Proto-Indo-European: The Mother Tongue

Trace English back past Old English, past Latin and Greek, past even the oldest written records, and the trail converges on a single language that no one alive has ever heard — a tongue spoken on the grasslands of prehistory some 6,000 years ago. We call it Proto-Indo-European, and its children now stretch from Iceland to India. This is the remarkable story of the mother of half the world’s languages.

The Language Behind the Languages

Proto-Indo-European (PIE) is the reconstructed common ancestor of the Indo-European language family — the largest in the world, with around half of all humans speaking one of its descendants as a first language. It was never written down; it vanished thousands of years before the invention of writing. And yet linguists can describe its sounds, words, and grammar in detail. How is that possible?

The answer is one of the great intellectual achievements of the modern era: the comparative method. By systematically comparing words across the daughter languages and identifying regular patterns of correspondence, scholars can work backward to reconstruct the forms that must have produced them. Reconstructed words are marked with an asterisk to show they are inferred, not attested — for example, *péh₂ter, the ancestor of father.

The Discovery

The breakthrough came in 1786, when Sir William Jones, a British judge in India, observed that Sanskrit bore a resemblance to Greek and Latin “stronger than could possibly have been produced by accident.” He proposed that all three had “sprung from some common source, which, perhaps, no longer exists.” He was right. That insight launched the science of historical linguistics and, eventually, the reconstruction of PIE itself.

The Family Tree

From its prehistoric homeland — most scholars favor the Pontic-Caspian steppe north of the Black Sea — PIE spread and split into the great branches we know today.

BranchMajor languages
GermanicEnglish, German, Dutch, Swedish, Norwegian
Italic (Romance)Latin → Italian, French, Spanish, Portuguese, Romanian
HellenicGreek
CelticIrish, Welsh, Scottish Gaelic, Breton
Balto-SlavicRussian, Polish, Czech, Lithuanian
Indo-IranianSanskrit, Hindi, Persian, Bengali, Punjabi
OthersArmenian, Albanian, and extinct branches like Tocharian and Anatolian (Hittite)

Cognates: The Family Resemblance

The proof of the family lies in cognates — words in different languages descended from the same ancestor. Line them up and the kinship is unmistakable, especially among the oldest, most basic words: family terms and numbers, which languages rarely borrow.

EnglishLatinGreekSanskrit
fatherpaterpatērpitar
mothermatermētērmatar
threetrestreistrayas
isestestiasti
nightnox / noctisnyxnakt-
newnovusneosnava

Grimm’s Law: Sound Change With Rules

If the words are related, why don’t they sound identical? Because sound change is regular — it follows laws. The most famous, Grimm’s Law (described by Jacob Grimm, of fairy-tale fame), explains how certain PIE consonants shifted in the Germanic branch. PIE p became Germanic f, and PIE t became th. That single rule explains why Latin kept the p in pater and pisces while English shows f in father and fish, and why Latin tres corresponds to English three. The differences aren’t random — they’re fingerprints of ancient, lawful change.

One Root, Many Descendants

The most striking thing about PIE roots is how a single ancient form can fan out into a crowd of modern English words that no longer look related at all — some inherited directly through Germanic, others borrowed back in through Latin or Greek. The root *bher- (“to carry”) gave us native bear (to carry), Latin-derived transfer and ferry, and Greek-derived metaphor — all the same 6,000-year-old idea.

PIE rootMeaningEnglish descendants
*pēd-footfoot, pedal, pedestrian, podium, octopus
*dent-toothtooth, dental, dentist, mastodon
*ğneh₃-to give birthkin, kind, gene, genus, nature, pregnant
*reĝ-to rule, straightenright, rich, regal, regular, rajah
*kerd-heartheart, cardiac, cordial, courage
*me-to measuremoon, month, meter, dimension

Look closely at the heart row: the native English heart, the Greek-derived cardiac, and the Latin-derived cordial are all cousins — the very same root passed down three different branches of the family and then reunited inside English. Grimm’s Law explains the consonants: PIE k stayed k/c in Latin and Greek but shifted to h in Germanic, which is exactly why cord and heart begin differently.

How Linguists Reconstruct a Word

Reconstruction is detective work with strict rules. Suppose we want the PIE word for “father.” We line up the attested forms — English father, Latin pater, Greek patēr, Sanskrit pitar, Old Irish athir — and look for systematic correspondences. The non-Germanic languages show p where English shows f; by Grimm’s Law we know f is the innovation and p the original. Comparing the vowels and endings the same way, we arrive at the reconstructed *péh₂ter. The asterisk is a mark of intellectual honesty: it says “no one wrote this down, but the evidence requires it.”

The method’s power was proved spectacularly in the twentieth century. Linguists had predicted that PIE contained certain mysterious “laryngeal” sounds based purely on internal patterns. Decades later, when the long-lost Hittite language was deciphered from clay tablets, it turned out to preserve exactly those sounds — a prediction confirmed by buried evidence, the linguistic equivalent of finding a missing fossil.

A Window Into a Lost World

Reconstructed vocabulary even hints at how the PIE speakers lived. They had words for horse, wheel, yoke, and wool, suggesting a pastoral, wagon-using people who had domesticated animals. They had no shared word for sea but did for snow and wolf, pointing to an inland, northerly homeland. Words are not just tools of communication — reconstructed carefully, they become a kind of archaeology, letting us glimpse a civilization that left no other trace.

Frequently Asked Questions

If PIE was never written, how do we know it existed?

Through the comparative method: regular, systematic correspondences across dozens of related languages can only be explained by descent from a common ancestor. The reconstructions are hypotheses, but extremely well-supported ones.

Yes — distantly. Both are Indo-European languages, which is why basic words and numbers show ancient cognate links despite thousands of years and thousands of miles of separation.

Keep Exploring