I was enjoying Lauren Collins’s New Yorker piece “How a Hazelnut Spread Became a Sticking Point in Franco-Algerian Relations” (archived) but of course kept wondering about the origin of the name of the spread, El Mordjene. Then I got to this key passage:
Cebon, which now employs eight hundred people, has three factories. The one that manufactures El Mordjene is only a few miles from the Mediterranean. The sea inspired the Fouras to give the product its name, which means “red coral” in Arabic.
With that information, I was able to discover that the Arabic word is مرجان ‘small pearls; corals,’ which has a very interesting etymology:
From Classical Syriac ܡܪܰܓܳܢ (margān, “pearl-like”), from ܡܰܪܓܳܢܺܝܬܳܐ (margānīṯā, “pearl”), from Ancient Greek μαργαρίτης (margarítēs, “pearl”), an Iranian borrowing.
At that μαργαρίτης link, we find:
Borrowed from Indo-Iranian.[1] According to Beekes, possibly from Proto-Iranian *mŕ̥ga-ahri-ita- (“oyster”, literally “born from the shell of a bird”).[2] Compare Middle Persian [script needed] (mwlwʾlyt’ /morwārīd/) (whence Persian مروارید (marvârid)), Sogdian [script needed] (marγārt), Sanskrit मञ्जरी (mañjarī), and Avestan 𐬨𐬆𐬭𐬆𐬌𐬌𐬀 (mərəiia).
Among the list of descendants they give Aramaic מרגניתא, מַרְגָּלִיתָא (margālīṯā), as well as Classical Syriac ܡܪܓܢܝܬܐ and the Hebrew loan word מַרְגָּלִית (margalít); English margarite; and Latin margarīta (see there for further descendants), but they neglect to add “see there for further descendants” to the Syriac, which richly deserves it — someone who edits Wiktionary should add the parenthetical.
The OED (entry revised 2000) has this fairly chatty etymology, which disputes the idea that the Greek is an Iranian borrowing:
< Anglo-Norman margarite pearl, Old French margarite, margerite pearl, daisy (12th–13th cent.; Middle French, French marguerite marguerite n.) < classical Latin margarīta pearl < Hellenistic Greek μαργαρίτης (already in ancient Greek, denoting an Egyptian plant; also Hellenistic Greek μαργαρῖτις λίθος and μαργαρῖτις, μαργαρὶς λίθος and μαργαρίς) < Hellenistic Greek μάργαρον pearl-oyster, pearl, μάργαρος pearl-oyster (in medieval Greek also ‘pearl’) + ‑ίτης (see ‑ite suffix¹). The further etymology is unknown; Pliny refers to the Greek word as ‘barbarous’, probably implying that it is regarded as of eastern origin. Although there are parallels in the Iranian group of languages (compare Parthian murgārīt, Middle Persian morwārīd (Persian marvārīd), Pashto marγalara pearl) it is unlikely to have originated there (Iranian scholars consider it a loanword < Greek); compare also Armenian margarit (a loanword < Persian), Sanskrit mañjarī cluster of flowers, (in lexicographical sources) pearl, and Syriac margānīt pearl.
At any rate, it’s one of those words that really gets around, from Georgian მარჯანი (marǯani) to Hausa murzani to Tagalog margarita.
one of those words that really gets around
Even Mooré: muizãanfo, plural muizãana (noun-class assignment as for seeds, logically enough.)
You didn’t mention that other edible grease, named after margaric acid, itself named for its pearlescence.
Mooré: muizãanfo
The rather odd change r > i [j] is probably by analogy with mui “rice” (singular muiifu “grain of rice.”)
Western Oti-Volta has a historical change *r > j, but I can imagine no scenario where the loan could possibly be old enough to have participated in that.
Kusaal has pɛɛl, doubtless from some lost substrate language.
Also the European Jewish family name Margolies/Margolius/any number of other spellings.
I recall a (decluttered or lost) doorstop of a translated Hungarian novel with a female character named `Bioodpearl’.
That’s the extent of my memory of the book but I continue to wonder about the name.
Rather to my surprise, my dictionaries turn up quite a lot of Oti-Volta words for “pearl.” Less surprisingly, not much sign of any cognate sets. I suspect that quite a few basically mean “bead.” The stem of Gulimancema laanbu plural laani “pearl” matches Kusaal lan, Mampruli lanni “testicle” well formally … semantically, it’s good enough for Afro-Asiatic work ..
The Toende Kusaal dictionary has simii “pearl, bead, necklace”, which looks as if it should be the exact cognate of Agolle Kusaal Simiig “Fulɓe person.” The connexion of ideas is not evident to me …
A search on [“Bloodpearl” Hungarian novel] turns up nothing.
Well, according to Beekes, the word is “mostly assumed to be from Iranian” and he points to critical discussion of this hypothesis. And the derivation from Proto-Iranian *mŕ̥ga-ahri-ita- is not his own but a “suggestion by Gershevich” which Beekes simply quotes without further comment. Beekes’ treatment is more tentative than Wiktionary implies.
That was a good article!
Another derivative familiar to readers of the Arabian Nights is the name of a character in the story of Ali Baba -“Morgiana”.
Worth noting: French perle usually just means “bead”. I suspect that ambiguity is what’s behind some of the Oti-Volta attestations.
That would certainly make sense. The English glosses are often pretty erratic, and sometimes outright wrong, even in Niggli’s generally very good Mooré dictionary.
re Y: Hence the spelling margarine, though everybody I know has always pronounced it as if margerine.
Seems that the origin of “pearl” itself is also unclear, once you get beyond French:
https://en.wiktionary.org/wiki/pearl#English
(The link to Old French perle actually suggests yet another speculative etymology.)
Looking at the Avestan and Sanscrit besides the other Iranian words mentioned makes me think that it is a loan into Iranian and Sanscit as well, from an unknown language.
The Toende Kusaal dictionary has simii “pearl, bead, necklace”, which looks as if it should be the exact cognate of Agolle Kusaal Simiig “Fulɓe person.” The connexion of ideas is not evident to me …
In a spirit of deep scientific enquiry, I was just looking up pictures of random Fulɓe people in traditional dress on the intertubes. Lots of bead necklaces and bangles and beaded braids feature in pictures of the ladies, especially. Maybe the Toende word really is a metonymous use of “Fulɓe.” No example sentences given, alas.
On the other hand, the Toende New Testament has sĩwĩi, which looks like compound of some “cran-” morpheme with wĩi “rope” (Agolle nwiig.) Folk etymology? In which direction?
@ LH, re Hungarian novels
It was `Parallel Stories’ ( ~ 1100 pages) by P\’eter N\’adas. It must be somewhere in the first third or so of the book because that’s as far as I got.
Ah, this is what you’re remembering:
It’s from “A Genteel Building.”
Avestan 𐬨𐬆𐬭𐬆𐬌𐬌𐬀 (mərəiia)
What???
This looks as if a Wiktionary editor misread transliterated Avestan mərəγa- ‘bird’ as *mərəya- (as an element in Gershevitch’s proposed **mr̥ga-ahri-ita- ‘proceeding from a shell that looks like (the outstretched wings of) a bird’, i.e. ‘oyster-born’), and then “corrected” the Avestan spelling of *mərəya- to *mərəiia-, and then wrote that in Avestan script.
For the typology of ‘bone’ ~ ‘(mollusc) shell’, note the proposed relationship of Greek ὀστέον ‘bone’ to ὄστρεον ‘oyster, bivalve’ and ἀστακός, ὀστακός ‘lobster’. For the typology ‘bivalve shell’ ~ ‘wing’, note Latin pinna ‘feather, wing’ and Modern Latin Pinna ‘genus of pen shells’ (due to Linné?).
If I recall correctly, Gershevitch’s account of the morphology of the element -ahri- (‘bone’ > ‘shell’) is very complicated and does not immediately impose itself.
Short comment because I am on holiday, on my tablet, on bus station wifi.
yiddish has מאַרגאַריטקע | margaritke for “daisy”, which wiktionary has as a specifically romance sense of the word*. which could make it a quite early arrival to the lexicon (especially if one sides with weinreich about the “romance component”), or a quite late one (especially if it always had the slavic suffix), or anywhere in between (especially if it passed through polish or romanian or german or something else on the way). i’m damn sure not gonna do it, but it would be wonderful if someone turned schaechter’s full material for Plant Names in Yiddish – or even just his complete source-list (only א and ב were printed) into a no-frills lexicon – it wouldn’t be comprehensive, but would be such an improvement.
.
* because the colored center looks like a pearl in the oyster of the petals? because there is or was a coral-centered variety of daisy in a relevant area?
Same in German. Even large wooden beads are Holzperlen.
One of the most baffling facts about English spelling or pronunciation, right up there with the mendacious -e of determine or intestine. “Hey! We didn’t have enough Featherstonehaughs and Worcestershires yet! Let’s have one more!”
This is why I hate the use of unaltered γ with Latin letters. Especially in italics. The IPA gave you a perfectly clear ɣ, it’s in Unicode, it’s used in a few actual orthographies, use it.
That, BTW, is a Middle Chinese loanword in Proto-Turkic that was passed on to Hungarian.
The botanists still maintain the legal fiction that when you invent a new taxonomic name that wasn’t already a Latin word, you thereby create a new word in the Latin language. The zoologists have long given that up.
French marguerite > German Margerite for the really big species. (The small ones are Gänseblümchen.)
Ah, yes. Das Glasperlenspiel …
No pearls are harmed in the playing of this game. In fact, one gathers that no glass beads are involved either. That’s just the name of the shop, love.
https://genius.com/Bonzo-dog-doo-dah-band-shirt-lyrics
I am confused by the “margarine pronounced as margerine” claim. The second syllable is reduced so the vowel is a schwa. Both “a” and “e” can be used in English orthography to represent a vowel-reduced-to-schwa. What contrary-to-fact pronunciation is the “a” rather than “e” supposed to be implying, holding the stress pattern constant?
FWIW, Donovan’s song “Marjorie Margarine” puts secondary stress on the final syllable, thereby unreducing the vowel to FLEECE. I don’t know if that’s a standard feature of some dialect or just an artistic-license thing that made the rhymes with “queen” and “scene” work better.
ETA BONUS FUN ORTHOGRAPHIC FACT: If you dig deep enough into Donovan discography trivia you will learn that the song “Marjorie Margarine” was spelled “Marjorie Margerine” on the German and Italian releases of an album containing it. Not from a point in his career where he was doing very well commercially, so cross-border quality control may not have been very extensive.
What contrary-to-fact pronunciation is the “a” rather than “e” supposed to be implying, holding the stress pattern constant?
Nothing to do with stress: the pronunciation of the written g as [d͡ʒ].
In fact, I have primary stress on the final syllable in “margarine”, which has the FLEECE vowel. Pretty sure that’s UK-standard (though my own idiolect is fairly Donovanian.)
I do know one person who has /g/ for the normal /d͡ʒ/. It always sounds weird to me. Spelling-pronunciation stuff. He’s always said it like that.
Ah. Like the distinction between “purge” and “purgatory.”
I once read a novel, in which the hero, who I eventually realised must have rejoiced in the given name “Roger”, was consistently called “Rog” by his intimates. Took me right out of the moment every time. I must admit that “Rodge” looks pretty silly. Probably the author should just have called him Bill.
@DE: See also “Reg”, and British national native meal “meat, potatoes, and two veg”, or so I’m told.
@rozele: I’d have guessed daisies are “pearls” because of the pearly white “petals”.
@David Eddyshaw: The protagonist of the Space Quest game series is named Roger Wilco. The narration often calls him Rog.
One Fat Englishman, by any chance?
Consider the minimal pair “Rog” (clipped “Roger”) and “rog” (clipped “interrogatory”*), which modulo capitalization are homographs but pronounced differently. Or the parallel “Reg” (clipped “Reginald”) and “reg” (clipped “regulation”).
*In the lawyer-jargon sense. I don’t use the “rog” clipped form myself but I have colleagues who do.
I keep thinking The Flaming Lips’ “She Don’t Use Jelly” has the word “margarine” in it. It doesn’t.
@Xerîb: Thanks for explaining the Avestan. The Sanscrit still doesn’t really fit, except if there was folk etymology involved.
@rozele: Russian also has margaritka for “daisy”. Polish and Ukrainian have stokrotka, Belarusian ramunak, Lithuanian ramunė.
Fowler 1926 sv “gaol, gaoler, jail, jailor, &c.”
@DE:
The OED agrees with you, and in fact only lists ultimate-stress pronunciations for BrEng, but I’m pretty sure that information was already obsolescent when the entry was last revised in 2000. I only ever hear /ˈmɑː.dʒəˌɹin/ from my (East London born and bred) partner and their family, even from folks born in the 1950s.
(A good indicator of just how obsolescent is that the OED includes a pronunciation with /g/ for orthographic <g>, without even marking it as “dated”.)
without even marking it as “dated” — do they ever use labels like “dated” in the pronunciation section? I don’t recall seeing any. For margarine the change is noted in the etymology section, which says that the /g/ pronunciation “became rare in the second half of the 20th cent.” It also says that the stress was originally on the first syllable, and shifted to the last syllable “outside North American English” in the 20th century. Maybe the initial-stress pronunciation never completely went away, and your partner’s family are holdouts?
I listened to the first few dozen clips on Youglish for UK “margarine” and heard several with first-syllable stress (one speaker did it both ways, first and last) and none (so far) with /g/.
This 5 Apr 2024 BBC podcast has four well-spoken Brits saying “margarine” several times from the 50m30s mark on. The stress seems variable.
Re Fowler’s puzzlement at “mortgagor”: it’s part of a matched set with “mortgagee,” and people often learn them as a pair, with the symmetry thus presumably overriding a default orthographic pattern. Consider an example going the other direction: obligor and obligee, where at least in my experience (some online references make contrary claims but I’m suspicious of what their actual source is) both use the “hard” /g/ in pronunciation, i.e. both pattern with obligation rather than oblige.
The spelling of “gaol” is separately odd because the letter-sequence “ao” rarely (not immediately thinking of another instance although I’m not claiming there aren’t) matches with the FACE vowel. So the one oddity makes the other more tolerable since it’s just an all-around peculiar spelling/pronunciation pair?
@ LH last night:
THANKS!
… the newborn was registered with her mother’s last name and given the first name of Gyöngyvér. …. She persistently and darkly hated her unknown unmarried mother because of this name, because of the gyöngy, meaning pearl, and because of the vér, meaning blood. … And the {\bf words with ambiguous meaning} must have rattled in her head, because her forehead was throbbing with pain….
This is much more than I remembered, but I see now why it caught my attention. Gyöngyvér didn’t sound like a run-of-the-mill name but I never got far enough to understand its nuances.
https://en.wiktionary.org/wiki/Appendix:English_words_where_G_is_pronounced_exceptionally of course there is
@j. W. B.: The word that “gaol” immediately reminds me of is “chaos”, with the FACE vowel.
I didn’t see any puzzlement about “mortgagor” in Fowler’s mention of it.
Thanks for “obligee”. I would never have imagined a /g/ in that word.:
https://en.wiktionary.org/wiki/Appendix:English_words_where_G_is_pronounced_exceptionally of course there is
Besagew (/ˈbɛs.ə.ɡju/) is a word new to me:
Etymology:
And there’s a nice image of one.
That list has “Dodgson (one pronunciation)”. What other pronunciation of Dodgson is there, and does it really include /g/?
There is the pronunciation identical to that of my own name, Dodson, which is originally a pronunciation spelling.
“Obligee” is interesting because online references pretty much consistently disagree with me. Possibilities: 1) the online references are not based on anyone actually going out and listening to people use the word but on sort of a lazy/automatic assumption that the usual rule based on the following vowel must apply; 2) the context in which I would have had any occasion to hear the word said aloud or say it aloud myself (NYC legal/financial jargon subcultures) has somehow deviated in recent decades from a more general historical pronunciation and fieldworkers haven’t caught up with that development*; or 3) I am actually hallucinating my own lived experience with the word.
I guess Fowler noting something as “very anomalous” need not logically entail that he was puzzled.
*I feel like I have even heard people joke about it – the word not its pronunciation – because who is whom as between obligor/obligee in a given situation often gets people muddled up in a way that understanding who is whom in comparable legal-jargon pairs like bailor v. bailee or vendor v. vendee or trustor v. trustee does not.
gaoler v. prisoner
Further special pleadings from algae and Doctor Georgios Georgakis
On those occasions when as a chemist I said “decyl” or “decene”, I used /k/ as in “decane”.
@ki
I suppose that prevented you from mixing up a litre of decyl and a decilitre…
In Modern Hebrew, מַרְגָּנִית marganit is used colloquially for ‘daisy’ (perhaps under the influence of the Yiddish word rozele discussed.) Normatively (“correctly”) it is used for Lysimachia, formerly Anagallis, pimpernels (including the scarlet pimpernel).
pimpernel saus WP, <. MedL pimpinella, “most likely from piper (‘pepper’) because its fruit resembled peppercorns”.
in “ahem, ahem” whichever clancy brother chants the rhyme has “margarine” with final-syllable stress (to rhyme with “clean”) in his early/mid-20thC irish english (presumably reflecting the norms of tipperary children around 1930).
@rozele: That sounds to me like Paddy.