Origins
The English verb 'say' descends from Old English 'secgan,' from Proto-Germanic *sagjaną, which traces to the Proto-Indo-European root *sekʷ- meaning 'to say' or 'to utter.' It is one of the most frequently used verbs in the English language, consistently ranking in the top ten across all corpora of spoken and written English, and its meaning has remained virtually unchanged for over a millennium.
The Proto-Germanic form *sagjaną is well-attested through its descendants: German 'sagen,' Dutch 'zeggen,' Old Norse 'segja,' Old Frisian 'sedza,' and Gothic 'saga' (a narrative, report). The consonantal differences between these forms — English 'say' versus German 'sagen' versus Dutch 'zeggen' — reflect regular sound changes: the West Germanic gemination of *-gj- produced the doubled consonant visible in Dutch, while English simplified the form through palatalization and eventual loss of the medial consonant.
Outside Germanic, the PIE root *sekʷ- appears in several branches, though with less certainty than some other etymological connections. Latin 'inquit' (he/she says), a defective verb used almost exclusively to introduce direct speech, is generally traced to *sekʷ- with the prefix *en-. Lithuanian 'sakyti' (to say) is a strong cognate from the Baltic branch. Some scholars also connect Old Irish 'insce' (speech, discourse), though this is debated.
Germanic Development
One of the most culturally significant words derived from this root is 'saga.' Old Norse 'saga' (a narrative, a story, a history) comes from the same Proto-Germanic root, formed as a feminine noun meaning literally 'something said' or 'a telling.' The great medieval Icelandic sagas — the Saga of the Volsungs, Njáls saga, Egils saga — were oral narratives before they were written texts, and their very name encodes this oral origin. The English word 'saga' was borrowed from Old Norse in the eighteenth century to refer specifically to these Norse narratives, and later generalized to mean any long, detailed account.
The compound 'soothsayer' preserves the Old English word 'sōþ' (truth, reality), making a soothsayer literally a 'truth-sayer.' The word 'sooth' itself is archaic in Modern English (surviving mainly in 'forsooth' and 'soothsayer'), but the compound demonstrates how 'say' could be combined with other elements to create specialized vocabulary.
Another notable derivative is 'gainsay,' meaning to deny or contradict. This comes from Middle English 'gainsayen,' from 'gain-' (against, from Old Norse 'gegn') + 'say.' To gainsay someone is literally to say against them. The word has a somewhat formal or literary register in Modern English but remains in active use.
Middle English
The past tense 'said' (/sɛd/) shows an irregular vowel shortening that occurred in Middle English. The expected development would have produced a past tense with a long vowel (rhyming with 'played'), but the extreme frequency of the word led to phonological reduction — a common pattern where the most-used forms undergo special erosion. The spelling 'said' preserves the older vowel quality that the pronunciation has long since abandoned.
The phonological development from Old English 'secgan' (/setʃ.ɡɑn/) to Modern English 'say' (/seɪ/) involved several steps: the palatal consonant cluster was simplified, the final syllable was lost (as happened to most Old English infinitive endings), and the remaining vowel underwent the Great Vowel Shift, diphthongizing from /eː/ to /eɪ/.
In modern usage, 'say' functions not only as a speech verb but as a discourse marker ('say, did you hear about...'), a rough approximation marker ('about, say, fifty people'), and a hypothetical introducer ('let's say you had a million dollars'). This functional expansion beyond its core meaning of verbal utterance demonstrates the tendency of high-frequency words to develop grammaticalized functions that transcend their original lexical meaning.