Where English came from

English is often described as a language that borrows heavily, which undersells it. English has a documented history of arriving somewhere, taking the vocabulary and leaving. The result is a language with an unusually large word stock, an unusually irregular spelling system, and a habit of having two or three words where one would do — each carrying a slightly different social flavour. Almost all of it is explained by history.

It is a Germanic language

Despite appearances, English belongs to the Germanic branch of the Indo-European family, alongside German, Dutch and the Scandinavian languages. Its ancestor arrived in Britain from around the fifth century with settlers from the North Sea coast — the groups later described as Angles, Saxons and Jutes.

The Germanic inheritance is not visible in the size of the dictionary but in the parts that get used most. The commonest words in everyday English are overwhelmingly Old English in origin: the, and, man, house, eat, drink, sleep, good, water, hand. So is the grammatical machinery. You can construct a plain English sentence entirely from Germanic material; you cannot construct one from the borrowings alone.

The irregular verbs are a fossil of the same inheritance. Sing, sang, sung and drink, drank, drunk follow an old Germanic pattern of marking tense by changing the vowel. Regular verbs taking -ed are the later pattern, which is why new verbs always join that group.

Then the Norse arrived

Viking settlement from the eighth century onward left a distinctive mark, because Old Norse and Old English were related closely enough to be partially mutually intelligible. Speakers could get by, and the mixing that followed went unusually deep.

Norse gave English sky, egg, window, knife, husband, leg and a great many others. More remarkably, it supplied the pronouns they, them and their — grammatical core vocabulary that languages very rarely borrow. That they were taken at all indicates contact far more intimate than trade.

Sometimes both words survived with a split in meaning. Norse skirt and native shirt come from the same root; so do skill and native craft. English kept both and gave each a job.

Then 1066

The Norman Conquest did not replace English, but it moved it downstairs. For roughly three centuries the language of the court, the law and administration was Norman French, while English continued as the speech of the majority.

This produced the most distinctive feature of English vocabulary: paired words where the Germanic term is plain and the French-derived term is formal. You ask or you enquire; you can be kingly, royal or regal, the three ascending in ceremony as they move from Old English through French to Latin.

The famous illustration concerns food. The animals in the field carry English names — cow, pig, sheep, deer — while the meat on the table carries French ones — beef, pork, mutton, venison. The usual explanation is that English speakers tended the animals and French speakers ate them. That is probably too neat as social history, but the pattern in the vocabulary is real, and the class division underlying it is not in doubt.

Law kept French longest, which is why legal English still runs on words like plaintiff, tort, estate and attorney, and why legal phrases so often pair a Germanic and a French word for the same thing: breaking and entering, will and testament, fit and proper.

Then the scholars

The Renaissance brought a deliberate wave of borrowing from Latin and Greek, as writers reached for vocabulary to discuss subjects English had never needed words for. This is where the scientific and abstract register comes from.

Some of these borrowings were resented at the time as showing off — contemporaries called them inkhorn terms — and many did not survive. Those that did are now unremarkable: education, species, chaos, skeleton, atmosphere.

Greek in particular remains the default source for new technical coinage, which is why a word invented last year for a new medical procedure will still be assembled from roots two and a half thousand years old.

Why the spelling is like that

English spelling is genuinely irregular, and the irregularity has causes rather than being simple carelessness.

Printing arrived in England in the 1470s and began fixing spellings in place. Shortly afterwards, English pronunciation went through a substantial shift in its long vowels — the Great Vowel Shift — which played out over roughly two centuries. Spelling had already begun to set; pronunciation moved underneath it. The result is words spelled for a pronunciation that no longer exists.

The silent letters in knight, knee and gnaw were once pronounced. The gh in light and night represented a sound English has since lost, which German still has in Licht and Nacht.

Some irregularities are worse than accidental. A few spellings were deliberately altered by scholars to display Latin ancestry, sometimes wrongly. The b in debt and doubt was inserted to reflect Latin debitum and dubitare, having never been pronounced in English. The s in island was added on the mistaken belief that the word was related to isle; it descends from Old English igland and has no business there.

What the mess is worth

The obvious question is whether a language this irregular is simply badly built. The trade is real: English spelling is harder to learn than most, and its vocabulary is unusually large.

What it bought is precision and register. Having three words for a concept, arriving from three directions, means being able to choose the plain one, the formal one or the technical one, and to signal something by the choice. Most of the effects English writers reach for — plainness, grandeur, irony — depend on that layered vocabulary and on readers hearing which layer is being used.

None of it was designed. It is the accumulated residue of invasion, occupation, scholarship and trade, and every awkward spelling is a small piece of evidence about where the language has been.