I recently saw a reference to a “Maghrebi doughnut” called sfenj: “a light, spongy ring of dough fried in oil. Sfenj is eaten plain, sprinkled with sugar, or soaked in honey.” Sounds delicious! That Wikipedia article is full of interesting stuff (e.g., Moroccan Arabic idioms like “Give someone a sfenj and he’ll say it’s ugly” and “As if hitting a dog with a sfenj”), but of course I wanted to know about the etymology, for which I had to go to Wiktionary, which gave the overall meaning of the Arabic word as ‘sponge’ and said “From Ancient Greek σπόγγος (spóngos).” The best part is the list of Descendants:
• Maltese: sfinġa
• Libyan Arabic: سفنز (sfinz)
• Moroccan Arabic: سفنج (sfanj)
• → Persian: اِسفَنج (esfanj)
• → Hindustani:
Hindi: इस्फ़ंज (isfañj)
Urdu: اسفنج (isfanj)
• → Sicilian: sfincia
• → English: sfincia
• → Italian: sfincia, sfincione
Note that none of them is spelled or transliterated “sfenj”; ah, the joys of rendering Arabic into the Latin alphabet! (That “English” sfincia is pretty marginal; it seems to be used only of the Sicilian dessert, and the Wiktionary link goes to the Sicilian entry, just like the one above it. The OED knows nothing of such a word.)
Sfenjes sate sphinges.
It’s a long shot, but I wonder if somehow words for sphinx and their relatives get implicated. Italian (or Sicilian) sfincia is superficially at least quite close to sfinge (“sphinx”). Heh. Both involve squeezing!
O, is Y working with the same line of thought? “Sfenjes sate sphynges.” I don’t get the meaning.
I misspelled sphinges. I had an image of passers-by offering the sphinx a sfenj, in lieu of answering its riddle.
Why is the Sphynx cat spelled with a y?
Israeli Hebrew סופגנייה (sufganiya) ‘jelly donut’ was inspired by Mishnaic Hebrew סופגנין
(sufganin) (some details here: https://www.sefaria.org/search?q=%D7%A1%D7%95%D7%A4%D7%92%D7%A0%D7%99%D7%99%D7%94&tab=text&tvar=1&tsort=chronological&svar=1&ssort=relevance), which likely belongs to the family noted in the post, and, since the word occurs in the Mishna, that Hebrew word may well be the currently oldest-known member of the family.
In my part of Algeria we call them xfaf (“light ones”) instead – good comfort food, best eaten hot with no toppings. Moroccan and Algerian have basically lost contrasts between short vowels, so a phonemic transcription of the usual word would be /sfəndʒ/ or /sfənʒ/ (depending on area).
In Moroccan Tamazight (SE Morocco) the /s/ becomes /š/ : /šfenž/
/šfenž/
This is completely regular, by the way: sibilant harmony is productive not only in Tamazight but in some neighbouring Arabic varieties as well.
I’m guessing simple ignorance. For centuries, it seems to have been the majority opinion that good old Latin silva was sylva…
@M: My understanding was that name of the sufganiyot we eat every year was an Israeli Hebrew neologism, merely inspired by the Greek loanword in the Mishna. Apparently, before the dish was brought back to the Levant with the first waves of Zionists, the dough was normally fried in schmaltz or tallow, which sounds… unappealing.
Sufganiya, the diminutive of Mishnaic sufgan, was invented by David Yellin, in המוכיח מוקפילד, his 1896 translation of The Vicar of Wakefield (here): “If the cakes at tea eat short and crisp, they were made by Olivia” (Like all names, he translated “Olivia” into a biblical equivalent, in this case אָהֳלִיבָה ʾŏhºlîḇâ, Aholibah of Ezekiel 23:4.) Those were pancakes, I think, not donuts.
The word for ‘sponge’ is a manifestation of a Wanderwort of unknown origin, related to L. fungus. Martirosyan discusses it at great length.
The word for ‘sponge’ is a manifestation of a Wanderwort of unknown origin, related to L. fungus.
I’ll be damned! It was worth making the post to learn that.
the dough was normally fried in schmaltz or tallow, which sounds… unappealing.
I had custard tarts in Guangdong back in the early 90s that had been baked in chicken fat. Unappealing is right.
i’d go for fried dough cooked in shmalts! not sugared, or filled with jelly, though.
When I see an obscure-looking 5-letter English word I often rush to the Wordle allowed guess list to see if it’s there – I try to make a collection of real 5-letter English words that aren’t on that list. (Some, such as “abjad”, used to not be in Wordle but had since been added.)
I was only mildly surprised to not see “sfenj” in the Wordle list; I was significantly more surprised to see nothing starting sf- at all. The list (conveniently alphabetized) goes straight from “seyen” to “shack”.
(In fact it appears that the consecutive letters “sf” do not occur anywhere in the list. I’m tempted to check which other letter pairs this is the case for.)
I’m tempted to check which other letter pairs this is the case for.
Apparently enough of the consonant-consonant pairs that the full list would probably be very long. (A few do show up, sometimes in unexpected ways; the only -bm- word in the database is “abmho”.)
My collection actually contains over 20 words at the moment; it used to be more but (as I said) a few got added in the August 2022 update. (I have confirmed that neither of the 2023 updates added any of the remaining words.) Perhaps the least obscure are “eroge”, “ecchi”, “unfun”, and “yoink”, though the audience here would probably also be familiar with “hanzi”.
Admittedly after some point it becomes unclear what actually qualifies as an English word. (Does “knyaz” count? Google does find instances of apparent use within English-language text.) My list includes oggtt but in a special section for words so excessively rare that it’s unclear whether they should have been included in Wordle.
If the cakes at tea eat short and crisp
Wait, you can form a middle voice of “eat”? Does this sentence read as weirdly to everyone else as to me?
Wait, you can form a middle voice of “eat”?
Mentioned only a week ago.
And in the subjunctive, too. I think it’s the combination.
Lameen: Not a native speaker, but this is indeed very weird, and in my anglophone neck of the woods (northeastern Ontario and Western Canada, mainly) I have certainly never heard anything remotely like this. Indeed I suspect many L1 anglophones in these parts (and others too…which ones, I wonder?) would find the sentence flatly ungrammatical.
I suspect many L1 anglophones in these parts (and others too…which ones, I wonder?) would find the sentence flatly ungrammatical.
I doubt it. It’s certainly not common, but I’ve heard it and perhaps used it on occasion — it fits into a productive pattern (cf. “This bed sleeps two”).
Short here means ‘crumbly, friable’, so I was likely wrong, and some kind of crumbly cake is discussed (I had been swayed by a mention of pancakes elsewhere.) The OED has a separate entry (under short, adj., IV.20.b), “Of fruit, meat, etc. to eat short: to break up or crumble in the mouth.” The two quotations given which are in the middle voice are from Evelyn’s 1699 Acetaria: a Discourse of Sallets, of radishes: “The bigger Roots (so much desir’d) should be such as being transparent, eat short and quick, without stringiness, and not too biting”; and from London and Wise’s 1706 The Retir’d Gard’ner, discussing a variety of pear: “The Angelique of Bordeaux is something like the Winter Bon-chrétien, only ’tis broader, and not so large; its Pulp eats short, and its Juice is sugar’d as much as the Bon-chrétien‘s: It may be kept a long time.”
Fat makes baked goods crumbly, hence shortening.
I am not sure what quick means here.
Yellin was probably puzzled by “short and quick” and even by “tea”. His translation, retranslated, is something like, “If the sufganiyot brought to the table were made nice, Oholiva’s hands had made them.” (I’m not doing justice to the pretty and convoluted biblical-style syntax.)
The Retir’d Gard’ner is a translation of Gentil François’ Le jardinier solitaire. The phrase translated as ‘its pulp eats short’ is “elle est cassante.”
The W-ary entry for fungus summarizes Martirosyan. It mentions Georgian, Tsezic, and Mordvinic lookalikes (after Gamk’relidze & Ivanov). I’d like to know what Christopher Culver and J Pystynen think of the Mordvinic ones.
Wait, you can form a middle voice of “eat”?
You can do this in Kusaal:
Dakaŋa la nuud sʋ’ʋŋa.
“That beer drinks well.”
You can actually do this with almost any agentive ambitransitive verb in Kusaal, but in the imperfective aspect (as here) it’s only grammatical in the “habitual” sense: you can’t say “the beer is drinking well” in a progressive sense.
This seems to be a common thing in West African languages, though the details seem to vary a lot. (Most grammars don’t go into a lot of detail about it.)
For the eat of sentence under discussion, it may just be a past tense. I don’t know the particulars of Goldsmith’s English (born in Co. Roscommon or Co. Longford, educated in Dublin and Edinburgh), but eat was a possible past tense in the 18th century (see the OED here). Cf. this passage from Dafoe here, or from Boswell’s journal (21 December 1762):
Later editions often amend eat to ate in passage from Goldsmith.
Short comment because I am napping after eating.
(Interesting the way The Vicar of Wakefield was brought to press.)
emend and Defoe. Voice capture on my phone…
In my experience, “voice capturing” works as well as sending an dog to fetch a newspaper without chewing on it. All the phonemic variability that folks here discuss with enthusiasm, will throw voice capture for a loop. It fails when I’ve got a cold or am chewing gum.
Or when I speak German, because – depending on whether I’ve been reading a lot of English recently – my accent tends to wobble around what the software expects. It does its best, which is reliably more than the circumstances warrant.
The W-ary entry for fungus
I can’t judge most of it, but the references to “Carl Meißner; Henry William Auden (1894) Latin Phrase-Book” must have been added by someone who has no understanding of Latin or by a robot; the quotes all relate to the verb fungor, not to fungus.
This wine drinks well.
Middle voice is very often invoked here. I have long argued against that (including at the Log when I still visited there; and here too, I seem to recall). The sense is clearly passive, not middle. Many verbs in English do have uses that are well classified as middle voice, but I think drink and eat are not among them.
Turn serves well to illustrate an unconventional but rational way to apply the terms active, middle, and passive (along with transitive and intransitive) for English:
Active voice
1. [transitive] Clive turned the door handle.
2. [transitive, reflexive] Clive turned himself around to face the window.
3. [intransitive] The rusty old handle could only be loosened by our combined efforts: I hammered while Clive turned, and eventually it was freed.
Middle voice
4. We watched aghast, as the handle turned all by itself.
5. Earth turns through one revolution every 24 hours.
6. Clive turned around to face the window.
Passive voice (or simple passive voice)
7. We had to hammer the rusty old handle, which turned only with the greatest difficulty.
With other verbs, for clarity:
8. Such weak enamel scratches when you use too hard a brush.
9. The floor cleans more easily with ammonia.
10. This wine drinks best with brie or camembert.
Periphrastic or composite passive voice
11. The rusty old handle could be turned only with the greatest difficulty.
12. Such weak enamel is scratched when you use too hard a brush.
13. The floor is cleaned more easily with ammonia.
14. This wine is best drunk with brie or camembert.
The existence of a well-established English periphrastic or composite passive voice does not justify avoiding the application of passive in analysing other verb constructions – or seizing on middle as an inept alternative because passive is “already taken”. If the sense is middle, call it as it is; if the sense is passive, call that as it is.
Xerîb: I think you are right. The bit with the cakes is followed by “…if the gooseberry wine was well knit, the gooseberries were of her gathering…” So both are in the past tense.
must have been added by someone who has no understanding of Latin or by a robot
Any text-producer ignorant of Latin might add the “ambiguous” qualifications there in order to conceal how disastrous that ignorance is, by affecting a higher sensitivity to meaning:
(ambiguous) to perform the last rites for a person: supremo officio in aliquem fungi
As you suggest, both people and robots can do plausibility. That’s why the uproar about AI is so silly. AI “assistants” demonstrate that intelligence is a matter of assimilation and imitation. As is the case with people. Once again, people are shown to be machines. As if we didn’t know that already.
One obvious, reasonable way to respond to this is by upgrading the notion of “machine”, so that people won’t believe they must feel inferior or superior. This is a bot-eat-bot world. Only confident bots survive.
“Intelligent machine” is a pleonasm. Some (“trivial machines” in von Foerster’s terminology) are more predictable than others. They all produce plausibility – except for those that don’t. These spend their life as decommissioned lawnmowers or asylum inmates.
“If the cakes at tea eat short and crisp” is itself in the OED under eat, sense I.5. “intransitive with passive force (chiefly with adj. or adv.): To have a certain consistence of flavour when eaten.” (Unrevised from 1891; “consistence of flavour” in OED2 appears to be a typo for “consistence or flavour” in OED1.)
See also It eats salty at Language Log (by Ben Zimmer), with several links to valuable background reading. And yes, there were objections to that sentence (said on “Top Chef”) because the complement was an adjective. Some commenters felt it was cooking-show jargon and better analyzed as a verb of perception like “looks” or “sounds”.
As further evidence that Xerîb was right that eat is a past tense in Goldsmith’s sentence, elsewhere in The Vicar of Wakefield can be found:
and
For those who don’t know, that past tense is pronounced “et.”
I was just wondering how Goldsmith pronounced the past tense eat. There are three possibilities: eat as in seat, eat as in great or eat as in death. The last is still a variant in Dublin
Has the past tense “eat” historically been pronounced in all those ways, or are the first two used by people unfamiliar with the form and assimilating it to what they know?
Has the past tense “eat” historically been pronounced in all those ways
I am away from my library at the moment, but should think the past tense eat regularly continues Old English strong class V past indicative ǣt, 3 pl. ǣton, a Germanic inheritance (Gothic (fr)et, etun; Old Norse át, átu; German aß, aßen), with regular West Germanic *ā > Old English ǣ > ME /ɛː/ > Early Modern English /eː/. I wonder if Modern ate represents an Old English or Middle English analogical past tense formation, with short OE *æ, ME a, in the root. See the note at the OED entry I linked to: “but a form æt, with short vowel, must also have existed, as is proved by the ME. form at, mod. ate”. If anyone knows more please elaborate, as I cannot research adequately that matter at the moment.
Most strong verbs remaining in English that have /iː/ in the present have collapsed onto the weave, wove; steal, stole; freeze, froze; etc. paradigm for the past tense, regardless of what strong verb class they originally came from. The remaining ones don’t seem to follow any other pattern in Modern English: beat, beat; see, saw; and eat, ate, or et in the past, although they do all revert to /iː/ in the past participle.
That turns out to be a past tense, too. A fuller quote:
That’s how I was taught to pronounce ate; I only learned later that the pronunciation you’d expect from that spelling, i.e. 8, also exists.
eat as in death
I take that to be “eat as in debt”; that is, eat would end in an alveolar stop, not a dental stop.
I take it as “eat with ea as in death“, i.e. ea as a spelling for the DRESS vowel.
@dm, jc
Yes, dm is right, I am bad at using IPA and lazy when I want to dash a quick comment off. Also as Hat says, there is no reason to think the vowel in eat in the quoted passage was not the DRESS vowel (or the alternative vowel indicated by the spelling “ate”). I was curious if Goldsmith used the “Roscommon long a”, where ate (or eight) could sound like hay yet without the initial h.
Speaking of literary references: right across the street from Dostoevsky Museum in St. Petersburg, there’s a small bakery with a variety of basic pastries, and hot pyshki (which are closer to sfenj than to other types of doughnuts). Frying continues for the whole day because they are sold for 15 cents a piece (certainly the cheapest option in the whole city), and half a bucket of powdered sugar is generously dropped on each serving. It’s usually quite crowded. Unlike other retro cafes, it’s not famous among tourists; and there’s exactly one video on YouTube made two years ago by some delivery guy visiting it to have a snack.
Lameen : “In my part of Algeria we call them xfaf (“light ones”) instead – good comfort food, best eaten hot with no toppings. Moroccan and Algerian have basically lost contrasts between short vowels, so a phonemic transcription of the usual word would be /sfəndʒ/ or /sfənʒ/ (depending on area).”
In my part of Bulgaria we call them мекици — “soft ones”. Eaten with powdered sugar.
Vasya: so you also eat them with powdered sugar.
@V: Only vaguely related, but your comment reminded me that I once wondered how the aliens (“hard ones” and “soft ones,” as well as the three genders if the latter) were rendered in translations of The Gods Themselves.
Brett : “a hard one” was translated as “Твърд”, “a soft one” as “Мек”, always capitalized. I’ll check about the genders.
Part of a wider well-known Uralic cluster (the Nganasan verb ‘to be drunk’ only listed in the comments there is by now recognized as direct inheritance from Common Uralic too) and compared with fungus & other IE parallels since the 19th century, too. Holopainen 2019 remains with the view it’s a cluster of Indo-Iranian loans, from a source closest related to Sanskrit bhaṅgá- ‘hemp’, regardless of if Middle Persian bang ‘henbane’ is or isn’t cognate.
Recent IE sources that I’ve seen or seen referenced seem to all reject the also old comparison of the IIr. words with fungus (or Greek σπόγγος) though. Maybe there’s no good enough cultural basis to consider here too the semantic shift ‘toxic / narcotic plant’ > ‘mushroom’? Or just too few explicit semantic breadcrumbs — at least Mari, Mansi and Khanty attest this specifically as ‘fly agaric’, a mushroom well known as a shamanic narcotic and already good reason to think these don’t simply come from a neutral general PU word for ‘mushroom’.
It’s interesting that in Mari and Mordvin the word means both ‘mushroom’ and ‘sponge’. It’s an unusual colexification (presumably because of the texture of both), and supports connecting the word with the Mediterranean. The distribution (SE of the Urals) looks more areal than genetic, suggesting later diffusion rather than a single early Proto-something loanword, no?
Maybe the exotic word for ‘mushroom’ was borrowed in some places along with the narrow meaning ‘medicinal mushroom, fly agaric’? Is fly agaric known to have been used around the time and location of PU?
Y: an unusual colexification
Norw., Da. sopp “mushroom”, svamp “sponge’, Sw. svamp “mushroom” < ON svõppr ~ sõppr “mushroom, sponge”; Ger. Schwamm “sponge”, Du. zwam “mushroom” PGmc ~*swambaz “mushroom, sponge”.
Similar enough to fungus, σπόγγος et al to suggest a wanderwort or substrate. Anyway, we have a colexification covering half of Eurasia.
NB that Schwamm(erl) vel sim. is the general word for “mushroom” in many (mostly Southern) German dialects, and in the Standard, besides meaning “sponge”, can also refer to various types of mushrooms growing on wood.
Yes, there are more examples all over Germanic. I just picked enough easily found forms to make the point.
Eng. swamp belongs here too. That’s colexification for yah!
I was going to say that there’s no trace of a sense “intoxication” in Gmc., but wait a second…
Swamp made me think of a basic meaning “soak(ing)”, which made me think of the verb swim and its somewhat unruly set of cognates, which in turn made me think of Norw. svime “dizziness, unconsciousness” and cognates. Not as exciting as a colexification with svamp and svamp and swamp, but it might perhaps be indicative of something at the root level. Maybe it’s as simple as “be soaked”.
Oh, and it’s been 2024 for an hour here. Happy new year, whenever it comes!
Russian губка “sponge” is apparently a diminutive of a now-dialectal word meaning “mushroom”. As it happens, the non-diminutive form survives in the meaning “lip” – now that‘s colexification!
(…Not sure if it beats my favorite pair “environment, Wednesday”, but it’s probably close.)
I agree on the “mushroom/sponge” side we might be talking about a colexification all across Europe-to-the-Urals at least; “half of Eurasia” might be overreach unless there’s something further east and/or south, but it’s definitely a huge area.
Some mushrooms (especially wood-growing ones) do look remarkably like sponges, though I’m not sure if any had the right kind of texture to be used in bathing.
(TIL that “loofah” is an actual plant, closely related to pumpkins, whose overripe fruits form so many fibers that they can be used as bathing sponges.)
Lion’s mane mushrooms (Hericium erinaceus) are pretty sponge-like.
Take a look at a map again: Mordovia and Mari El are nowhere near the Urals, instead respectively a bit south and east of Nizhni Novgorod. Like all IIr. loans it’s still likely yes at least narrowly post-Proto-Uralic, i.e. at least a few parallel loans; some explicit reasons for this too in minor details of vowel development I believe (e.g. Samoyedic *pëŋkå- should have been instead either *pïŋkå- or *pëŋkə- if actually inherited from PU).
There’s no known candidate for any old Uralic general ‘mushroom’ term it could have ousted though. Finnic has *seeni which comes from PU ‘polypore’ judging by its cognates (some go from there towards ‘tinder’), & a more likely late arealism are Samic *kōmpër and Permic *gob, coming from somewhere in the direction of Slavic and/or Bulghar.
I wouldn’t know if there is, or even feasibly could be, any archeological evidence for Amanita usage millennia ago but it’s ethnographically recorded pretty widely, also on the IE side I believe (I recall even some speculation of that as one of the candidates for the Vedic soma).
…sometimes I also wonder if the #bonga and #gomba etyma for ‘mushroom’ have themselves a deeper connection via metathesis, but that’s getting into the wild speculation zone.
J.Pystynen: There’s no known candidate for any old Uralic general ‘mushroom’ term it could have ousted though. Finnic has *seeni which comes from PU ‘polypore’ judging by its cognates (some go from there towards ‘tinder’)
Agarikon, a conk on pine trees, has been found in Shaman graves. It has known medicinal uses, but (in spite of the name) none resembling the hallucinogenic properties of the agarics.
I wouldn’t know if there is, or even feasibly could be, any archeological evidence for Amanita usage millennia ago
It should conceivably be recoverable, at least as DNA, from shaman graves, and especially from inside purses or boxes with (one must presume) special content. I can’t find anything in the searchable literature, though. The technology probably wasn’t available when the currently known shaman burials were excavated.
Romanian burete too is both ‘mushroom, bolete’ and ‘sponge, kitchen sponge’.
Stress on the second syllable: bu-RE-te; dexonline says the singular has been refashioned after the plural bureți and would originally have been *buret(u) from Latin boletus.
January First-of-May : “Russian губка “sponge” is apparently a diminutive of a now-dialectal word meaning “mushroom”. As it happens, the non-diminutive form survives in the meaning “lip” – now that‘s colexification!”
That’s the normal word for both mushroom and sponge in Bulgarian — гъба, diminuteve гъбка.
J Pystynen: thanks for correcting me. But now that I think of it, a more basic problem: would (±Post-) Proto-Uralic speakers have been familiar with sponges? For that matter, for how long have e.g. recent Mordvins or Mansy been familiar with them? Have sponges ever been a significant long-distance trade item anywhere? Could it be that the colexification in these Uralic languages (and maybe in Tsezic too) is due to recent Russian contact?
That is all aside from the issue of the word for ‘mushroom’ alone.
Can confirm.
…and in Ukrainian it means “mushroom” and “lip”. I happen to know that because it was mentioned in the long response to A. Vovin’s long attack on the Etymological Dictionary of the Altaic languages, where Vovin presented this polysemy as obviously absurd.
“Conk” could be probabably be a good starting point for a polysemous word meaning “mushroom; porous material”. The sense “mushroom” could be extended to “intoxication”, and “porous material” be concretized to “sponge” and “swamp”. For “lip” it seems more likely to me as the original sense, and that conks were named for their lip-like shape.
But with more than one language involved, things turn complicated. An intuitive semantic extension in one language may be followed by the opposite extension in the other.
The kind of mushrooms that grow on trees in the wild do look like lips. But you can cultivate lots of kinds of mushrooms on cut down trees.
I had previously noticed that the words for “sponge” and “mushroom” were the same in some languages. Wondering about it, I guessed that might make more sense if one’s prototypical mushroom was a morel.
Well, the bathroom sponges of old were actual Porifera, an animal phylum.
In Danish, svamp covers both Porifera, modern imitations in other materials, and the kingdom of Fungi (especially mushrooms, though dry rot and some molds are covered as well). We don’t have any terms for the tinder fungus that involve lips, though.
Oh, tinder fungus in German: Zunderschwamm. (Or just Zunder.)