Those of us who stood before dusty blackboards intoning "amo amas amat" often wondered why we were doing it. What was the point? In my class a favourite declension was bellum - "war" - which we chanted in the nominative, vocative and accusative plural as "bla bla bla". Our teachers assured us that Latin would help us in studying other languages. So why didn't we just study those other languages and have done with it?

If only we'd had Nicholas Ostler's fascinating new book, we would have found innumerable reasons to be interested. Enlarging on a mere detail from his magisterial history of the world's languages, Empires of the Word, Ostler tells the story of Latin's life. With lightly worn but hugely impressive scholarship, he pursues Latin from its birth and subsequent fate at the hands of invaders, plagues, grammarians, poets and priests, through to its present place, not buried in the graveyard of history, but perhaps reclining in its retirement home.

Latin the language, Ostler argues, formed Europe, and therefore also inspired the Americas that sprang from it. Lasting three times longer than Roman domination, it was written unchanged by courtiers, clerics and merchants for two millennia, and is still echoed today in scientific terminology, law codes, grammar, wizarding spells, and - until 40 years ago - the standard liturgy of the Catholic Church. Like the air we breathe, Ostler says, it is almost too central to our culture to have been noticed.

So where did Latin originate? Thought to be an Indo-European language that arrived in Italy some time between the 6th and 2nd millennia BC, it settled on the west coast in Latium, hence its name. Words - particularly those connected with architecture, such as atrium ("forecourt") and fenestra ("window") - were borrowed from neighbouring (but not Indo-European) Etruscan. And from the 2nd century BC, Greek loan words enriched the grammars of medicine, music and food (oliva or "olive" and olem or "oil").

From the Greeks, the Romans also discovered how a language could enable a person to express him- or herself fluently. For example, the notion of a structured sentence, ending with a "period" or full stop, was Greek, the word periodos meaning "circuit", as in "lap". Aristotle defined a sentence as "an utterance with a beginning and an end in itself, and a length that can be easily taken in".

How many of us have ever considered the origin of the humble sentence? Before this periodic system was adopted, an utterance could (and often did) ramble incomprehensibly ad infinitum. Thus, Roman scholars adapted these Greek rules to the rough Latin vernacular, thereby improving its style and comprehensibility, and making it easier to teach.

Oddly, as Ostler points out, it was another millennium and a half before Europeans realised they could apply these rules of grammar to their own vernaculars, thereby hastening the decline of the very language that had spawned them.

The Romans quickly discovered that a language could also be an instrument of power and commerce. Not through any official language policy, but through the army, the courts, and the merchants travelling the empire's long straight roads, Latin proliferated.

By the time Rome had fallen, Latin had replaced Greek in western Europe as the language of the church, the Greek texts having been translated into what was then the language of the street. By the end of the 1st millennium, through missionaries and by means of books, Latin had become the language of clerici - meaning both clerics and clerks.

Following the breakdown of Roman infrastructure and centralisation, Latin mutated into different branches of the language family known as Romance. Only in Britain was it displaced altogether. Why? Plague? War? Isolation? There is no answer, but in a characteristically intriguing footnote, we learn that plagues in general cut a swath through British courts, cities and monasteries, thereby militating against languages spoken by the elite. In 14th-century England, that language was Latin (along with French), and so, after the devastation wreaked by the bubonic plague, a great many vacant jobs had to be filled by English-speakers from the countryside. The use of French and Latin at court, by the law and by the church never recovered.

Yet Latin itself survived, and through 80 generations of grammar lessons its stable rules have been maintained. It is the one constant in the cultural history of the west, beyond Rome, beyond even Christianity. Its 2,500 years of documented history make it unmatched as the language of Europe's memory; thus, within the linguistically fragmented European Union, perhaps it could once again be used as the glue that binds Europe together. Then Latin would be the only language that European schoolchildren need learn, and "amo amas amat" would live again. If anything could inspire EU functionaries to adopt such an idea, it is this wonderfully learned and entertaining book.

Helena Drysdale's "Mother Tongues: Travels Through Tribal Europe" is published by Picador