The Living Languages
On the invented death of Latin and Greek, the century that taught us to fear them, and what their revival tells us about the atrophy of public speech
I have never understood why Latin and Ancient Greek are often called “dead languages.” The phrase has always struck me as something worse than inaccurate. It felt like a subconscious command, an instruction: approach this language as one approaches a body on a morgue table, with instruments, at a clinical distance, expecting and hoping that you get nothing back.
This is teaching for an autopsy, not for connection.
Nothing about walking into a room to greet a cherished relative, however changed by the years, resembles walking into a morgue. The two postures could not be less alike, and it has always seemed to me that we owe Latin and Greek the status of a relative we share.
The reasoning behind this is not sentiment. A language dies when the community who spoke it disappears without issue: no descendants, no continuation, nothing carried forward. That is not what happened to Latin or to Greek, and it is worth being exact about why. Latin did not vanish. It changed, as living things change, into Italian, Spanish, French, Portuguese, Catalan, Romanian, and a wide constellation of regional tongues still spoken by hundreds of millions of people who inherited its case system in their mouths as children, long before they ever met it as a subject on a timetable.
Greek did not vanish either. Modern Greek is not a reconstruction or a scholarly resurrection; it is the same language, spoken by the same people in the same landscape, evolved continuously for at least three thousand years.
Peoples do not, as a rule, simply disappear, and neither does the language in which they conduct the business of living. This is transformation, the ordinary work of time on any living tongue, Latin and Greek included.
Call a language dead and you license the autopsy. Call it a language survived by numerous grown children, and you are obliged to go and visit the family.
Did The World Grow Stupid?
There is a second thing I have never been able to explain, and it is closely related to the first.
If Latin and Greek are now so widely dreaded, approached by pupils, with something closer to fear than curiosity, why did people spend so many centuries wanting to learn them?
Renaissance humanists competed for the prestige of a Ciceronian sentence. Monastic and cathedral schools for a millennium treated fluent Latin as the passport to the whole of learned Europe. Diplomats, scientists and clergy conducted the ordinary business of their lives in it. It cannot be that the grammar of Cicero or of Homer suddenly became more difficult sometime in the last hundred and fifty years, nor that whole populations suddenly became less capable of learning it.
Languages do not change their difficulty. Something else changed. And it was in the company of scholars who share both of these suspicions, about the false metaphor of death, and about this inexplicable reversal from longing to dread, that I found some of the clearest answers I have yet encountered.
On an October morning in 2025, forty people gathered in the Faculty of Classics at the University of Cambridge, and some two hundred and thirty more logged in from elsewhere in the world, to spend a day discussing how children should be taught Latin and Ancient Greek. The occasion was modest, even niche, by the standards of academic conferences. And yet buried in the papers that emerged from that day is an argument that answers both of my suspicions at once and carries implications well beyond the classroom, bearing, in fact, on the health of the polis itself.
Decoding Is Not Reading
The pedagogical debate in question concerns what linguists call Comprehensible Input: the proposition that a language is acquired not by memorising its rules but by encountering it, again and again, as meaningful communication pitched just above one’s current level of understanding.
For decades the opposite premise governed classical education. Latin and Greek were taught as codes to be cracked, declensions drilled, sentences dismantled into their grammatical parts, comprehension arrived at only after translation had been completed.
The result, as more than one contributor observed, was generations of pupils who could parse a subjunctive with precision and yet could not read a paragraph of Cicero for pleasure or sense. One researcher found that pupils beginning Tacitus, having passed their GCSE, recognised barely a tenth of the vocabulary in his opening sentence. They had learned an apparatus for decoding Latin. They had not learned to read it.
What is being diagnosed here, if one steps back from the specifics of case endings, is a particular and very modern impoverishment: the treatment of language as information to be extracted rather than as a medium of shared understanding. The comprehensible input movement is explicit that this is not merely inefficient but psychologically corrosive. English students, brave and willing enough to learn the Classics, can analyse Latin silently but feel a visceral, almost bodily reluctance to speak a single word of it aloud, because the language had only ever been theirs to solve.
The Century That Taught People to Fear
This is where the second suspicion I began with, finds its answer. The dread was manufactured, not inherited to the languages. Nothing intrinsic to Latin or Greek requires that a pupil meet them with anxiety rather than desire, and, as I have already noted, for most of their history nothing did.
The languages did not suddenly become harder somewhere in the last hundred and fifty years. What changed was the apparatus built to teach them.
The nineteenth century’s new national school systems, and above all the German gymnasium tradition of exhaustive grammatical dissection transmitted into Flanders, for instance, through Adhemar Geerebaert’s widely used grammars, reorganised both languages around the parsing of forms rather than the reading of sense. Pupils devoted “substantial weekly hours” to Latin under this regime and “frequently emerged unable to read fluently or compose effectively.”
The German reformer Georg Rosenthal, writing from inside that very tradition in 1924, gave the failure its sharpest epitaph: Gesehenes Latein ist kein Latein: seen Latin is no Latin. A century of schoolchildren were not defeated by Cicero or by Homer. They were defeated by a method built, to elevate the teacher and to stand between them and them.
This transformation did not occur in isolation. It belonged to a broader nineteenth-century confidence that education could be made scientific, standardised and measurable.
The same age that built the modern bureaucracy also rebuilt the classroom in its own image. Knowledge became something to be classified, examined and certified; language became an object to analyse. It produced administrators at the expense of citizens capable of judgment.
We inherited schools remarkably efficient at teaching pupils to identify the parts of a sentence, yet increasingly uncertain for the far more difficult task of entering a conversation about the common good.
The reduction of Greek and Latin to grammar was not an isolated pedagogical mistake. It was one expression of a much larger civilisational habit: replacing participation with procedure, understanding with technique, and wisdom with administration.
It is worth pausing on how familiar this impoverishment now feels outside the classroom. The instinct to treat language as an extractable payload (skim for the headline, mine the transcript for the quotable line, let the algorithm summarise the argument so that one need not sit with it) has migrated from the Classics classroom into the entire architecture of contemporary public discourse.
We have built a civic culture preocuppied with decoding for usefulness rather than reading and understanding. We decode speeches for scandals, skim interviews for headlines, debates for social media clips and books for sharable quotations. We no longer linger long enough inside the language for the meaning to land and possibly change us.
The current civic culture processes political speech for its retrievable content while remaining strangers to the deliberative act that speech is meant to enact. The parallel is not incidental. It is, I want to suggest, the same reduction wearing different clothes.
Before “Word” Meant Word
The Greeks did not have a separate term for language as opposed to reason, or reason as opposed to public speech. Logos did the work of all three. For Aristotle, the human being was not simply zoon politikon, the political animal, but was political precisely because it was zoon logon echon : the animal possessing speech, capable of declaring not merely pleasure and pain, as other animals can, but the just and the unjust, the advantageous and the harmful, in common with others.
The polis, in this account, is not first a set of institutions. It is a community constituted by the shared act of meaningful speech about how to live. Citizenship is inseparable from the capacity to use logos with others, not merely to decode it alone.
This is why the recovery of “living” Latin (Rouse’s Direct Method at the Perse School in 1902, the interwar dream of Latin as a neutral, unifying European lingua franca, the Jesuit school dialogues in which pupils actually greeted, argued with, and teased one another in the language rather than merely construing it) carries a weight beyond antiquarian curiosity.
Each of these movements, in its own century, was an attempt to restore language to the condition Aristotle assumed it should have: a medium held in common, not a corpus held at arm’s length. Cato’s motto, beloved of Latin revivalists, captures the same intuition in six words: rem tene, verba sequentur : grasp the matter, and the words will follow.
Meaning precedes and generates language; language does not precede and generate meaning. Reverse that order, as grammar-translation pedagogy did, as much of our public discourse now does, and one produces fluent-seeming output that has never actually been understood by anyone, least of all its author.
Small Republics of Speech
What is striking about the practitioners of this new old approach to Greek and Latin is how deliberately they have reconstructed the conditions of a small polis within the classroom in order to solve this problem.
At the Open University, students are given “incomprehensible” literary Latin they cannot yet translate and asked to make what sense of it they can through community, discussion, and parallel translation, before returning to it years later with earned competence, a pedagogy that treats interpretive struggle in company as more valuable than premature, solitary certainty.
At Haileybury, texts are read aloud in chorus, so that understanding is verified not by silent test but by the audible confidence of a room of pupils speaking together.
At the University of Oxford, Cressida Ryan’s inclusive-design argument makes explicit what the others imply: that a “high face-saving” environment, one that does not punish the vulnerable act of trying language aloud with the twenty thousand small humiliations that a certain kind of pupil has already absorbed by adolescence, is not a nicety but a precondition for anyone to risk speech at all.
None of these are technical fixes. They are re-descriptions of the classroom as a place where meaning must be built together, aloud, at the risk of error, rather than assembled privately and submitted for marking.
It is hard not to notice that this is precisely the capacity that contemporary political life has been shedding: the willingness to risk one’s half-formed view aloud, among others, in the expectation that understanding will be reached through the exchange rather than possessed before it.
We have not lost the capacity to transmit information. We have lost, in alarming measure, the practised comfort of using logos together in public, at the pace and vulnerability that genuine deliberation requires.
The Real Corpse
Latin was never the body on the table. What we have actually been performing an autopsy on, these past two centuries, is our own willingness to use language together, in public, at any risk to ourselves, and that patient, unlike Cicero’s tongue, may not have living descendants waiting in the wings.
The irony is that this recovery arrives precisely when machines have become extraordinarily good at the very skills schools spent two centuries cultivating. Artificial intelligence can now translate, parse, and summarise faster and more accurately than any pupil. The machine is not wrong to do this, but the pedagogical task has shifted in consequence: what remains for the human learner to cultivate is not retrieval but interpretive judgment, the capacity to sit with a text or an argument, and ask what it means, what follows from it, what it asks of the reader.
This line of thought applies with equal force to citizenship. A public that has outsourced the labour of understanding to a summarising machine, like the TV, while retaining none of the practised capacity to interpret, deliberate, and speak for itself, has reproduced in the whole of civic life the very failure that Comprehensible Input was invented to correct in the Latin classroom: fluent-seeming output, uninhabited by comprehension.
It is the same discussion as to whether a fourteen-year-old should meet the ablative absolute through a graded story about Ariadne or through an isolated grammar drill. The underlying wager, that a shared tongue is either lived and spoken into common understanding or merely decoded and left inert, is the same wager on which the health of the polis ultimately rests.
A civilisation that has forgotten -or suppressed- how to use its own language together, in the vulnerable, provisional, communal way, should not be surprised to find that its political speech, however voluminous, no longer persuades, no longer binds, and no longer means very much at all.
Rem tene, verba sequentur.
The recovery of Greek and Latin is not, in the end, about Greek or Latin.
It is about remembering what language is for. Communication.
Before there can be institutions there must be speech. Before there can be politics there must be logos.
A civilisation is held together not only by its laws or its markets, but by its shared capacity to search for meaning aloud.
That is why the Classics matter.
Not only because they teach us ancient words and concepts, but because they remind us how a free people learns to speak.