One of the words in the Scripps National Spelling Bee 2016 (Guardian liveblog) was chremslach, the plural of chremsel; your curiosity about what the word represents can be satisfied by this lively Haaretz column by Liz Steinberg (thanks, Paul!): “Admittedly not the most common of Jewish foods, chremslach are flat, fried fritters made by some Ashkenazi Jews for Passover or Hanukkah.” But what if you’re curious about where the word is from? It’s not in the OED, NOAD, or M-W Collegiate; Webster’s Third New International has it (which is why it was eligible for the spelling bee), but the etymology given there is just “Yiddish chremzel.” Well, yes, it practically screams “Yiddish,” but where is the Yiddish word from? So I did a little digging in Google Books and found this on p. 393 of Alexander Beider‘s Origins of Yiddish Dialects: “Influence of French is also quite likely in AlsY frimzl and SwY fremzl ‘noodle’/EY khremzl ‘Passover pancake.'” The relevant footnote reads:

[…] This word is related in some way to Italian vermicelli though the immediate etymon for the WY word is uncertain. Starting with Kosover (1958:63), several authors wrote about the link between the WY frimzl and the Old French word that in turn was loaned from Italian. However, no linguistic argument corroborating this hypothesis was ever suggested, while the proposed French etymon is actually anachronistic. In French, the earliest reference to a word derived from vermicelli dates from the sixteenth century only (Wexler 1992:54). Looking into the early Jewish references collected by Kosover (1958:61–5), one can observe the existence of two series: (1) with the initial gimel […], the oldest date from the twelfth century; (2) with the initial vav or double-vav […], mainly present in sources from western Germany of the twelfth to fifteenth centuries, but also known in Normandy in the thirteenth century. The first series is clearly of French origin. It was in certain Romance dialects in the territory of France that Germanic (Frankish) initial /w/ gave rise to /gw/ that later turned into /g/. The last process ended in Old French during the twelfth century. […] The second series (to which WY frimzl is related) may also be of French origin. Indeed, Frankish /w/ remained unchanged in northern (Wallony, Picardy) and eastern (Lorraine) dialects. […] Both series have two idiosyncrasies in common: (a) the letter zayen /z/ for what was “c” in Italian; (b) the introduction of a vowel between /r/ and /m/, most likely as a result of the metathesis between the first vowel and the second (liquid) consonant. Such characteristics could not appear independently. Either both series had the same ancestor that already possessed these features, or one of these series influenced another. The feature (a) is typical for France only. It is regularly found in French and Occitan: compare modern French plaisir “pleasure’ < Latin placere, oiseau ‘bird’ < aucellus (Bourciez 1921:153–4). For the feature (b), a close parallel can be found in French fromage ‘cheese’ whose initial sounds underwent in Old French the change from /furm-/ to /frum-/ (Pope 1934:178).

So there you have it: chremslach is related in some way to vermicelli, though the details are frustratingly unclear. Me, I’ll stick to latkes.


  1. I’ve never thought about the etymology, but I have always considered “chremsel” to be a very ordinary Yiddish word. I am actually surprised that it doesn’t appear in more English dictionaries.

  2. An ordinary Yiddish word, but not an ordinary English word.

  3. I was looking at the Grauniad’s live coverage. Most of these final-round words I have no clue about, which surprises me. But what the hell do they mean, “silent e in Mischsprache“? What silent e?

  4. That surprised me too; I assumed the dictionary gave some ferschlugginer two-syllable pronunciation like MISH-shprahk, but no, looking it up I find the only pronunciation given is three syllables, with a final schwa for the -e. So either the Graun misreported or the presenter mispronounced it (grounds for a lawsuit!).

  5. A bit off topic, but here is Al Beider’s enjoyable interview (in Russian)

  6. I was surprised the interviewer used ты with him, but maybe they’re friends?

  7. maybe they’re friends?

    Looks like classmates at Fiztekh

  8. marie-lucie says:

    chremslach is related in some way to vermicelli

    I hope Etienne can help us here, because I find this relationship quite unconvincing without having all the necessary elements to evaluate it.

    Vermicelli is the Italian word, a diminutive of the word meaning “worm”, from Latin ver, vermis which refers to the worm-like shape of the pasta strands. It and its French adaptation are only attested in the late Midle Ages or Renaissance period. The Yiddish word (occurring in two presumably dialectal forms) does not refer to any type of worm-shaped pasta but to a flat, circular shape, a sort of crêpe. The text refers to “Old French” without precising the actual period, and brings up two phonological rules which belong to a fairly old period (eg Latin w- becoming gw- under Frankish influence, long before the emergence of “Old French”, let alone the influence of Italian).

    Since at least the f-initial Yiddish word appears to have some relationship with a word attested in Normandy, the Normand word might reflect a borrowing from Old Norse, dating from the early occupation of Normandy by the original “Northmen” (there are a number of other such borrowings in the Normand dialect, some of them now in more general French use).

    It seems to me then (barring other, more precise, better attested data) that the Romance and the Yiddish (Germanic) words are only coincidentally related.

    As for the “metathesis” for the word meaning ‘cheese’, the root of the Latin word was form- in formaticum, a (variety of) cheese made in a ‘form’ or mold. This root survives in Italian formaggio and the Auvergnat la fourme (a local cheese) but at some point French changed formage into fromage. This sort of “metathesis” is common, with the liquids r and l starting as syllabic and later acquiring a separate vocalic component, either before or after the liquid, according to the particular “phonotactic” rules of a language. (Similarly, English bird was once brid).

    Finally, I am not competent to comment on the f/ch alternation in the Yiddish words.

  9. “It’s not in the OED, NOAD, or M-W Collegiate”

    but it’s in the Scripps National Spelling Bee 2016. He who makes the rules, drives the debate. A very obscure Yiddish word in a US “national” spelling bee. Oy Vey

  10. Trond Engen says:

    I can’t think of any potential ON cognate (the closest I can think of is hleifr “loaf, round cake; round cheese”, which just adds more confusion), but I’d like to know the Normand word.

  11. marie-lucie says:

    Trond, the text refers to “Normandy in the thirteenth century”!

    Googling “dialecte normand” yields a number of studies and dictionaries of local dialects, most of them from the 19th century when the dialects were general in the rural population. On a more general site I tried “galette” (a type of round flat bread) and “pain”, without success. I don’t plan to go through all the references at this time! But a Norse origin for the word in question is likely.

    The Italian influence which brought people, their culture and their words to the court of France (Caterina dei Medici, known as Catherine de Médicis, brought hundreds of Italians with her when she came to France to marry the king) dates from the 1500’s.

  12. I believe that these early citations in Rashi, etc. were first collected in Mordecai Kosover’s יידישע מאכלים , which you can download here. See pp. 62-63.

  13. Trond Engen says:

    with the initial vav or double-vav […], mainly present in sources from western Germany of the twelfth to fifteenth centuries, but also known in Normandy in the thirteenth century.

    marie-lucie: Trond, the text refers to “Normandy in the thirteenth century”! […] I don’t plan to go through all the references at this time! But a Norse origin for the word in question is likely.

    On the face of it, yes, but since some of the western German attestations are even older, the word would have to be common Germanic, separately inherited in western German and whatever Scandinavian dialect left this word in Normand (or the Jewish community of Rouen had roots in Western Germany, or the other way around).

    Anyway, an ON origin can’t explain the variation ch-/v-, not even to mention f-. Initial ch- might be from ON h-, but the h- of hr- was retained in much the same area that lost the v- of vr-. Searching wider within Germanic, if ch- is High German for k-, then it might point to a pair **krimsel/**vrimsel. As it happens, the Da.-No. word vrimle “crawl, teem (with)” is explained as a borrowing from Du. wriemelen, a contamination of wiemelen with krimmelen with the same meaning. **chrimseln und wrimseln could be a jocular decription like “worms and germs”. But it’s a longshot. And I still haven’t touched f-.

    MMcM: you can download here. See pp. 62-63.

    Nice. This oligalphabet can’t read it, but someone will come along. But the Latin alphabet forms quoted inbetween do indeed seem to point to ‘vermicelli’. Or *wurmis.

  14. marie-lucie says:

    MMcM, Trond: without downloading, I can’t find the relevant forms, but WHY would a word for ’round flat bread’ be from one meaning “little worms”? Was the dough perhaps divided and rolled into balls, then into long pieces, then flattened (as for making strips across the top of a pie)? That seems like a lot of manipulation for an apparently basic food word.

  15. The Kosover book is quite a treasure. Even without knowing Yiddish, I can figure out enough to feel my mouth watering and my arteries clogging.

    Grimslich~gremslich appears in some modern-day Jewish cookbooks as another variation, a kind of a fried bread pudding.

  16. Trond Engen says:

    I can imagine a semantic path “little worms” -> “noodles” -> “dumplings” -> “loafs” -> “flat bread”. But it’s not very attractive. I’d rather derive it from an ON word for round dough. Some of my best friends are ON words for round dough. But I can’t find anything that fits.

  17. marie-lucie says:

    Yes, one can imagine such a path, but considering the real life counterparts of those words, flat bread types usually start with a ball of dough that needs to be flattened to be useful (for instance, flattened over a heated stone griddle to cook it, before the oven was invented, allowing three-dimensional loaves).

  18. Kosover mentions as a parallel the names of the city of Worms in early Jewish sources, vermaiza ~ garmaiza.
    In the medieval references this food is described as ‘bits of dough’. It’s not clear to me whether they were boiled, like spätzle, or fried, like donuts.

  19. Etienne says:

    Hmm. This looks messy. I sense a decent monograph could be written on the history of this word.

    I won’t go over every argument, but would like to offer a thought for hatters to consider: it concerns the /f/ ~/x/ alternation: could a Western Yiddish form with initial */fr/ have entered Eastern Yiddish indirectly, i.e. via some Slavic language (Czech, perhaps?) which, lacking /f/, substituted /x/? I believe Russian does this with some Hellenisms which originally contained /f/.

  20. Yes, I was thinking the same thing; there is occasional substitution of /x/ for /f/ in Russian, e.g. хранец /xranets/ ‘syphilis, (lit.) the French disease’ from Dutch frans.

  21. Trond Engen says:

    So those words I could read were all different forms of the name of the city of Worms? Oh, well.

    I’ve been pondering if ch- and v-, even f-, might be reconsiled as ON hv-. We’d need metathesis of a form like **hvermsill, which in turn would have to be formed from hvarm “rim of the eyelid; puss in the eye”. Not very attractive either.

    If I handwave away all variation in the initial cluster as internal French and Yiddish developments, I can concentrate on wr-. Supposing metathesis *wrem- < *werm-, we might see a form vermsla “the act of warming, what is warmed up”, parallel to e.g. veizla “the act of allowing; econ. allowance; etc.”, geymsla “the act of taking care of something, hiding” and many more. A noun vermsl is known in the meaning “well or spring that doesn’t freeze in the winter”.

    (The relation between feminine singulars and neuter plurals is long and interwoven. A pair like kensl “recognition; accusation” and kensla f.s. “recognition; instruction” might indicate a vague aspectual nuance.)

  22. Trond Engen says:

    Vaw and double vaw! There’s not only verma “warm up” but ferma v. “load” and a feminine noun ferma “feeding up; food” used in poetry. ” A derived noun **fermsl(a) could be construed in the meaning “the act of filling (with food)”. (I believe more in vermsla but it might perhaps have contributed to the alternation.)

  23. Trond Engen says:

    On the other hand: The Norman conquest of Sicily in 1061 came after Norman mercenaries had been active in Italy for almost a century. Italian cooking may well have been brought to Normandy before anywhere else in France.

  24. occasional substitution of /x/ for /f/

    Paul Wexler also pointed this out in The Balkan Substratum of Yiddish.

  25. marie-lucie says:

    Trond: The Norman conquest of Sicily in 1061 came after Norman mercenaries had been active in Italy for almost a century. Italian cooking may well have been brought to Normandy before anywhere else in France.

    The Norman mercenaries came down at the request of local princes, they were not nobles led by the Duke of Normandy, and there is nothing that suggests that they went back to Normandy, at least not in enough numbers to influence their country of origin. Instead they remained in Southern Italy, Sicily, Malta and even North Africa, conquering more and more territory, which would not have been possible in Normandy or France.

    Meanwhile the new Duke of Normandy (Guillaume) had his eye on the throne of England, which he was able to secure for himself in 1066.

  26. Hat: Me, I’l stick to latkes.

    But do you grate your own potatoes?

  27. My wife does the grating, but I wash the grater.

  28. marie-lucie: I think the semantic path went like this. First, the term meant noodles, then pasta more broadly. Eventually, Eastern Yiddish developed the Slavic-derived lokshn for noodles, so fremslekh, now mutated to khremslekh, remained only as a term for a kind of non-noodle pasta. In the rabbinic sources as traced by Kosover, this seems clear (as far as I can tell; the Kosover material is admirably detailed and therefore hard to get a handle on, at least for me).

  29. marie-lucie says:

    Ben, thanks for the suggestion. I don’t know enough about the history of pasta-making in various countries.

    One thing occurs to me: the word sometimes mean “bits of dough”. Those “bits” could be the leftovers from laying a large piece of flattened dough at the bottom of a mold and cutting off the excess, for instance when making a pie.. Those bits of course are as edible as the rest of the dough and could be fried or otherwise treated for use as a snack or in soup. Noodle-making may have started in this way, from deliberately making long strips in imitation of the leftover pieces of dough.

Speak Your Mind