Hilaire Belloc claims, in The Path to Rome, that he spoke only Latin throughout his epic walk from the Moselle to Rome -- and that he was generally understood by those he met along the way. It was an affectation but it contains a kernel of truth. The point is that to Belloc the modern romance languages and Latin are so closely connected that mutual intelligibility might be possible in certain circumstances.
Sceptics will rightly argue that French, Italian, Spanish, Portuguese and Romanian each possess their own well-developed systems of syntax and phonology as well as their own rich literatures. And it’s a fact that no native speaker of any of those vernaculars can understand classical Latin without first submitting to long years of instruction.
So how would we answer the sceptics?
Think of language in terms of registers. This word register has emerged as a technical term in the field of linguistics for the various styles or levels of speech and writing within a given linguistic community.
For example, in English we recognize a youth register, the idiom of pop culture and advertising and other forms of media aimed at young people. Next there is the level of educated discourse, the kind of speech that one would expect to hear in ordinary conversation between more cultivated people. Then there are technical registers: the language of sailors, or motor mechanics, or cricketers. This leads on to literary registers - the language of books – that may themselves be subdivided chronologically: Johnsonian, Shakespearean, Chaucerian English for example. Are all these registers, then, aspects of the one English language?
If that is true, why do we describe Italian, Spanish and Latin as different languages, while claiming that we ourselves share a common linguistic citizenship with Shakespeare and Chaucer? We agree that 'English' is the native tongue that we hold in common with Johnson and Chaucer and even King Alfred, but our approach to the Latin/Italian divide, for example, is inconsistent with that. The accepted narrative has it that Dante rejected Latin, chose to write his Divine Comedy in the vernacular and thereby bestowed independent respectability on modern Italian. Spanish and French, according to the same received wisdom, sprang fully grown from their respective native soils, distantly related to Latin perhaps, but quite distinct.
But Dante probably saw it differently: he thought of himself as using one form of Latin (which we would now call Italian) to write the Divine Comedy, and another (he called it gramatica – we would just call it Latin) for his rather more scholastic works, such as the de vulgari eloquentia. He thus distinguished between the literary and vernacular registers of the same language. It is we who have seized on the distinctions he made, prized them apart and set them in concrete. Nowadays a chronological chasm yawns between what Dante saw as two registers of the same Mother Tongue, a consequence of the disdain most of us have for the “Middle Ages”: if a piece of writing is old and dead we call it Latin, if it’s fresh and new and vivid we call it Italian!
So what is the real relationship between Latin and the modern romance languages? Are they little more than dialects of the same tongue? If this is true, how is it that the Italian schoolchild has almost as much trouble learning to read Virgil as the young anglophone? And if this is not true, how did the independent vernaculars such as Italian and French spring into existence, and what was their relationship with Latin during those long centuries when both Latin and the vernaculars existed side by side?
To answer this question we need to remind ourselves that language is really about sound and utterance, and that writing is merely notation. In other words, accurate judgements about language cannot be based on spelling alone, for appearances can be deceptive.
In every alphabetic language such as English certain words are spelled one way and pronounced quite differently. Think of ordinary English words as night or knob or though, where notation and pronunciation have long since parted company. This is not just a question of whimsy. Words are rather like archaeological artefacts, for they point us to circumstances and conditions that have long since disappeared. There really was a time when the gh in night and though were not silent, but had a strong guttural sound rather like the ch in Scottish loch or German Bach. It was once easy to distinguish between night and knight, because the k of the latter was sounded, as it still is in German words like Knabe or Knie.
 Such a situation might also apply in other languages. If we encounter the word bonum in one text and buono in another, is it completely safe to say that the first text is written in ‘Latin' and the latter in 'Italian'? It may not be that simple. It may well be that both words were pronounced identically, and that only the notation is different. If this seems like a stretch, consider that even in classical Latin, final m was definitely not vocalized before a following vowel, and probably never pronounced as we pronounce it (it was almost certainly nasalized). Moreover there is evidence that final consonants such as the –t of the third person singular verb and the –s of the second declension masculine noun might be dropped in ordinary speech. There is also evidence that the letters c and g, always ‘hard’ in classical Latin, had at a quite early date assumed the palate-alveolar pronunciation (i.e. a ‘soft’ sound) before the front vowels i and e that we find in modern Italian. We could summarize this by saying that spelling is often centuries out of date, and is therefore an extremely unreliable guide to the reality of verbal communication.
If this is true, the orator or the preacher speaking aloud in Latin in, say, the middle of the first millennium, may have been producing sounds that were startlingly at variance with the spellings in his written notes. This may go some way to explain how it is that congregations could understand St Augustine’s sermons, though written down and transmitted to us in ‘classical Latin’.
If we could go back to the last days of the Roman Empire I think we would be surprised to find how Italian Latin sounded even then. The tendency towards the convergence of the oe and ae diphthongs with long e would have been well advanced, as would the characteristic modern Italian treatment of c and g. Syntactically, too, I think we would have detected a preference, among ordinary people, for a possessive de followed by what might have been distinguishable as an ablative, in place of the more formal genitive. Of course the ablative would have sounded like an accusative anyway, because the final m of the accusative would have been silent. It is not hard to imagine a kind of bastard register of spoken Latin in which prepositions were employed to carry more of the burden of sense, and in which final noun/adjective endings were sloppily fudged. Speak this patois according to the Italian rules of pronunciation and what have you got? Does it matter whether we call it Vulgar Latin or Proto-Italian? Would its speakers have understood Cicero and Virgil? Up to a point, if they were spoken clearly, with all the supportive clues of tone and gesture.
Latin continued to be spoken by ordinary people in western Europe long after Rome “fell”, but the available registers of the language no doubt multiplied as the education system weakened and the several regions of the old Empire lost contact with the centre and with each other. Only towards the end of the first millennium, little more than a thousand years ago, do we first start to hear complaints that ordinary people could no longer readily understand their clergy when they preached to them! At about the same time new forms of notation appeared, not universally, but according to need. The Strasbourg Oaths are usually regarded as the first surviving document in French, but it is probably more accurate to describe them as the first attempt to express a particular military/legal register of Latin in phonetic notation, and to write down the universal speech as it sounded, rather than in accordance with traditional and conservative convention.
To anyone who loves language in general, and the Latin/Romance tradition in particular, this is exciting stuff for it brings with it that delightful frisson that comes from having one's assumptions overturned. To those who find language a tedious but necessary tool it will perhaps offer less delight but much valuable instruction in the nature of linguistic fluidity: to assume that French, for example, sprang into existence with the Strasbourg Oaths, to take no interest in anything that preceded it, nor any cognizance of the Latin that was written after it, is to impoverish and even cripple one's capacity to judge and appreciate literature.
As an undergraduate nearly 50 years ago I recall the professor of English at my university declaring that a knowledge of Latin was essential to the full understanding of any English literature written before about 1900. This startlingly arrogant claim raised no eyebrows then but it would today: fewer people than ever (as a proportion of the population) trouble to become proficient in a language other than English and there is a general assumption, at least in the anglophone world, that everything worth reading is available in translation.
Leaving aside the question of whether any translation can be regarded as a primary source, that assumption is itself far from well-founded. We are in a sense prisoners of our own culture. The very size of our prison may deceive us into thinking that it has no walls.
It is not easy to bridge the wide semantic and syntactic gulf between Latin and modern English – I cannot pretend otherwise – but modern pedagogic methods and aids probably make it easier to achieve than ever before. All that is lacking now is the will. David Daintree is Director of the Christopher Dawson Centre for Cultural Studies, in Hobart, Tasmania. For further reading Any reader interested in pursuing the historical and linguistic aspects of this question is referred to an excellent little collection of 16 essays, edited by Roger Wright, entitled Latin and the Romance Languages in the Early Middle Ages, Pennsylvania State University Press, 1996.