Learn etymology · Guide 1 of 3

How etymology works

Etymology is the history of words: where a word came from, what it first meant, and how it reached the form and meaning it has today. The word is itself a good example. It comes from Greek etymología, built on étymon, “the true sense of a word”, from étymos, “true”. The ancient Greeks thought a word’s origin revealed its real meaning. Modern etymologists don’t, as this guide explains, but the name stuck.

The evidence: what was written down, and when

An etymology is a historical claim, and like any historical claim it rests on evidence. For most words that evidence is writing: every surviving letter, chronicle, poem, law, sermon and newspaper in which the word appears. Large historical dictionaries, such as the Oxford English Dictionary, collect these quotations and arrange them by date, so that you can watch a word appear, change spelling and change meaning across the centuries.

The earliest quotation found for a word or a sense is called its first attestation. When an entry here says a word dates from 1588, it means that 1588 is the earliest known written use. The word may well have been spoken for years before that, and an earlier example may yet turn up in some forgotten document. First dates are regularly pushed back as more old texts are searched.

Writing only goes back so far. English was first written in the 7th century, Latin by the 6th century BC, Greek in the 2nd millennium BC. To go further back, etymologists compare languages, which is how we can say anything at all about languages that were never written, such as the ancestor of English, Latin and Greek. That method is the subject of the third guide.

Words change their sound

Pronunciation never stands still. Old English stān, bān and hām became stone, bone and home, while in Scotland the same words went a different way and became stane, bane and hame. In the 15th and 16th centuries the long vowels of English moved again, in the change known as the Great Vowel Shift: before it, bite sounded rather like “beet” and house like “hoose”, a pronunciation Scots still keeps.

The crucial point, discovered in the 19th century, is that these changes are regular. Old English ā did not turn into o in a few words chosen at random: it changed in practically every word that had it. That regularity is what makes etymology possible. It is explained in the guide to sound laws.

Words change their meaning

Meanings drift too, and less predictably. A few examples from English:

Nice comes from Latin nescius, “ignorant”. It entered English meaning “foolish”, then passed through “fussy”, “delicate” and “precise” before settling, in the 18th century, on “pleasant”.
Silly was Old English sǣlig, “happy, blessed” (German selig still means “blessed”). It slid to “innocent”, then “pitiable”, then “feeble”, and finally “foolish”.
Meat once meant any food at all, a sense that survives in sweetmeat.
Deer once meant any wild animal, like its German relative Tier; hound once meant any dog, like German Hund.
Girl in Middle English meant a child of either sex.

Linguists give these shifts names. Meanings can narrow (meat, deer), broaden (Old English docga was one particular breed before dog took over from hound), improve (nice) or worsen (silly), or move sideways by metaphor and association. Unlike sound change, meaning change follows no laws, which is why every step needs written evidence.

Where new words come from

A language gets new words in a handful of ways:

Inheritance: a word handed down from parents to children since before records began. Father, night, water and two have been in English, in one form or another, as long as there has been English.
Borrowing: taking a word from another language. English has borrowed on a huge scale: sky, egg and even they from Old Norse; beef and pork from French (while the animals kept their English names, cow and pig); algebra from Arabic; shampoo from Hindi.
Compounding and derivation: building words from existing parts, as in bookcase or kindness.
Shortening and blending: bus from omnibus; brunch from breakfast and lunch.
Back-formation: edit was formed from editor, not the other way round.
Names: the sandwich is named after an Earl of Sandwich.

Relatives: cognates and borrowings

Words in different languages that descend from the same ancestral word are called cognates. English night, German Nacht, Latin nox (stem noct-), Greek nýx (stem nykt-) and Sanskrit nákt- are cognates: none of them borrowed the word from another. Each inherited it from a common ancestor language.

A single ancestral word can reach English twice, once by inheritance and once by borrowing. Father is the native English descendant of the same ancient word that gave Latin pater, and paternal was borrowed from Latin many centuries later. Likewise hound is native, while canine comes from its Latin cousin canis. Such pairs are why English often has a plain native word and a learned Latin one for the same idea.

Looks can deceive. Words that look alike are not necessarily related, and related words often don’t look alike. English bad and Persian bad mean the same thing and are unrelated; so are Greek theós and Latin deus, both “god”. Meanwhile English wheel and Greek kýklos (as in cycle) are the same word, and so are English two and Armenian erku. What decides the matter is not resemblance but regular correspondence, as the next guide shows.

Folk etymology and origin myths

People like words to make sense, and when a word stops making sense they sometimes reshape it. This is folk etymology. Old English brȳdguma meant “bride-man”; when guma “man” died out, the word was refashioned as bridegroom, as if the groom looked after horses. Hangnail was once angnail, a painful nail, and has nothing to do with hanging. Female comes from Latin femella, “young woman”, and was respelled to match male, to which it is not related at all.

The same urge produces origin stories. Many are told with confidence and repeated everywhere, and many are false. Posh does not stand for “port out, starboard home”; golf is not an acronym for anything; and very few English words before the 20th century began as acronyms. The tell-tale signs of a myth are a story that is too neat, that first appears long after the word did, and that no early written use supports. Entries on this site report such stories only to say whether the evidence bears them out.

A word’s origin is not its “true” meaning

It is tempting to think that the original meaning of a word is the real one: that decimate should only mean “kill one in ten” because that is what it once meant. Linguists call this the etymological fallacy. A word means what the people who use it mean by it. Nobody today calls a pleasant person “ignorant” by saying they are nice. Etymology explains how a word got where it is; it doesn’t say where the word ought to be.

How to read an etymology

Etymologies use a few conventions you will meet in entries here and in other dictionaries:

An asterisk (*) marks a form that is not recorded in any text but has been reconstructed, such as Proto-Germanic *fadēr, the ancestor of father.
From, or the sign <, points back to an older form: nice < Old French nice < Latin nescius. The sign > points forward: Old English stān > stone.
Inherited means passed down within the language family; borrowed means taken from another language; a calque or loan translation copies a foreign word part by part (skyscraper became French gratte-ciel).
Of uncertain origin and origin unknown are honest answers. Many common words, dog among them, have no agreed etymology, and a good dictionary says so rather than guessing.