language family
A language family is a group of languages that all descend from one common ancestor — they are, quite literally, relatives. Just as cousins share a great-grandparent, English, Hindi, Spanish, Russian, and Persian all trace back to a single language spoken thousands of years ago. We call that family Indo-European, and it is the largest by number of speakers — nearly half the world speaks one of its languages as a mother tongue. (Counted a different way, by sheer number of distinct languages, the Niger-Congo family of Africa is the biggest, with well over a thousand.)
How do we know languages are related if no one wrote the ancestor down? By comparing them carefully. When the words for basic things — "mother," "two," "night" — line up in regular, repeating patterns across many languages, that match is too systematic to be coincidence or borrowing. Those patterns are the family resemblance, and linguists use them to reconstruct the lost parent tongue.
One thing to clear up: languages can sound alike or share words without being family at all. English borrowed thousands of words from French, yet a true family bond shows up in the deep, ordinary vocabulary and the regular sound correspondences — not in fancy loanwords. And living near each other doesn't make languages relatives: Finnish and Swedish are neighbors but belong to entirely different families.
"Mother" across the Indo-European family: English mother, German Mutter, Latin māter, Sanskrit mātṛ, Russian mat' — one word, many descendants.
Regular correspondences like this — not chance resemblance — are the evidence for kinship.
The family-tree image comes from 19th-century scholars like August Schleicher, who literally drew languages as branching trees. The breakthrough was Sir William Jones's 1786 observation that Sanskrit, Greek, and Latin were too alike to be unrelated — the spark that launched comparative linguistics.