Did educated people in the late Middle Ages and Renaissance use Latin routinely (Medieval Latin and Neo-Latin), rather than a regional vernacular, to conduct real-life conversations about ordinary, everyday matters? Were they taught how to do this in the schools of the day with the help of specimen written dialogues (colloquia)? Did their teachers use a Renaissance equivalent of the ‘direct method’, and did they teach Latin in the way that modern foreign languages are taught today? Or was spoken Latin, with a simulacrum of practical relevance to everyday life, a way of ‘bringing the subject to life’, an enjoyable diversion from the standard pedagogical fare (the ‘grammar grind’)? These are the questions that this article addresses. I argue that Latin was not generally used for everyday conversations, and that students were not taught how to conduct them outside the classroom any more than they are today, though spoken Latin was used as a medium for teaching and learning Latin, as it is to some extent today. Since Latin was not the first language of any native speaker, and since it was learned as a language primarily for reading and writing, comparisons with the teaching of modern foreign languages are specious. I also argue that spoken Latin today, as a pedagogical tool, is best kept out of the classroom and used, if it must be used, as a hobby or a pastime. It has limited usefulness as a means of learning Latin to a meaningful level (a level at which the learner can engage with original Latin texts). And the kind of Latin that is spoken in the classroom, an attempt to render a spoken form of Classical Latin, however ‘correct’ it may be grammatically and phonologically (and the grammar and phonology even of Classical Latin changed over time), is most unlikely to have been spoken routinely in the same kind of informal situations by an educated (one who is adept in Classical Latin) native speaker of Latin. In fact, the more ‘correct’ it is, the less likely it is to resemble authentic everyday spoken Latin, even of the educated elite that learned Classical Latin. This is even more the case after Classical Latin came increasingly to be different from the contemporary Latin that anyone spoke, and had increasingly to be learned from grammar books as if it were a second language. What Quintilian says of written Latin may be said of educated spoken Latin too: aliud est Latine, aliud grammatice scribere.

I continue with some extracts (the italics are mine), to which I append my comments:

‘When, around AD 800, Medieval Latin came to be clearly differentiated by speakers of Romance languages from the Late Latin assemblage of various written and spoken forms, it ceased to be perceived as a language spoken for everyday purposes. Spoken Latin was hereafter used only by the most educated classes, even though it might still be used by them in casual conversation. It was still spoken by such men until the end of the Middle Ages and after, though how often and in what contexts would vary. There were probably a good many churchmen who could adequately follow a sermon or a speech delivered in Latin but who could not have carried on a conversation. In the universities of the later Middle Ages it was naturally used in lectures and debates, but its use as the language for chatter between students at mealtimes had to be artificially maintained by rules. The value of speaking and understanding spoken Latin was preserved, but as a learned tongue, most often used for formal purposes and never with the same colloquial fluency as the vernacular languages.’ (Richard Sharpe, Medieval Latin: An Introduction and Bibliographical Guide (1996), p. 315)

[The extract maintains that if spoken Latin was used in the Middle Ages for ‘casual conversation’ it was only by educated people, and that it was beyond the capability of most of them. There is no suggestion that spoken Latin was used routinely for the same purposes, and with the same competence, as the vernacular languages. My belief is that most people, even highly educated people, would not have conducted most conversations in Latin.]Footnote ¹

As for the scant use of spoken Neo-Latin generally, and reasons for this, see the chapter on Neo-Latin by David Butterfield in the Blackwell-Wiley A Companion to the Latin Language (2011). (See also what he has to say about the ‘Direct Method’ of teaching Latin.)

‘Latin has come to epitomise a language of hard work, analysis, and logic. The humanists, by contrast, did everything they could to make Latin a living language learned primarily by hearing and speaking, like any other mother tongue. In their emphasis on using the language in life situations, the humanists came remarkably close to the practices used to teach foreign languages in the twentieth century: to the extent possible, learning a foreign language should replicate the way children learn their mother tongue in the first place.’ (Jürgen Leonhardt, Latin: Story of a World Language (2013), p. 224

[Note ‘any other’, implying that Latin was a mother tongue, not just like one. Latin was no longer a ‘mother tongue’ in the age of the humanists, nor had it been for centuries. And when it was a mother tongue it was not the Latin of the humanists nor the Latin that their Neo-Latin was based on, i.e. written Classical Latin. It is safe to say that no child acquired this form of Latin from its mother, literally or metaphorically.]

‘Collections of dialogues, which, for students, served as models of conversational Latin, were among the more interesting innovations of Renaissance pedagogy. The best known of such works was Erasmus's Colloquia Familiaria …’ (Leonhardt, p. 219)Footnote ²

‘…one has to admire Erasmus's way with words. However, one question remains: for what actual conversational encounter was it preparing students?’ (Leonhardt, p. 219)

[The same question could be asked of all the dialogues, not just of the present one between the two female characters, Xanthippe and Eulalia. I don't think Erasmus envisaged students, especially schoolchildren, conversing outside the classroom like the characters in his dialogues]

‘In short, Latin was taught as if it were a completely normal language and students were preparing to meet real Romans (of course, only Classical Romans from Cicero's times …)’ (Leonhardt, p. 223)

[Note ‘as if’: it wasn't actually a ‘completely normal language’, meaning presumably a first language used mainly for speaking. The reason is precisely that the language and style of Neo-Latin (the language that was taught by the humanists) was (ultimately) from Cicero's times. And Ciceronian Latin, even the Latin of Cicero's letters, was hardly the language of actual everyday conversations of ‘real Romans’, even elite Romans, including Cicero himself. As I say later, Classical Latin was a form of the language that was developed and elaborated (as opposed to evolved, as spoken languages do) essentially for writing literature and certain types of subliterary documents. Except for certain formal occasions, the Latin that was spoken was not Classical Latin, nor any later derivative of it, including Neo-Latin.]

We must distinguish between (a) literary sketches in Latin of characters engaged in conversations about everyday matters; (b) students reading such sketches as a novel aid to learning Latin; (c) students engaging in conversational Latin as a (more enjoyable?) aid to learning Latin; (d) students simulating real-life conversations in Latin as an aid to conducting such conversations outside the schoolroom; (e) educated people in general (the conversations are in Latin, now a learned language) holding actual real-life conversations in Latin about everyday matters; (f) educated people holding conversations in Latin about weightier matters; (g) people in general, including educated people, holding conversations in a vernacular about everyday matters; (h) the use of spoken Latin as a medium of instruction by teachers (especially Jesuits) in the teaching of Latin (the Renaissance equivalent of the ‘direct method’).

It is easy to get one's wires crossed! There is no doubt that (a) existed, and we know of other writers of such sketches apart from Erasmus. We also know that (f) existed, both regionally and supraregionally. It is probable that (b) existed, and we know that pupils of Erasmus himself were the intended readership of the earliest dialogues. It is likely that (c) existed, and we know that (h) existed. I find it very difficult to accept that (d) and (e) existed. If I am right about (e), then (g) existed, for educated people as well as uneducated. The only scenarios that resemble the modern practices of teaching Latin through the medium of spoken Latin are (c) and (h), to which I return later. I suggest above that (c) may have been a more enjoyable way of learning Latin, but that is not the universal experience of present-day students, who also find conversational Latin just as hard, if not harder, than more traditional methods. After all, composing correct Latin orally and spontaneously that has a scintilla of intrinsic interest is more demanding than written prose composition. And the less spontaneous it is the more stilted and artificial it becomes — hardly ‘conversational’ Latin.

Before the end of the first millennium Latin had been spoken by everyone for whom it was a first or a second language. Most of these people were illiterate or had basic literacy only. A minority was literate, of whom an indeterminate proportion was highly literate and wrote the literature that we read today. We do not know how the Latin actually spoken by these different groups of Latin-speakers differed, but we know that the written Latin of the (barely) literate differed from that of the highly literate. We assume that the spoken Latin of the highly literate differed to some extent from that of the literate and the illiterate. To what extent we don't know, but we assume, arguing sociolinguistically by analogy with other language communities for which we have actual evidence, that there was constant interaction between the spoken Latin of all these groups. Things were to change by the end of the first millennium, for the vast majority of Latin speakers would no longer speak Latin at all, leaving only those who were able to keep Latin alive, or at least in use, by learning it as a second language. What sort of Latin did these people speak, and for what purposes? Bear in mind that Latin was no longer a first language, and that whatever kind of Latin they spoke must have been a form of the Latin they learned as a second language, with no input from, or interaction with, any other kind of Latin speakers.

When Latin ceased to be a first language with native speakers it continued to be learned as a second language, as Medieval Latin and later from about 1300 as Neo-Latin. Both were based on and derivative of written Classical Latin, Neo-Latin more so than Medieval Latin. Throughout this period Latin was also spoken as well as written. But the Latin that was spoken was no longer the first language of anyone, but a form of the written language that was learned as a second language. For centuries Latin had been the only language used (apart from Greek and a few minority languages that survived) in most parts of the empire and former empire. But by the end of the first millennium other languages had emerged in these areas, especially the Romance vernaculars, acquired as first languages, spoken at first and later both spoken and written, and eventually with literatures of their own alongside Latin literature. The people who learned Latin as a second language — a minority of the population, of course — now had an acquired first language as well, and were effectively bilingual, if not wholly so (not all would have been equally competent in both). One might suppose that, given a choice, the language of choice for ordinary, everyday conversation would be the first language vernacular, not a learned lingua franca learned principally for the purposes of reading and writing works of literature, scholarship, science, law and administration. As such a lingua franca, there would have been occasions when the spoken form was useful, even necessary, though few practitioners of this spoken lingua franca were as competent as Erasmus, or as much-travelled in lands with different vernaculars where a common language was an asset.Footnote ³ But this is a far cry from using such a language as a means of everyday spoken communication in one's own country, for which the regional vernacular would have been adequate for all but recherché conversations.

The educated people of the Middle Ages and Renaissance learned Latin for the same reason that we do, to read what had been written in Latin from about 200 BCE onwards, insofar as texts were available for them to read (most texts of the better known writers from classical antiquity, and many of the lesser known writers, were available after about 1500). But, unlike us, they also learned Latin in order to write it, especially to write works of literature (as well as many other genres of writing, formal and less formal). If we want to write a novel or a collection of poems, we write it in our given language, not in Latin; and if our given language is a world language, we can be assured of a large audience capable of reading it. The given language of educated people in the pre-modern era was not Latin but a regional vernacular. Apart from the fact that Latin was a ready-made language for writing literature, the only one available at a time when the vernaculars were not sufficiently developed to encompass all the genres of Latin, Latin, if not a world language, was a supraregional one, the only language that was understood by educated people whatever their vernacular might be. If you wanted to command a wide audience for your work, it was essential to write in Latin. And the texts that were studied were studied as much for their usefulness as models of composition as for their literary value. The focus of learning Latin was therefore on the ability to read and to write it, not to speak it, certainly not to speak it in situations where the vernacular would have been a more appropriate medium of communication. The need to converse with one's peers supraregionally, therefore in Latin, was not widespread if it is true that most people, even highly educated ones, did not tend to travel frequently or extensively outside their own language area. Erasmus was the/an exception in this respect.

So what was the spoken Latin of the Middle Ages and Renaissance used for? Was it really used routinely as an alternative to a vernacular for everyday conversations, a kind of secondary first language, as some people seem to think? And was spoken Latin used in schools as a means of teaching Latin, which had to be learned as a second language of course? (See the extracts at the beginning of this piece.) If so, was it used only to enable children to acquire basic literacy in Latin, perhaps in a more attractive and agreeable way than the conventional grammar grind? We must remember though that Latin, however learned and for whatever purpose, was a second language, even (eventually) for Italians.

Was the learned Medieval Latin and Neo-Latin (the only forms of Latin now available: the Latin of the ordinary people no longer existed) used for everyday conversation any more than Classical Latin had been? On the whole, we do not know on what sort of occasions and in what contexts the spoken form of Classical Latin was used in the Classical period; nor in what way and to what extent it differed from the written form with which we are familiar; nor how it differed from the spoken Latin of the mass of the people, most of whom did not have access to written Classical Latin. It is questionable though whether the Classical Latin that we are familiar with was used much in everyday conversations, even among the educated elite. ‘’Quintilian at least (12.10.40-43) remarks that people like him tend to use the everyday language of the ordinary people when speaking with friends and wives (and with slaves, note), and that this language is different in its nature from that of an eloquent man, i.e. a highly educated person like himself: quae sit cotidiano sermoni simillima, quo cum amicis, coniugibus, liberis, servis loquamur … nam mihi aliam quandam videtur habere naturam sermo vulgaris, aliam viri eloquentis oratio. (And Tacitus (Dialogus 32.3 — a work probably published within a few years of Quintilian's), has Messalla saying of cotidianus sermo that it contains foeda ac pudenda vitia.) In the late Middle Ages and Renaissance, educated people, i.e. people who had learned Latin, presumably carried out conversations with uneducated people in the vernacular — the everyday language of the ordinary people. I find it bizarre to suppose that they would have carried out the same kinds of conversation in (learned) Latin with other educated people. Obviously, conversations about weightier matters would have been carried out in the lingua franca of Latin, regionally where the vernacular was deficient, ultraregionally where the vernaculars were different as well as deficient.

Writings such as the Colloquia Familiaria of Erasmus are literary sketches written in Neo-Latin, not in some non-existent ‘colloquial’ Latin that might have been learned orally by a Renaissance schoolboy and used as a model for real-life conversations. I doubt whether anybody had real-life conversations like those of the characters in Erasmus's dialogues. The only way in which the sketches differed from other, more conventional writings in Neo-Latin was in novelty of subject matter and accompanying vocabulary, and of course the grace and playful humour of Erasmus. The fact that the dialogues show ordinary characters having conversations about this, that, and the other in polished Neo-Latin does not mean that they are snapshots of real conversations between real people. They are no more real in this respect than the plays of Aristophanes or Plautus or the sketches of Lucian. Reading or reciting the dialogues might have been one way (perhaps a more agreeable way) that helped students to learn Neo-Latin; it would not have helped them to hold real conversations, the vast majority of which, if I am right, were not held in Latin at all, anyway, but in the regional vernacular. The reason they were written in Latin and not a (surely more lifelike) vernacular that more closely mirrored real conversations is that, like Erasmus's other writings, they were aimed at a supraregional audience (an educated one, of course) for which only Latin was available. In fact, beneath a thin veneer of realism they are as artificial as can be, not least because they are written in Latin, by then a wholly artificial language, a ‘cultural artefact’ (Clackson).

How exactly were these dialogues of Erasmus used in a classroom? Were they read aloud by pairs of students? Were they memorised and acted out as playlets? They could not be classed as exercises in ‘conversational Latin’ if that entails spontaneous oral composition. Were they used rather as models for conversational Latin exercises? Or were they written simply as a language learning aid, to be read privately like other texts? Erasmus, like other Humanists, had reservations about traditional pedagogy and favoured innovative methods (spoken and conversational Latin) and materials (texts that depicted vignettes of contemporary life) that he thought would encourage the more active involvement of learners and enliven the learning of Latin. This was surely what motivated his Colloquia, But it had all to do with learning Latin more enjoyably and effectively, not with preparing students to carry on everyday conversations outside the classroom in Neo-Latin. Erasmus was surely aware that this did not happen in real life.

Renaissance ‘conversational Latin’ school exercises (as opposed to the dialogues of Erasmus) would no more in themselves have equipped learners to read authors such as Persius or Manilius than do their modern counterparts today; at least it has not been demonstrated that they did. Nor would they have helped them to write sophisticated Latin, as they needed to do (unlike today). The ‘immersive experience’ or the ‘communicative’ approach — call it what you will — does not get one very far towards reading or writing grammatically and stylistically complex Latin. And if such teaching and learning methods are not able to do this, of what practical use are they? We do not learn this difficult language, in strictly curtailed time allocations, in order to engage in artificial and banal pleasantries. They are essentially meretricious diversions that purport to bring these (allegedly) dead languages to life. They convey the misleading impression of presenting learners with re-creations of authentic spoken exchanges, based as they are on Classical Latin and on the assumption that everyday conversations were carried out in the kind of elite written Latin with which we are familiar. One is put in mind of slaves and other lower-class characters speaking flawless Classical Latin in the Cambridge Latin Course, a course that prides itself on putting its users in touch with the historical and cultural reality of the ancient Roman world.

The written representations of spoken Latin and Greek in the mouths of literary characters should not, in the absence of independent evidence, be taken as necessarily representative of the actual language used by real people in real-life situations. Literary language did not on the whole conform to ordinary spoken language, even of the elite. This is as true of a Renaissance humanist as of a Greek or Latin writer in antiquity.