived and unhelpful question at the centre of the argument. But if this question can be put to one side, some genuine questions come into view.