This chapter will focus on patterns repetition in speech fragments from Cato the Elder to C. Gracchus, as well as the speeches quoted in the Rhetorica ad Herennium, with a view to understanding their composition and intended effects. Repetition provides a systematic framework for many of the traditional rhetorical figures, such as anaphora, alliteration, homoeoteleuton, antithesis and polyptoton (see D. Fehling, Die Wiederholungsfiguren und ihr Gebrauch bei den Griechen vor Gorgias, Berlin 1969; cf. J. Wills, Repetition in Latin Poetry, Oxford 1996). Using repetition as a lens allows analysis not only of longer extracts but also of very short fragments. These patterns will be used to test the thesis that Roman oratory continued to respond to the ancient Latin form of the carmen even while being influenced by Greek rhetorical ideas (cf. on this point E. Sciarrino, Cato the Censor and the Beginnings of Latin Prose, Columbus 2011). The transmission of the fragments under consideration is itself heavily influenced by the rhetorical and grammatical tradition, and my discussion will accordingly take account of the screening effects which this transmission has on the evidence.

What is ‘early Latin’? The main contention of the present volume is that this question does not have a single answer. Rather, ‘early Latin’ is one of those ubiquitous labels (like ‘old’ or ‘archaic’ Latin) which have been used by classical scholars to denote different linguistic entities, and above all to describe a variety of linguistic features, in an often confusing and potentially contentious way. ‘Early Latin’ is above all a linguistic construct, which evokes frameworks of periodisation (often diverging), and posits a distinction between a supposedly discrete and cohesive linguistic variety (‘classical Latin’) and another one, equally discrete and cohesive, belonging to an earlier time period (‘pre-classical Latin’, a notion which has often carried negative value judgments since antiquity). Far from aiming to replace one theoretical framework with another, the studies presented here contribute, through a fresh analysis of specific linguistic phenomena and stylistic trends, to challenge the myths of periodisation and standardisation, and to expose the limited usefulness of evolutionary models to explain language change.

Cicero is well known to provide information about early Roman drama through his frequent references to performances, biographical details and characteristics of playwrights, motifs in dramas, language and style. Most of these comments are integrated into a specific context and therefore reflect Cicero’s argumentative aim. Yet, at the same time, they reveal insights into the nature of Roman Republican drama and its assessment in Cicero’s time. This chapter explores Cicero’s comments on the language, style and rhythm of early Republican dramas as well as his taking up of linguistic features of these plays. By looking at a selection of representative passages, this contribution examines what Cicero says about the language and idioms of early Roman playwrights and analyses whether Cicero takes up any of the linguistic features highlighted or instead opts for alternative versions. Such a study enables a better understanding of the Romans’ own view of the language of their early dramas as well as of any differences and developments between the various playwrights and dramatic genres.

This chapter focuses on morpho-syntactical phenomena that are typically brushed off with labels such as ‘archaism’ or ‘colloquialism’ (e.g. siet; medio-passive infinitives ending in -ier; med; familiai; absence of subordinator ut; absence of subject accusatives). These labels provide neither an explanation nor an assessment that takes their functions and distributions into account. The analysis of distribution patterns shows whether a type of scansion, a morpheme, or a construction that stands out from the angle of an Indo-Europeanist or that of a classicist is in fact normal in early Latin, or marked in some way. For example, the discrepancy between the different conjugations shows that Plautus and Terence do not pick forms in -ier and then try to fit them into the line; rather, they write their lines and only pick forms in -ier if it cannot be helped. This has implications for assessing the stylistic value of forms in -ier in contemporary inscriptions or in Lucretius. This analysis, in turn, can help us to understand early texts in their own right, and aids us in reconstruction or understanding how a Roman of the classical period would have felt about a specific phenomenon.

The interest in early Latin developed mainly outside the field of normative grammar, particularly in authors who belonged to the tradition of scholarly or antiquarian writing. Varro’s encyclopaedic works testify to a unique effort to save uestigia of the cultural and historical past by means of linguistic operations. His approach soon lost its institutional character and was replaced by curiosity for rare minutiae, as in Pliny’s Dubius Sermo. Grammarians like Probus and Caper, whose orientation was philological rather than didactic, considered that the auctoritas of literary models made divergences from the norm or contemporary usage acceptable, and viewed uetustas as the area of experimental variation, both lexical and morphological, with respect to the usage of Republican writers. The inclusion of an immense corpus of literary quotations in comprehensive works (artes grammaticae) facilitated the adoption of an overall perspective that embraced the evolution of the linguistic system at all levels, and kept alive an awareness of the diachronic dimension of the language, which became increasingly profound in scholars like Priscian who read Terence in sixth-century Constantinople.