English eight names the integer after seven and before nine. Speakers also use it as a noun for anything counted as a set of eight: a racing crew, an eight-cylinder engine, the eight of spades. The number sense has barely moved since the Old English period. Spelling, pronunciation, and a few later labels built on the count are what changed.
The origin of the word eight is not a French or Latin loan. It was already English when the earliest vernacular texts were written.
Inherited, not borrowed
Old English had eahta and æhta, with regional variants such as ehta and ahta. That form is a direct continuation of West Germanic, not a taking-over of Latin octo, even though the two numerals are related at a much older level.
By about 1200, Middle English spellings include ehte. By the late fourteenth century eighte is well established, and that is the spelling that settles into the modern word. Some dictionaries date first use to the fourteenth century because that is when their citations show the familiar written form. The spoken and written word is older: it is attested in Old English, well before the year 1000. Historical dictionaries class it as inherited from Germanic.
The line is therefore continuous: Proto-Germanic into Proto-West Germanic into Old English, then Middle English, then the modern language. Contact with Norman French supplied English with thousands of other everyday words. It did not supply this one.
Germanic and Indo-European ancestry
Comparative reconstruction gives Proto-Germanic ahtōu (also cited as ahtau). From that stage come Old Saxon ahto, Old Frisian ahta, Old Norse átta, Gothic ahtau, Dutch and German acht, Swedish åtta, Danish otte, and English eight. The West Germanic form is ahtō. Scots still has aucht beside the southern type.
Behind the Germanic numeral sits Proto-Indo-European oḱtṓw. Some scholars write a laryngeal in the first syllable (h₁oḱtṓ(w) or h₃eḱtō); the consonanatal skeleton is not in dispute. That reconstruction accounts for a wide set of relatives: Latin octo, Greek oktṓ, Sanskrit aṣṭau, Avestan ashta, Old Irish ocht, Welsh wyth, Breton eiz, Lithuanian aštuoni, Old Church Slavonic osmi. From Latin come Italian otto, Spanish ocho, and French huit (through Old French oit). Tocharian has forms such as okt.
English eight and Latin octo are cousins, not parent and child. They split long before either language existed. That is why German acht still looks closer to English than French huit does, even though French and English later spent centuries in contact. The French word looks odd because Latin octo went through a series of Gallo-Romance sound changes; the English word never passed through that mill.
The same PIE numeral also sits behind learned English formations that look Latin rather than native: October (the eighth month in the older Roman calendar), octagon, octopus, octet. Those were borrowed or coined on Latin and Greek octo- / okto-. They stand beside eight as parallel descendants, not as derivatives of the English numeral.
Compounds inside English are native. Old English already had eahtatīene for eighteen, eahtatig for eighty, and eahteða or eahtoða for eighth. Modern eighteen, eighty, and eighth continue that pattern rather than switching to a Latin oct- stem.
Why the word is spelled with gh
The letters gh in eight are a Middle English scribal habit, not extra sounds that speakers later invented. Scribes used gh for a velar fricative — the kind of sound still heard in German acht — especially before t. The same convention produced night and fight. In late Old English the middle consonant could also be written with a yogh.
In later English that consonant faded in this environment. Once it was gone, the preceding vowel developed into the modern diphthong, so eight is now pronounced like ate. Homophones include ate and, in some accents, ait. The spelling froze while the sound kept moving. The gh is a fossil of a real consonant, not a lost extra syllable.
A dual shape, and an unsettled analysis
The PIE form oḱtṓw ends like a dual, the grammatical number used for natural pairs (the same slot that appears in the word for two). That has tempted linguists to read eight as “two fours,” perhaps originally two sets of four fingers. Avestan ašti- can denote a breadth of four fingers, and a similar stem has been compared with Proto-Kartvelian otxo- “four,” sometimes taken as a loan from an early Indo-European form meaning four fingers.
The suggestion is not a consensus. The ordinary PIE word for four is kʷetwóres, which does not match oḱtṓw in a transparent way. Reference works sometimes note that the numeral looks dual in form without agreeing on a dual of what. Eight continues an old dual-shaped numeral whose further breakdown, if it ever had one, is not recoverable with certainty.
Tamil eṭṭu and Malayalam eṭṭŭ look similar by accident. They are not part of the Indo-European set.
Labels built on the count
The cardinal meaning has been stable for as long as English has been written. What the language added later are names for things that come in eights, or for shapes and situations that the digit suggests.
From about 1600, figure of eight describes a doubled loop: a racecourse, a knot, a path on ice. In the 1690s piece of eight named the Spanish dollar, a coin worth eight reales and often marked with the numeral 8. The rowing sense — an eight-oared shell or its crew — is recorded from 1847. Talk of an eight-hour working day as a fair length of labor appears by 1845. Behind the eight ball, “in a bad position,” is American pool slang from 1932: the black eight is the ball you do not want sitting in the way of a shot. An eight as an eight-cylinder engine is a later mechanical use of the same counting habit.
None of these senses redefines the number. They borrow the count as a convenient label. The numeral itself stayed put.
A family still visible in the consonants
Set English eight next to Gothic ahtau and Latin octo. The skeleton is still there: a velar, a t, and a rounded vowel. Sixteen centuries of English spelling have buried the velar under gh and turned the vowel into a diphthong, but the match is close once the letters are read as history rather than as a guide to modern sound. That inherited match, not a medieval borrowing, is why the word looks the way it does — and why its German and Dutch cousins still keep the older consonant in plain view.