That’s the subject line on the e-mail frequent commenter Zackary Sholem Berger sent me with the link I’m about to quote, and I can’t improve on it. Check out this 2006 post from the wonderful site Balashon—Hebrew Language Detective (which I’ve finally gotten around to adding to my RSS feed), featuring Mike Gerver’s impressive etymology for Hebrew אתרוג etrog (the fruit of the hadar tree):
Etrog, on the other hand, is listed in the same book [Ernest Klein's Etymological Dictionary of the Hebrew Language for English Speakers] as borrowed from Persian turung or Mandaic trunga. (The form etrunga is found in Kiddushin 70a.) The Persian word, according to Chaim Rabin’s article “Lexical Borrowings from Indian Languages as Carriers of Ideas and Technical Concepts” (in Between Jerusalem and Benares: Comparative Studies in Judaism and Hinduism, page 25, edited by Hananya Goodman, SUNY Press) comes from Tamil, and is related to matulankam and matulai which mean pomegranate or lemon. (In modern Tamil, pomegranate is matulanpazham, where pazham means ripe fruit.) Rabin says that there is no similar word in Sanskrit, suggesting that etrogs were originally found only in southern India where Tamil and other Dravidian languages are spoken, and only spread to northern India and Persia in a later period (after Sanskrit). I’m not sure what this implies about the question of whether pri etz hadar always meant only the etrog, and whether the etz hadaat could have been an etrog. It is quite possible, of course, that trunga did not mean an etrog, but a different kind of fruit, at the time the word was borrowed from Dravidian, and that it was this other fruit that was only found in southern India. The kam at the end of matulankam (and hence the nga at the end of trunga) are presumably related to kaay meaning “fruit” in modern Tamil. The same root is apparently found in the Persian word naranga (source of naranja in Spanish and hence orange in French and English), which was also borrowed from a Dravidian language. In modern Tamil, naru means “smelly,” so naranga could mean “fragrant fruit.” (Words that mean “fragrant” tend to evolve to mean “smelly” in any language.) Oranges are thought to have come to the Middle East and Europe from northern India, according to the Encyclopedia Britannica, and to there from southern China and Indochina, so the question arises as to why the word would be borrowed from a Dravidian language. One possibility is that the word dates back to the period before the Indo-European conquest of India, when Dravidian languages were spoken in Northern India as well. So the g in etrog would be cognate with the g in orange.
Zackary said confidently “you will like this,” and of course I do. (I thought either MMcM or I had done a post on the tangled history of orange, but apparently not.)