onion and union

Phonologically this is straightforward, but what the heck happened semantically?

Latin ūniōn- meant "oneness, unity" - understandably, since it's derived from ūnus "one". But it also meant "a single large pearl", and "a kind of single onion". Onion is from Anglo-Norman vngeon, oignon, oinion etc, in the 12th-13th centuries, while union is borrowed from French union in the 15th century.

So what's the connection between "oneness" and "onion" and "pearl"? The OED says

According to the classical Latin agricultural writer Columella, the peasants used ūniō for a certain variety of onion because it put forth no shoots, i.e. it represented a single entity. The application of the word to a pearl may represent an independent derivation from ūnus one, alluding to the fact that it was worn alone, or it may be a transfer from the sense ‘onion’, with reference to the similarity in shape.

pashmina and peculiar

*peḱ- "pull out (wool)" in the form *peḱ-s-men- became Persian پشم pašm which means "Wool. Fleece. Hair on the privities, pubes" or if you prefer, "Hair, wool, fur, down; the pubes; anything insignificant or of no moment, anything worthless". This word is found in پشمینه pašmīnah "woolen cloth", borrowed into English as pashmina.

The extended form *peḱ-u-, meaning "herd" and then "wealth", became pecūlium "riches in cattle, private property" and pecūliāris "of or relating to a person's peculium, belonging to a person, one's own, personal, private". Borrowed into Middle English as peculier "Distinguished in nature, character, or attributes from others; unlike others", becoming English peculiar.

Also pecorino and fee

I originally had written that the root *peḱu- "wealth" shifted to "cattle" in other languages but to "wool" in Persian. I changed this based on Glen Gordon's comments. Thanks Glen!

hijra

I just found out that a few months ago New Delhi's court decriminalized homosexuality. In honour of this, the word today is hijra, a term for a transvestite who is considered neither male nor female.

From the OED, I see that Hindi हिजड़ा hijaṛā "eunuch, impotent man" is perhaps borrowed (via Marathi हिजडा hijaḍā and Oriya ହିଜଡ଼ hijaṛa) from Kannada ಹೆಣ್ಣಿಗ heṇṇiga "impotent man, coward", ultimately from Tamil பெண்டன் peṇṭaṉ "hermaphrodite, eunuch". This is the masculine form of பெண்டு peṇṭu "woman".

Wikipedia notes (there is no citation) that Kannada initial h- is a reflex of earlier p-. It seems likely that the Kannada and Tamil words are both derived from the same proto-form.

No relation to Arabic هِجْرَة hijra, the migration of Muhammed to Medina.

Hijras are men who sacrifice their genitalia to a goddess in return for the power to confer fertility on newlyweds and newborn children, a ritual role they are respected for, at the same time as they are stigmatized for their ambiguous sexuality.
- Gayatri Reddy, With respect to sex: negotiating hijra identity in South India

Portraits of hijras.

the story of Verner's Law

It's an amusing tale of the early discoveries in Indo-European linguistics. And parts 2 and 3. Seen on Mr Verb.

mouse, muscle, mussel, musk, nutmeg

Proto-Indo-European *mūs- became the word for "mouse" in English and many other languages. In Latin it became mūs "mouse", then mūsculus, literally "little mouse", borrowed into English as muscle. The OED tells us that the word for "mouse" also has the meaning "muscle" in many Indo-European languages, because of the resemblance of a flexing muscle to the movements of a mouse.

Weirdly, mūsculus is also the source of mussel, I guess because of the resemblance of the shellfish to mice in size. The Greek cognate μῦς was used for mice, muscles and mussels - and also "a large kind of whale", bafflingly.

It wasn't just muscles and molluscs that were compared with mice. *mūs- became Sanskrit मुष्क muṣka "testicle". Related to this is Persian مشك mušk "musk", since the musk-bags of deer were thought to resemble testicles. The Persian word was borrowed into Greek then Latin as muscus, which is the source of English musk. Muscus is also the source of the -meg in nutmeg, thru Old French mugue "musk", which was used to refer to a fragrant herb.

Welsh and walnut

Proto-Germanic *walχo-z "foreigner" shows up in Old English wealh "Celt, Briton", Old Norse Valir "Gauls", French Wallon. It was borrowed into Slavic, for instance Czech vlach, apparently meaning "Italian". The Anglo-French Waleis shows up in the name Wallace. With the adjectival -ish suffix, *walho-z became English Welisc, then Welsh.

Old English walhhnutu is wealh plus hnutu "nut". Etymologically it "meant the nut of the Roman lands (Gaul and Italy) as distinguished from the native hazel" according to the OED.

According to Skeats, wealh in the plural was wealas "foreigners", which became Wales.

turban and tulip

Persian دولبند dūlband "turban" was borrowed into Turkish as tulbant/tuliband, then found its way into Italian as tolipante, making its way into English as turban. It is not known exactly when the change of tul- to tur- took place.

The Turkish word was borrowed again as modern Latin tulīpa, Italian tulipano, French tulipan, tulipe, English tulip. According to the OED, the tulip was introduced into Europe from Turkey in the 16th century. It was so called because of its likeness to the turban in shape.

According to Yule, the dūl- of dūlband is from Arabic dul "to roll". I wonder if the -band is from Proto-Indo-Iranian bandh- as in Hindi bāndhanā "to tie" and bandanna.