Marcus Caelius Rufus (c. 88–48 bc) is the author of seventeen letters to Cicero preserved in Book 8 of Cicero's Epistulae ad familiares, totalling thirty pages in the Oxford Classical Text. We also possess nine letters of Cicero to Caelius (Fam. 2.8–16, eighteen pages OCT), three of which are reactions to preserved letters of Caelius. I will be concerned with those letters that were exchanged between the two when Cicero was proconsul in Cilicia (fifteen letters by Caelius – total number of words 5,275 – and eight by Cicero – 2,011 words), written between May 51 bc and November 50 bc. The size of the two corpora is large enough and the circumstances in which they were writing were sufficiently stable to see whether there are differences between the language of the two men.

Caelius had a good reputation as orator, as is testified by Cicero himself and by Quintilian:

INTRODUCTION

The Bellum Alexandrinum, Bellum Africum and Bellum Hispaniense, which have been transmitted to us as a sort of continuation of Caesar's Bellum civile, are not only important sources for our knowledge of the historical events of the 40s bc. They also have long been recognised as precious pieces of evidence for the stylistic diversity of Latin in the first century bc. At least since the end of the nineteenth century the three pseudo-Caesarian Bella have been interpreted as a reflection of colloquial/substandard Latin and stylistically classed as second-rate literature. This is particularly true of the Bellum Hispaniense. Already humanists such as Lipsius, J. J. Scaliger or G. J. Vossius qualified its style as ‘horrid’ (horridus) or ‘somewhat harsh’ (duriusculus); the early editor Goduinus thought that the author's mother tongue was not Latin; Clarke (1753: 457) and Oudendorp (1737: ii.940) speculated that the work was a soldier's diary, and Madvig, Norden, Klotz, Pascucci, Diouron and others have sketched the image of an author who tries to write in an elevated style but constantly fails and reveals his lack of education.

When looked at more closely, this traditional characterisation must seem rather implausible. First of all, there are several features that contradict the hypothesis of a hastily written soldier's diary. It is commonly agreed that the speeches in 17.1–3 and 42.4–7 and Gnaeus Pompeius' letter in Chapter 26 are written in polished Latin and show no signs of negligence or incompetence.

INTRODUCTION

Possessive pronouns such as meus ‘my’ or tuus ‘your’ indicate that there is a connection between two entities. This connection may be possessive, as in my book, or it may be of a different nature, as in my friend. Lexical items often belong to different stylistic registers, for instance the poetic ensis ‘sword’ and its prose equivalent gladius. Grammatical items, on the other hand, tend to be stylistically neutral. Nevertheless, possessive pronouns have found a place in several stylistic studies. There are four reasons for this:

(i) Possessive pronouns can be emphatic; in English we can use his own rather than simply his in this case. Latin has different means of emphasis. One of them is hyperbaton. Adams (1971: 1–16) has demonstrated that verbal hyperbaton (type magnam habeo gratiam ‘I am very grateful’) is still rare in early prose, but becomes more frequent in the classical period, especially when the register is elevated. In Plautus, possessive pronouns are frequently separated from their head nouns and one may wonder what their register is. Does Plautus use hyperbaton for emphasis and stylistic reasons, or is he simply forced to do so by the metre?

During the early Middle Ages, ecclesiastical legislation required of all monks that they converse in Latin (during those few hours of the day when conversation was permitted). There was accordingly a strong incentive to learn to speak Latin properly; and this knowledge, once acquired, brought an additional benefit, namely that it facilitated travel for the Latin-speaking monk, enabling him (for example) to seek food and lodging from foreign monasteries as he journeyed from non-Latin-speaking countries such as Ireland, Britain or Germany, through the former Latin-speaking provinces of Gaul and Italy, on the pilgrims' route to Rome. Spoken Latin was learned by means of conversational manuals known as colloquia (the name was apparently coined by the great German humanist Beatus Rhenanus (1485–1547), better known to classical scholarship as the editor of Tacitus and discoverer of Velleius Paterculus). These colloquia consist of imagined conversations in syntactically simple sentences, framed so as to inculcate the vocabulary necessary for conducting the business of daily life. They constitute an interesting, but minor – and accordingly neglected – genre of Latin literature.

THE HISTORY OF SCHOLASTIC COLLOQUIA

The history of scholastic colloquia has never been written, and this is not the place to attempt it; but a few general remarks may be helpful by way of introduction to the text which will form the focus of this study. Broadly speaking, there were three periods of creative activity when imagination and effort were expended in the composition of scholastic colloquia.

Cato's Origines marked the beginning of Latin historiography, as earlier authors, beginning with Fabius Pictor, had written in Greek. It was composed in the latter part of his long life (234–149 bc) and consisted of seven books, of which the first three dealt with the foundation of Rome, the regal period and the origins of the cities of Italy (hence the title), while the remaining four contained an account of Roman history from the First Punic War to 149. He included in the Origines (something unique in ancient historiography) at least two of his speeches, one, delivered in 167, arguing in the Senate against declaring war on Rhodes (95), the other, delivered shortly before his death, supporting a bill to set up a special court to try Ser. Sulpicius Galba for his treatment of the Lusitani (108–9); the remaining fragments of his speeches (there are 254 in Malcovati) are not considered here.

Cato wanted to write impressively, and to that end looked for appropriate vocabulary and stylistic devices wherever he could find them. He took words from poetry and he neologised (see Briscoe 2005: 60), and, in principle, there is no reason why he should not, on occasion, have made use of features derived from the spoken language. To identify such elements, however, is very difficult: the only substantial texts earlier than Cato are the plays of Plautus, which certainly contain much that belongs to the spoken language, but also elements which are high-register.

Petronius uses many linguistic devices to characterise his narrative and the persons he describes, employing resources from the fields of vocabulary, morphology, syntax, code-switching, rhetoric and pragmatics; the only major linguistic possibility that he leaves underexploited is the description of pronunciation. Many studies of Petronius' language deal with the idiosyncrasies of his expressions, but the subject is still not fully understood, in part because his work is a literary creation that deliberately violates the literary conventions of classical Latin (see Adams 2005b: 77–8; Herman 2003: 139). The difference between this prescriptive or normative good literary Latin and the Latin of Petronius is notable. So, how can we define the language of Petronius?

The study of Petronius' language begins with the understanding that it varies both by genre (e.g. between dialogue and narrative) and by social context (e.g. between the speech of one character and another). Much has already been written on different aspects of this variation; here my aim is to examine a few specific usages and see what light they can shed on Petronius' linguistic and literary technique.

The Satyricon involves many different genres (see e.g. Petersmann 1977: 26; Callebat 1998: 10–2, 25–6; Biville 2003: 50–2), and Petronius writes according to his conceptions of the respective genres (oratory, epic poetry, tragic drama, derisive poetry) and also tries to use different registers for lower-class dialogue, middle-class dialogue, first-person narrative and, perhaps, ‘foreigner speak’.

INTRODUCTION

Roman tragedy has not received the same amount of careful examination as Greek tragedy – not even the tragic œuvre of Seneca and those few other tragedies that happened to survive in full. Partly the fragmentary nature of the textual corpus is to blame, but also, and this is perhaps even more important, the fact that Roman tragedy never managed to leave behind its smell of being secondary, of being derivative, and of being inferior to the Greek models. Nevertheless, some aspects of Roman tragedy, even early Roman tragedy, are reasonably well understood.

Performances of literary and subliterary tragic plays were an integral part of Roman dramatic festivals and were very much appreciated by all strata of Roman society. Leading Roman intellectuals felt that Roman tragedians and tragedies were roughly equal to their Greek counterparts (or even better still). And, in fact, one of the most important reasons why Romans felt so strongly about their tragedies and tragedians was the art and nature of the tragic language.

From a linguist's point of view, the language of Roman tragedy (and, in more general terms, of Roman drama) is a fascinating, artistic and artificial construction – but it could nevertheless have included colloquial features; the principal aim of this paper is to determine whether it did, but we shall also examine the nature of tragic language in a broader sense.

The present volume is to honour a man for outstanding contributions to scholarship, and to thank a friend for support and inspiration. Jim Adams is one of the very best and most important students of the Latin language who have ever lived. I attempt to sketch some of his achievements below, but first let me say this: I know that I can speak for the editors and many others besides (not only the contributors to the present volume, but generations of linguists, historians and classicists – not only Latinists – from many different countries), when I say that we are grateful to Jim not only for being the pre-eminent scholar that he is, for opening up and showing the way on numerous new or neglected aspects of Latin, and for publishing his findings so quickly, with such clarity and in such abundance, but also for inspiring us, pointing us in the right direction, and helping us to be better scholars. In my experience – and again I know that I speak for many – Jim has for decades been generous and unfailing in his readiness to share his learning, experience and approach, in answering questions, and in reading and discussing, and commenting and advising on, plans in germ and work in manuscript. For a linguist with work in draft, there are few things so beneficial as having it read by Jim, because he sees straight through the problems and tells you what the solutions are, and where to look to find the evidence to prove it.

GREETING AND FAREWELL EXPRESSIONS AS INDIVIDUAL ACTS AND SOCIAL PERFORMANCE

Greetings and farewells are among the most conspicuous aspects of interpersonal interaction in many different cultures and thus are a constant subject of anthropological, ethnological and sociological interest. From a linguistic perspective such expressions belong to colloquial language in its broadest sense, as they are inextricably connected to conversation and dialogue. Greetings and farewells are founded on a system of verbal interaction between individuals that varies according to cultural conventions, context, and the status and relationship of the interlocutors. To a much greater extent than most linguistic features, they require an interlocutor – though the interlocutor may not be actually present. Greetings and farewells also tend to come in clusters: the first interlocutor to utter one expects a reply or other reaction adequate or commensurate to it. Consequently these expressions are individual acts that belong to ritual performance governed by social conventions, meaning that speakers have a relatively limited freedom of linguistic choice (cf. Letessier 2000).

Indeed in modern western societies a speaker greeting or taking leave of a given person in a given context has a rather restricted set of options, such as ‘hello’, ‘hi’, ‘good morning’, and ‘goodbye’. These formulae often cannot be literally translated between languages because their meaning comes not from their lexical significance but from conventions that are strictly language-specific (cf. Cardona 1976: 205).

INTRODUCTION

Bede died in 735, at the twin monastery of Monkwearmouth-Jarrow near the modern Newcastle-upon-Tyne, where he had been a monk since his childhood. He was an Anglo-Saxon, living in an Anglo-Saxon kingdom, and will have spoken the local dialect of Old English. Shortly before his death, he completed his best-known work, the Historia ecclesiastica gentis Anglorum, by far the most important source for the early history of Christian Britain. Like the rest of his extensive oeuvre, it was in Latin. Bede's mastery of the language owed everything to the use he made of the astonishing library built up by the founder of his monastery, Benedict Biscop, during repeated visits to the Continent. What most influenced him were not classical texts, with the exception of Virgil, but the Latin Fathers and, it need hardly be said, the Bible.

Bede's narrative manner in the Ecclesiastical History is consistently grave and measured. But he follows his models, hagiographic as well as historiographic, in allowing direct speech to play a significant part. I propose in this contribution to make some observations on the style of passages where two of his characters are represented as conversing, usually in private but occasionally in public.

Such a topic seemed appropriate in a volume devoted to colloquial Latin. But in an author like Bede, the question of colloquialism is more than usually fraught. What could Bede know of conversational Latin? He had not (perhaps) read Terence, let alone Plautus or the letters of Cicero or Petronius.

INTRODUCTION

‘Hyperbaton’ is the name given, originally by rhetoricians, to the phenomenon in both Latin and Greek word order whereby words that are or seem to be syntactically connected (e.g. a noun and an adjective which agrees with it) occur some distance apart, separated by other words that are in grammatical terms less closely connected. A convenient example is given at Rhetorica ad Herennium 4.44: instabilis in istum plurimum fortuna valuit: omnes invidiose eripuit bene vivendi casus facultates, ‘Unstable Fortune has exercised her greatest power on this creature. All the means of living well Chance has jealously taken from him’ (trans. H. Caplan). Rhet. Her. defines the feature as quae verborum perturbat ordinem ‘[the figure] which disturbs the order of words’; its name, which the Romans translated by transgressio or traiectio, means ‘stepping over’, implying that it is a dislocation of some more usual order in which one would not need to step over anything. Standard grammars usually present it as a departure from ‘normal’ or ‘natural’ order, although (as it is unnecessary to point out in the distinguished company of the contributors to this volume and especially of its honorand) any such definition begs a large question about what is to count as normal or natural. Furthermore, the listing of hyperbaton among the ‘figures of speech’ (which has been conventional from ancient times to Kennedy's Latin Primer and beyond) tends to carry at least for a modern reader the implication that it is an artificial rhetorical feature, belonging perhaps to a formalised literary register at some distance from ordinary speech.

The engagement of the Satires of Horace with everyday Latin has generally been a given. After all, the poet himself programmatically referred to them as sermo merus, straight conversation (Sat. 1.4.48), not to be compared to the high-register language of an Ennius, for example (1.4.60–1). But it does not take much to see that this position is disingenuous, as studiedly disingenuous as Catullus' praise of Cornelius Nepos and diminution of his own work as ‘trifling’, or as selectively true as Cicero's distinction:

INTRODUCTION

Linguistic register has been characterised as the result of a speaker's choice in a given situation (see Müller 2001: 282–3). It is, nevertheless, impossible to draw a hard-and-fast dividing line between linguistic variables always used by certain speakers and those about which a speaker might choose (so Müller 2001: 283; cf. Coseriu 1980: 111–12). Moreover, speakers and writers do not choose between a finite number of discrete linguistic varieties but modify their language in a continuous way according to the situation. The question what register is, and how it is divided from sociolect, is discussed in Chapter 2. Here we address a simpler question: how did Roman authors conceive of Latin as including options between which speakers and writers had to choose? This question encompasses several more specific ones. What awareness do authors show of a gap between formal and less formal varieties of their language – for example written and oral or official and familiar? What, if any, legitimacy or correctness did they assign to the different varieties, especially the informal ones? What do we learn from the metalanguage of scholars describing linguistic varieties?

We thus leave aside in the first instance passages in which Romans divide Latin into different varieties used by different groups of speakers, since such distinctions reveal concepts of geographical or social variation rather than of register.

INTRODUCTION

Most of the poems in the Silvae have a personal addressee, usually a senator or an equestrian; sometimes the emperor. Depending upon the status of the recipient and the relative gravity of the topic, Statius may adopt a more – or less – jocular tone. Short parenthetical remarks that are characteristic of colloquial language usually lend an air of informality when they are employed in literary works. Hence we would expect to find them in a poem such as Statius' hendecasyllables to Plotius Grypus, complaining about the unsuitable present that Statius received from him for the Saturnalia; indeed, the climax of the long list of items that Statius would have preferred to Grypus' gift contains two colloquial parentheses within a sentence of four lines (4.9.42–5):

Parenthetic formulae of request (rogo, 42) are characteristic of colloquial speech, which favours parataxis over hypotaxis (H–S 472; Hofmann 1951: 129–30, 199).

Although (or because?) it is a truism that Lucretius admits colloquial features in his De rerum natura, there are surprisingly few studies explicitly devoted to the language and style of this author, and certainly no book-length treatment exists as we have for a number of Latin poets. Commentators tend to note what they regard as colloquial usages, but are usually not free to pause and justify their assessment. The most substantial single publication on colloquialism in Lucretius to date remains that of Diels (1922), which is concerned primarily with morphological and lexical colloquialism and promises another study on syntactic colloquialism (Diels 1922: 59). Diels did not live to complete it.

Diels's general view was that the instances of colloquialism in Lucretius reflect the language of the farmers around whom the poet supposedly grew up and lived. At the time, scholars disagreed about the extent of the colloquial element in Lucretius' language (see Heinze 1924: 47; Ernout 1923: 155), whereas modern scholars are also likely to question the general explanatory rationale which underlies it. Moreover, it is fair to say that a certain disconnect has become the norm between scholars who are interested in details of Lucretius' language, at least when considered within the context of Latin usage in general and poetic usage in particular, and those who work on Lucretius as a literary or philosophical text and on what one might call his ideological position, whether it is within the didactic tradition, within the intellectual landscape of the Hellenistic period, or narrowly within the Epicurean tradition.