Last night, a friend and I were talking about some Arabic linguistic stuff and, in the course of the conversation, he asked me about words for ‘letter’ (as in ‘letter of the alphabet’) in other languages. I first brought up Greek γράμμα, gramma, and γραφή, graphē, after which he told me that the Arabic word for letter is حَرف, ḥarf. I think both of us immediately realized how similar graphē and ḥarf sound. The discussion then shifted to how Greek gamma is often used to translate ḥet (Arabic ḥāʾ) (and vice versa) in ancient texts.
A little digging showed that one of the earliest uses of حَرف for ‘letter’ was by Sībawayh, a Persian grammarian (of Arabic) of the 8th century. A cursory look at the lexicons shows that there was no cognate Persian word at the time. Although we might then logically assume that ḥarf is a loanword from Greek, there’s an even deeper connection between the two words to be explored.
The Arabic word حَرف has a broader semantic range than just ‘letter’, and has cognates in nearly all branches of Semitic. For example, ḥarf also means “sharp edge, cutting edge of a knife or sword” – a definition that we can see as early as Q 22.11. This connects clearly with חֶרֶב, “sword, edge,” a word common in Biblical Hebrew.
The connection with γράφω, to ‘scratch, graze’ (also γραφή, “representation by means of lines”), can be seen even more clearly in texts like Exodus 20.25: “But if you make for me an altar of stone, do not build it of hewn stones (גָּזִית); for if you use a chisel (חרב) upon it you profane it.” Similar to this, Gamkrelidze and Ivanov, in their derivation of Greek γράφω from Proto-Indo-European *(s)k‘rebh, “scratch, scrape, draw,” cite Iliad 17.599-600, “but the spear of Polydamas had grated the bone’s edge.”
In light of other ancient Semitic cognates such as Ugaritic ḥrb ‘sword’ (and perhaps Akkadian ḫarbu, ‘plow’) we may have here another word to add to the early Indo-Semitic lexical koinē.