dies

Dies (DEE-ays, "day") is one of the load-bearing words of Latin, and the record shows it nearly everywhere: 11,754 occurrences across 347 works — among the widest distributions in the corpus. It lives first in the historians and encyclopedists. Livy's Ab urbe condita ("From the Founding of the City") alone holds 1,790; the Naturalis Historia ("Natural History") 854; then the Letters to Atticus (330), the De Medicina ("On Medicine," 321), the Historia ecclesiastica gentis Anglorum ("Ecclesiastical History of the English People," 300), and the Noctes Atticae ("Attic Nights," 235). The word measures out narrative time — the unit in which a history advances, a cure is counted, a letter is dated.

The record carries a single lexicon entry, Lewis & Short (A Latin Dictionary, 1879), and the digest preserves not a clean gloss but the entry's grammatical head: it catalogues an unstable declension, spelled dĭes with a short first syllable but also dīes with a long one, cited from Livius Andronicus, Vergil, Horace, and Terence — old poets in whose mouths the word already wobbled. What it records is a word so ancient and so constantly spoken that its case-endings never fully settled.

For its ancestry the record matches one etymology pointer: dies is treated in Ernout-Meillet, Dictionnaire étymologique de la langue latine (DELL, 1932; 4th ed. 1959), s.v. dies (scan p. 201, entry #3136). The pointer carries no reconstructed root, so only what it supports is stated — the derivation is the province of that dictionary — and no root is imported it does not supply.

The citable surfaces run entirely through Livy's first two books — die, dierum, dies, diebus, diem ("on the day," "of days," "the day," "in the days," "the day" as object) — declining through its cases as Rome's earliest days are told.

If a word's case-endings never quite settle, is it still one word, or the wear-mark of everything the language did with it?