Victor Mair has a post at the Log that is not only an impassioned ode to the Allium genus—”My uncontested favorite of all the Allium species, however, is garlic (Allium sativum)…”—but a treasure-trove of linguistic tidbits in both Chinese (he discusses jiǔcài 韭菜, “a kind of chive”) and English. I had known about ramps, “a kind of wild garlic… only available for six weeks in the spring,” but had not known the etymology, which is excellent. Bob Shackleton said in the comment thread: “And most wonderfully, we have here the English descendant of the original Indo-European word for onion (*krem-) – one of the few surviving relatives of Greek ‘kromion’!” The OED elaborates, in its entry for rams (which it takes as the original English form and which it refers you to if you look up ramp; the entry was updated in June 2008):
Cognate with Middle Low German rāmese, rēmese, Old High German ramese (German (chiefly regional) Rams, Rambs, Ramisch, Ramsche, Ramus, etc.; > Danish rams, †ramse (now chiefly regional, except in the synonymous compound ramsløg)), Norwegian rams (now regional, except in the synonymous compound ramslauk), Swedish rams, ramsk (c1580; now regional, except in the synonymous compound ramslök) < the same Indo-European base as Early Irish crem (Irish creamh), Welsh craf, cra, Lithuanian kermušė, Russian čeremša, all in sense ‘wild garlic’, and ancient Greek κρόμμυον, κρόμυον onion. Compare post-classical Latin ramusia (10th cent. in a German source probably ultimately of British origin), ramusium (c1025 in a British source), ramuscium (a1125 in a British source), all probably < Old English.
The β. forms [ram, rame] result from analysis of the α. forms as plural. It is uncertain whether the (rare) late Middle English forms show singular or plural forms; compare also ramsey n. and ramsons n.
In the thread, Piotr Gąsiorowski points out: “Ramson comes from the weak-noun plural hram(e)san, and ramsons is etymologically a double plural like children.” As for rampion (a name for several different plants), it’s entirely unrelated, with a complicated etymology of its own about which the OED goes on for paragraphs; a cognate is German Rapunzel—which I had no idea was primarily the name of a plant; I knew only the gal with the hair—and another is French raiponce, about which marie-lucie writes:
Looking up raiponce again on Wikipedia, I see that the Rapunzel of the story is called Raiponce in French, just like the plant. Having only read Grimm’s stories in English, I did not know the equivalent French name.
As for the origin of the French word, the TLFI (French online dictionary) gives raiponce as basically a borrowing from an Italian word attested as raponzo, raponzolo, raperonzolo, from Latin raponciolum (but the Latin word could be a learned reformation based on one of the Italian forms). The suffix –onz– seems to be unexplained in Italian, so the apparently basic word raponzo could be an adaptation of a word from another language, perhaps a Germanic one such as rapunze, the presumed stem of Rapunzel.
The OED says:
The Italian word has frequently been regarded as a derivative (formed with several suffixes) of rapa rape n.5 (see further Französisches etymol. Wörterbuch at rapum), but recent authoritative dictionaries of Italian consider this more likely to be a folk etymology (see e.g. M. Cortelazzo & P. Zolli Diz. etimol. della lingua italiana (ed. 2, 1999) at raperonzolo). Although post-classical Latin rapunculum, etc., has sometimes been suggested as the ulterior etymon of the plant name, it is much more likely to be an adaptation of one or more vernacular forms, given its late attestation and close formal resemblance to the latter.
A double plural?
Thank you for mentioning that post, which is not only of linguistic but also of botanical and culinary interest! More later.
Oh, interesting. Rapunzel’s hair-tail as a reflection of onion-greens, and her cell as the ground in which the onion is buried? I immediately remembered the Russian folk riddle, “Сидит девица в темнице, а коса на улице” – “A maiden sits in a dark cell ,but her braid sticks out” – but the answer isn’t ramps, it’s carrot. Which makes me think, is carrot also cognate to PIR *krem??
Shelley, the “r” in children is one plural, cognatewith “s”. And then the “en” is the second plural, like the “en” in “oxen.
Ramps – the wiki on allium has:
Allium tricoccum — wild leek, ramp and look ye here, Allium ursinum — buckrams, wild garlic, broad-leaved garlic, wood garlic, bear’s garlic
– “buckrams”!
A. ursinum is Eurasian and A. tricoccum is an American species. It’s a huge genus with dozens on species in Eurasia and North America. The wiki on a tricoccum discusses the connection with A. ursinum and pints out the double plural in the word “ramson”.
Rapunzel means “wild garlic”? I bet there’s a submerged myth hidden in there somehwere.
No, rapunzel is the German name of a different type of plant altogether, a kind of bellflower.
In Germany, the plant Rapunzel from the fairy tale is identified as either lamb’s lettuce (Valerianella locusta) which is the version that is familiar to me or Campanula rapunculus, a type of bellflower which wikipedia says is also known as rampion.
I tried ramps for the first time last week, fried on rice. I had heard that they are way beyond garlic in pungency, but these, while quite tasty in a not-quite-garlic not-quite-onion way, didn’t seem to be any stronger than garlic. I mentioned that to the guy at the farmers’ market who sold them to me and he said it’s the raw ones that are beyond garlic in pungency, so eating them raw will make your breath offensive for hours, if not days.
So: Does rampion affect one’s breath as much as ramps? That might explain why Rapunzel was locked up in a tower: no one could stand her breath.
As for the singular, I imagine asking for one ‘ramp’ at a farmers’ market in Appalachia would get the same kind of reaction as ‘My Cousin Vinny’ got when he was offered grits and asked “what is a grit?”.
Oh, sorry for conflating ramps and rampions and Rapunzel’s mother’s pregnancy food cravings. As I can see from the comments at the Log, I wasn’t the only one to fall for this.
Ramps are traditionally pickled in Russia (only leaf-stems, without leaves), but I ate them raw too, when, one a late-spring mountaineering trip to the now-off-limits Svaneti Range, we set up camp in middle of a wild ramps patch. Their taste and texture are great but one probably shouldn’t eat them in bulk quantities, judging by a tachycardia I developed that night.
I forgot to mention:
When I was in college 40 years ago, a classmate from West Virginia told me that her home-town newspaper once mixed a small amount of ramps oil with the ink as a joking homage to the local delicacy. As I heard the story, the Post Office told them if they ever did anything like that again, they would never have another paper delivered by mail, because the local PO stank so bad the mail sorters couldn’t stay inside for more than a few minutes at a time.
@Shelley:
The Old English plural of cild ‘child’ was cildru. The final -ru consisted of the original stem-final consonant -r- (which reflects Proto-Germanic *-z- from still earlier *-es-) plus the actual nom./acc.pl. ending -u. The stem suffix was lost in the OE singular, so the -r- was reinterpreted as part of the plural enging.
In Middle English, the plural of child developed regularly into childre ~ childer, bit since the form apparently did not look plural enough, an extra plural marker, -en, borrowed from the so-called weak declension (as in the modern survivor ox, oxen) was added to it for good measure.
PG: since the form apparently did not look plural enough
Yes, this sort of thing does happen when old types of inflection have become rare and are no long identified as such. In this case, (e)r (still used to form plurals in Norwegian, for instance) no longer had a recognizable function in English, so (e)n, which did, was added to the previously plural form. At the time, besides oxen, there were still recognizably plural forms in n in the words kine (the old plural of cow) and swine (compare singular sow), later replaced by the now regular plurals in s, cows and sows respectively.
Two corrections:
(1) kine is a secondary double-marked plural like children: Middle English ki, an umlauted plural of cow (like mice from mouse), plus an extra -n
(2) Old English swīn was a heavy-stem strong neuter with an unchanged plural (like “deer” and “sheep”). It was not a plural of a weak noun, and not a plural of OE feminine sugu (the ancestor of modern “sow”), despite their superficial similarity.
I can’t believe I’m the first person to have used the phrase “et alia allia,” but so it appears to be (as far as Google knows).
Ahoy, Hat: I’ve just seen a verb I’ve never seen before.
‘Ed Meyer, of Savills estate agent in Cambridge, nonsensed Mr Furlow’s predictions and said: “It’s a very difficult one to predict but we believe we are going to see a 10 per cent rise in house prices in Cambridge over the next three years and that sort of growth will continue after that.”‘
“Childer” is still found in Ulster English.
marie-lucie, the r in ‘children’ isn’t actually related to the Norwegian plural forms. The latter are actually cognate with the normal English s plurals, both of which are continuations ultimately of the normal PIE plurals of the common o and ā (=eʜ₂) stems.
In ‘children’, the r wasn’t a plural marker originally, but a stem marker, as Piotr says. Eventually this stem marker was reinterpreted as a plural suffix, but this was never especially common in Old English, which is presumably why it was eventually supplemented with the n-stem ending.
I’ve just seen a verb I’ve never seen before.
I hadn’t seen it either, but I figured it was too good not to have been used, and sure enough, Google Books finds:
Richard Baxter, Church-history of the government of bishops and their councils abbreviated: Including the chief part of the government of Christian princes and popes, and a true account of the most troubling controversies and heresies till the Reformation (1681), p. 368: “Some will think it is well that the Councils for above 1000 years had so few that understood the original language, or else they would have so tost and torn, and sensed and nonsensed the Scripture, that they would have made it quite another thing.”
James Russell Lowell, “An Oriental Apologue” (1849), stanza XVIII:
Robert Alexander Wason, Happy Hawkins (1910), p. 67: “I nonsensed a while, tryin’ to get her to laugh an’ cut up, but not her.”
Mathias Freese, The i Tetralogy (2005), p. 176: “My mind is nonsensed with debris from the past.”
Hmm, and now that I think to check the OED, of course they have an entry (updated December 2003), with the Lowell and Wason quotes, but not (obviously) the Freese or (surprisingly) the Baxter—I’ve antedated them by 181 years!
Er, 188 years. Their first transitive citation is from 1869 (R. F. Burton, Explor. Highlands Brazil II. xxiv. 355, “M. Millivet..has grammaticized and nonsensed the word to ‘Santa Sé’”).
PG, Nelson, I thought I had learned all this years ago and did not need to double check, but it looks like I was wrong on all counts. Thank you for providing corrections.
Do you know whether the “additional” (e)n was added in the same general period to give children and kine, or was there a substantial delay for one of them?
I also wonder about the r in OE cildru: if the stem for both singular and plural was once cildr, did this stem-final -r have a function in the first place? under what conditions did it get lost? are there other examples of OE r-final stems losing the consonant? and do cognates in other Germanic languages (eg Kind, etc) also show traces of an original stem-final r?
@Marie-Lucie
The hypercharacterised -en plurals were commonly formed in the southern dialects of Middle English, where the nasal ending enjoyed a period of increased productivity. Another common example is eiren, pl. of ei ‘egg’ (OE ǣġ, pl. ǣġru, of course akin to German Ei, Eier)
The nouns in question are old s-stem neuters, a type well known from most branches of Indo-European. We have it in Latin, for example: genus, genera from older *genos, *genes-ā. Note that the -us of the nom./acc. sg. in is not an inflectional ending (unlike the nom.sg. ending of lupus ‘wolf’) but part of the stem. Old *-os became Germanic *-az, which was regularly dropped in West Germanic if no inflection followed, that’s why we don’t see it in Old English. But in the plural the stem-forming suffix *-es- (> Germanic *-iz-) was always followed by a case ending which protected it from word-final reduction. That’s why pre-Germanic *-es-ā developed into Proto-Gmc. *-izō and eventually into Old English -ru. Final *r was not lost in the ancestor of Old English, but in this case the PGmc. consonant was *z, not *r, and final *z was lost in West Germanic before it underwent “rhotacism”, changing into *r in other positions. That’s why the most productive class of Germanic masculine nouns has no ending in the nom.sg. in Old English or Old High German: PGmc. *stain-az ‘stone’ > OE stān, etc.
PG, thank you, now I recognize what my problem was! It is not the r that has been lost (as I erroneously understood from the wording), but the earlier PGmc *z, itself from earlier *s which is attested in Latin as an old morpheme. In dealing with OE and later stages of English in a synchronic perspective (that of OE, not of Modern English), we can say that r is part of the plural stem, as its final consonant, which does not mean that it has ever been a plural morpheme in that language. My mistake was to interpret your “stem-final consonant” (which was correct) as “post-stem consonant”, making me wonder what its function was.
Great stuff! I think the first and only time I heard “ramps” was on Top Chef or another cooking show a year or two ago, and the word baffled me (like another favorite of that show, “sunchoke”).
My only regret is that Macbeth’s “rump-fed ronyon” seems sadly unrelated to any of this.
re: “nonsensed” –
So is the newer use more or less as a synonym for UK verb “to rubbish” (as in, disparage something previously said)?
Well, I don’t think there really is a “newer use” that one could describe lexicographically; as far as I can tell, it’s occasionally reinvented as a nonce word rather than being a continuous part of some group’s vocabulary, so each time it means what the user wants it to mean (which is usually clear from context).
each time it means what the user wants it to mean
Indeed, in Baxter it means ‘interpret nonsensically’, in Lowell it means ‘condemn as nonsense’, in Wason it means ‘talk nonsense’, in Freese it means ‘fill with nonsense’, in Burton it means ‘turn into nonsense’.
“I thought I had learned all this years ago and did not need to double check”
Hubris being an equal opportunity soporific better kept in the glass tower than in a glass cockpit et cetera.
Rapunzel is discussed in this 2018 thread, starting with marie-lucie’s comment.
Thanks LH for mentioning it. I did think that I had written something about it earlier, and I see that some of the information there is better than what I wrote here, especially about the plant(s) at the origin of the name Rapunzel.
Ramson soup – leaves plucked before the plant goes into bloom, boiled in water and cut or blended – is a thing of awesomeness.
Child-r-en is exactly like Dutch kind-er-en. Just Kind-er in German, where -̈er has gone on a rampage as a productive plural suffix/transfix.
I remember pickled anzur onions (Allium oreoprasum) sold in jars way back.
Child-r-en is exactly like Dutch kind-er-en.
Morphologically yes, but the roots are not cognate. Kind- is cognate with Latin genus ‘kind, type, sort’, plus the suffix that surfaces in English as -th, but the etymology of child is disputed. The only secure cognate is Gothic kilþei ‘womb’, which is one of the meanings of cild in OE, as the various Scandinavian synonyms in kul-, kol- are not regularly related. With a different root extension, child is probably connected with Gothic kalbo ‘calf’ and Latin glæba, glēba ‘lump’ (cognate with clump and the source of glebe ‘the land associated with a parsonage’).
JC: Latin glæba, glēba ‘lump’ (cognate with clump and the source of glebe ‘the land associated with a parsonage’).
This is also the source of the French word la glèbe, obviously a direct borrowing from Latin, meaning basically “a clump of earth”, especially one that sticks to the shoes, and by extension “rich agricultural soil”, used mostly in literary references to peasant activities. I was surprised that the TLFI also gives this word the same or very similar meaning as English ‘glebe’, since I had never encountered it in French.
John Cowan:
Could ‘child’ come from a word for ‘womb’? Something like that happens in Latin, where Horace (Epodes 17.50), Livy (1.34.3), and Justinian’s Digest (you weren’t going to check the references, were you?) use venter (‘belly’ and by extended usage ‘womb’) to mean ‘child’ (Horace) or ‘fetus’ (the rest). The Horace commentator from whom I get this (L. C. Watson, A Commentary on Horace’s ‘Epodes’, Oxford 2003) quotes D. Gourevich on Greek γαστήρ, which apparently shows a similar extension of meanings. The full reference for the latter is “Les noms latins de l’ estomac”, RPh^3 50 (1967), 85-110, if anyone wants to do further research (I don’t).
Could ‘child’ come from a word for ‘womb’?
Seemingly so. Our earliest records in English give the meaning ‘fetus’, which survives in the expression with child. Ælfric writes “Heo þa gelyfde his wordum & wearð mid cylde”, which word for word is “She then believed his words and became with child”, referring to the Annunciation. From ‘womb’ to ‘fetus’ is an easy transition, as your Latin examples show.
Indeed not, but thanks for the details! I hadn’t known the “lump” connection.
Eta confirmed! That must be why a scientific snail/slug is gastéropode in French, where based on, well, everything else including Gastropoda itself you wouldn’t expect the é.
Indeed, the Old Bad Translation of 20,000 Leagues speaks of “gasteropods”, and I remember being puzzled by it.
languagehat (April 30, 2013), on finding “nonsensed” as a transitive verb in a book from 1681: “now that I think to check the OED, of course they have an entry (updated December 2003), with the Lowell and Wason quotes, but not (obviously) the Freese or (surprisingly) the Baxter—I’ve antedated them by [188] years!”
It’s there now! No way of knowing exactly when they put it in.
There is now an Alia Allia on Facebook with no posts and no friends. Odd!
By pure luck, the US and England seem to have partitioned the rams-descended words into a relatively clean split, with the Americans taking (almost exclusively) ramp(s) for the New World species, while the English, or at least the southern English, have dropped ramp(s) while keeping the other words for the Old World species. Or so I conclude from the OED’s region labels (all revised 2008, except buckrams):
Allium tricoccum (New World)
Allium ursinum (Old World)
I’m assuming that if the OED has a region label it’s probably right, though if it doesn’t have a region label, sometimes it really should, e.g. spring onion. So the Irish, Scottish, and northern English do still have ramp(s), but the overlap is unlikely to cause any cooking problems, since you can’t get them both in the same place and anyway it sounds like they’d easily substitute for each other (sadly I’ve never had the good fortune to eat either of them).
As for the singular, I imagine asking for one ‘ramp’ at a farmers’ market
in my heavily trinibagonian neighborhood, you learn pretty quickly that the singular of “doubles” is “a doubles” (but the doubling in the name is pretty transparently about the pair of small flatbreads used to make them, not about the fact that you’ll want to eat more than one).
“the original Indo-European word for onion (*krem-)” may be a little bit of an overstatement, depending on who you ask. Mallory/Adams have *kremhₓus ‘(wild) garlic’ labeled as “West-Central” since there are no Indo-Iranian reflexes. But AHD has apparently rejected this PIE root. It was in the first edition (1969), with the root kerem- (from Pokorny) referenced in the entry for ramson, but from the 1992 edition onward that root has been deleted from the Indo-European Roots Appendix, and the etymology of ramson now takes it back only to Old English. I wonder what’s up with that.
Something like that happens in Latin, where Horace (Epodes 17.50), Livy (1.34.3), and Justinian’s Digest (you weren’t going to check the references, were you?) use venter (‘belly’ and by extended usage ‘womb’) to mean ‘child’ (Horace) or ‘fetus’ (the rest).
I presume it’s the Digest that is the source of partus sequitur ventrem, the Roman (and American) legal rule that a slave mother gives birth to slave children independent of the status of the father.
Dmitry Pruss asked in 2013: “is carrot also cognate to PIR *krem??” Surprisingly, nobody answered at the time. The answer is no, it’s from a different PIE root. (“Root”, heh.) See discussion of Carrot in 2020.
But AHD has apparently rejected this PIE root.
A root krem- ‘onion, wild garlic’ is still included in Calvert Watkins’ stand-alone book, The American Heritage Dictionary of Indo-European Roots (AHDIER). See page 44 in second edition (2000) of AHDIER on Google Books here (I hope), and the corresponding entry on p. 45 in the third edition (2011).
I remember Calvert Watkins lovingly mentioning the areal(?) group including Greek κρόμμυον, Middle Irish crim, Old English pl. hramsan, Old Lithuanian kermušis, Russian черемша, etc., when the topic of the Indo-European words for plants of the genus Allium came up in Hittite class, in relation to the Hittite word suppiwašḫar (‘onion’? some species of Allium, at any rate). This was in 1994 or 1995. He even bothered to mention the speculative etymology associating the city name Cremona to this family of words for ‘ramsons, onion’, although he pointed out a hitch in the gitalong: the securely attested forms of this family all appear to be recent reworkings of a stem *kremu(s)-, and that the absence of this *-u(s)- element is unexpected in a language with so archaic a cast as Cisalpine Gaulish.
(Oops, forgot to reply at the time.) Thanks, Xerîb. It seems unlikely that they rejected the root as invalid; could it be a random cut simply for space reasons?
It’s not in Wiktionary in any case.
It’s been requested; *kermus- or *kremus- or *ḱrem- is a red link in Wiktionary’s etymologies of several of the Germanic, Celtic, Ancient Greek, and Slavic cognates listed by OED. Count one for the list of PIE roots that are endorsed by the OED but not in Wiktionary.
With that distribution, it may not be a PIE root, but a substrate word or a regional extension of another root, maybe k’er-?