Corinna da Fonseca-Wollheim (great name) has a NY Times review of Martin Puchner’s The Language of Thieves: My Family’s Obsession with a Secret Code the Nazis Tried to Eliminate
When Martin Puchner was growing up in a rowhouse in Nuremberg in the early 1970s, strangers would show up at the door asking for food. His mother served them water and sandwiches, standing in the doorway while the visitors made conversation in a language the boy could not understand, even though the words were mostly German. Later, his uncle pointed out to him what it was that drew these strangers to the house: Carved into the foundation stone was the sign of a cross inside a circle. To those in the know, it signified that the house’s occupants would give you food.
Those in the know were all manner of vagrants: tinkers, knife grinders, peddlers, journeymen — people without a fixed abode. The pictograms they carved into fence posts or chalked on houses were called zinken, after the Latin signum, for sign. The language they spoke was Rotwelsch, a mix of Yiddish, Hebrew and repurposed German that had been used for centuries by members of the itinerant underground. Puchner’s father called them “people eternally on the road, escaping to nowhere.” […]
Both Puchner’s father and uncle were drawn to Rotwelsch and sprinkled words from it into their speech. As a boy, Puchner delighted in zesty phrases like “making a rabbit,” which meant making a quick escape. On hikes, his father taught him to spot zinken on roadsides and farmhouses. Though his parents were solidly middle class, Puchner writes, “I grew up feeling that I had a special connection to the road and the itinerant underground.” In his family, he felt, “Rotwelsch became our special possession, our secret.”
Puchner became a professor of English and comparative literature at Harvard and decided to investigate this icon of his childhood:
Rotwelsch developed in the High Middle Ages and spread across Central Europe in the wake of the Thirty Years’ War. “Welsch” meant “incomprehensible”; “rot” was derived from a word for “beggar.” Rotwelsch was thus the incomprehensible language of beggars.
Technically, Rotwelsch is not a language (because it doesn’t have its own grammar) but a sociolect — a system of communication that binds members of a community together and keeps outsiders in the dark. Among the many Rotwelsch terms for police are the German words for bull, lantern and moonlight. But a substantial amount of Rotwelsch is derived from Hebrew and Yiddish, such as gannef, for thief. This infusion probably reflects the large number of Jews forced into itinerant professions in the Early Modern period because of laws banning them from landownership and many trades. According to Puchner’s research, however, the great majority of Rotwelsch speakers were in fact not Jews. […]
What endures is his fascination with the resourcefulness and resilience of generations of travelers, like the ones who came to his childhood home in Nuremberg, drawn by a hidden zinken.
“Their words for police, for being arrested, their zinken about begging and stealing, the rich vocabulary of food, drink, sex and lice, all this spoke volumes about their lived experience,” he writes. “Rotwelsch was like a worn tool that bore the traces of its earlier use. By studying it closely, one could tell a lot about the bodies that had wielded it.”
I trust neither the review nor the book to contain actual facts, so if anyone knows more about the history of Rotwelsch, please share. (Thanks, Dmitry!)
In Swedish, “rotvälska” means incomprehensible speech. I didn’t know it was an actual language.
Rotwelsch is presumably what Rottweilers speak.
And here I thought it was the jargon of Marxist Welshmen.
As well, sure.
It’s one of several varieties of Continental thieves’ argot (such as Dutch Bargoens and Polish Grypsera). First things that come to mind are, are any of them related? And, what varieties did the Yiddish borrowings in these cants come from?
I’m pretty sure ‘welsch’ didn’t mean “incomprehensible” at the time. Maybe “foreign”, but quite likely “Romance” or “Italian”. That might even mean that the early forms of it had recognizably Romance or Italian elements.
That’s an odd definition of ‘sociolect’. It’s a secret register, a sector of the language that’s available only to the initiated. I’ve also seen the word ‘cryptolect’.
A register like that is inherently fluid. The lexicon must have been constantly changing as words and phrases got useless after leaking out into the wider language and as new groups joined the “itinerant professions”. There must have been different forms for different purposes, some meant to be transparent for outsiders, others truly secret. There must have been different groups in a hierarchy based on their command of (or defining power over) the deep code.
And somewhere on the outside, though the distinction could sometimes be blurry near the edges, were various forms of Romani.
There’s a world of mystery out there: Secret Languages of Afghanistan and Their Speakers.
my vague memory (doubtless from a source no more rigorous than puchner) is that Rotwelsch is particular to the german-speaking lands… so i’d assume the yiddish involved would be ‘western yiddish’, according to the traditional major divide. i’d hazard a guess that it might even be specifically “westernmost” (the dialects corresponding to ashkenaz proper – hamburg and west/southward – which manaster ramer and beider argue are a separate unit from both ‘eastern yiddish’ and the rest of ‘western yiddish’), since those are the most deeply integrated communities, where there might be more flow between jewish and non-jewish germans’ argots.
i wonder, though, whether there’s a relationship between Rotwelsch and yiddish ganovim-loshn, klezmer-loshn, and balegole-loshn – the argots of thieves, professional musicians, and wagoneers: the disreputable and mobile professions of jewish eastern europe.
(i like “argot” because these sociolects/cryptolects are tied to specific crafts/professions…)
what’s going to get me looking at the book, though, is the chance to see whether the zinken correspond to north american hobo signs (some of which are still in use).
Is this rooted in Romani?
@ David Eddyshaw
The Welsh were always known for talking a lot of rot.
“Ganeff” is of course conventionally spelled “gonif” in Yiddish-influenced varieties of AmEng.* See also this short clip about the idiom “America gonif” which seems to be sufficiently non-pejorative as to be non-compositional. https://www.yiddishbookcenter.org/collections/oral-histories/excerpts/woh-ex-0001017/meaning-america-gonif
*”Gonif” is one of many Yiddishisms which were unknown to me growing up but got added to my lexicon after I started practicing law for a living in NYC.
The Welsh were always known for talking a lot of rot.
Proud to maintain the ancient tradition of my people. It is our ancient way (reflected in our ancient language.)
The sign carved in the foundation stone is an odd detail. Must have been awkward for the family that moved in when the Puchners muved out. “Why do all these randoms keep showing up on our doorstep?”
Does anyone know whether the reviewer deployed “a hidden zinken” correctly? I might have guessed it was plural.
Hans den Besten 2006, Yiddish Jews in the Netherlands and Bargoens:
And more, from Philologos at the Forward.
Does anyone know whether the reviewer deployed “a hidden zinken” correctly? I might have guessed it was plural.
In this meaning it is a singular, der Zinken. It is co-cognate with die Zinke, die Zacke, der Zacken, die Zinne et al. These all mean different things nowadays.
Another meaning of der Zinken is “schnozzola”.
@ Trond: -welsch probably just meant ‘non-Germanic’ > ‘foreign’ or something like that, not necessarily Romance; it’s etymologically related to such words as Welsh, Walloon, Vlach, Gaul, etc.
Re the ‘welsch’ element in Kauderwelsch, DWDS has
—
welsch ‘romanisch, besonders italienisch’ …, das oft ganz allgemein eine unverständliche Sprache kennzeichnet (vgl. rotwelsch)
__
There is also a long and inconclusive discussion of the first element Kauder.
https://www.dwds.de/wb/Kauderwelsch
@ PlasticPaddy: I meant in the specific case of Rotwelsch; I’ve never read about any strong Romance element in it. It’s of course true that in German ‘welsch’ often referred to varieties of Romance.
@fred
Yes, the second clause from DWDS supports your point that welsch is used to denote any form of incomprehensible speech. Sorry for the confusion????
@Fred, Paddy: I know. But as with all things lexical, the definitions get circular. ‘-welsch’ meant “incomprehensible” in ‘rotwelsch’ because its attested use in ‘rotwelsch’ means “incomprehensible language”. What I meant was that at the time the word was coined, presumably around or before 1200 CE, ‘welsch’ would (also?) have had a clearly defined meaning of “the language of those from the southern parts of the empire”. It could presumably also have been used arbitrarily for “foreign and incomprehensible” in an informal coinage like ‘rotwelsch’, but why ‘welsch’ and not e.g. ‘windisch’ or ‘jiddisch’ ? I think there’s reason to suggest that the earliest mentioned forms of travellers’ cant contained a recognizable Romance element, facilitated by the relative ease of north-south travel within the Holy Roman Empire.
The very name “Welsh” shows that at least some Teutons just meant “not-Germanic” by it. I would think that the “Romance” sense is simply a consequence of the fact that the great majority of gibberish-speakers these Germanic types encountered in relatively peaceful contexts happened to speak what we call Romance languages. (Bear in mind too that although it’s obvious to us that the language of the Vlachs is Romance, this would presumably not have been at all obvious to your mediaeval German man-in-the-street; indeed, he would probably have been mystified by our category “Romance.” The Germans still hadn’t invented Comparative Linguistics at that point.)
I’ve a feeling we’ve discussed this before, and that someone suggested that “Welsch” etc originally meant specifically “of ex-Roman ethnicity”, with no direct linguistic implications; not sure if that isn’t just a way of finessing away the use of the term for the non-Romance-speaking Britons, though. But counterexamples is counterexamples …
Still, Slavs don’t seem to be Welsch at all. Dunno.Maybe they were too incomprehensible even to be Welsch. Or just too foreign.
Again, even if “Welsch” did imply Romanceness, Pig Latin is not Latin; so there’s no reason to deduce from the name itself that Rotwelsch was originally largely Romance. The name surely just reflects the “it’s all Greek/Chinese to me” trope.
Incidentally, I notice that Wiktionary claims that “it’s all Greek to me” in Danish is “it’s all Volapük to me.”
https://fr.wiktionary.org/wiki/%C3%AAtre_du_chinois
Is this True?
(I can well believe that in Esperanto the canonical gibberish-language is Volapük; catty!)
Last year at LH:
First comment by a certain Eddyshaw.
The Germanic map seems to have been: Welsh on (mostly) one side, Wends (mostly) on the other, and Finns up further.
Southern/western gibberish-speakers, eastern gibberish-speakers, and northern gibberish-speakers. Fair enough.
A bit reminiscent of
https://alphadesigner.com/mapping-stereotypes/world-according-to-ancient-china
There is the same meaning of Volapuk in Russian, albeit obsolete. Never bothered to learn what was its origin. The word seemed to be artificial, as made up as abrakadabra. In a sense it is, although now I realize that it was made for a different role!
https://rg.ru/2006/06/23/volapuk.html
“Proud to maintain the ancient tradition of my people. It is our ancient way (reflected in our ancient language.)”
Pray tell us why your language (modern Welsh, I assume, is what you refer to) is more ancient than, say, English, French, Swedish, etc., all of which acquired their modern forms gradually and piecemeal. Modern Welsh is not Middle Welsh, nor is it the language often called British or Brythonic spoken in Roman times, just as modern English is not Middle English or Old English or proto-Germanic. I think one may refer to a language as archaizing if it has changed more slowly than most others, like Icelandic, Greek or Lithuanian, but Welsh isn’t even an archaizing language, is it?
You may be missing some context here.
As you are not the first to miss heavily signalled irony, I’ll take pity on you and help:
http://languagehat.com/publishing-tzotzil/#comment-4026535
Det er det rene volapyk ~ “It’s pure nonsense”. (Note the nativized spelling, it’s not a proper noun any more). Not really about material in unknown languages, but more about stuff in known languages that makes no sense grammatically, semantically or logically. (Jabberwocky is not volapyk, though, it does make sense just not the conventional sort).
Barbaric language is kaudervælsk. Cp German, and pace Danish WP which claims the words are synonymous.
David Eddyshaw: about your comment-
…”although it’s obvious to us that the language of the Vlachs is Romance, this would presumably not have been at all obvious to your mediaeval German man-in-the-street; indeed, he would probably have been mystified by our category “Romance.” The Germans still hadn’t invented Comparative Linguistics at that point.”
-I respectfully disagree. In the Middle Ages “Romance” was definitely a real linguistic concept to non-Romance speakers in contact with Romance varieties, due to the fact that the “Romance languages” at the time were much closer to one another, in practical terms were mutually intelligible -far more then than today- and were part of a dialect continuum which precluded treating them as separate languages in the same way we do today. In fact it has been pointed out that a pervasive problem in Romance Philology/Historical linguistics has been and remains a ubiquitous tendency to project backwards in time the sharp boundaries separating the various national and/or standardized languages of the present (or, more tellingly, of the late nineteenth century, when Romance linguistics arose and when nationalism as an ideology was uncritically accepted by all too many). I suspect similar problems/biases are at work in Germanic or Slavic historical linguistics as well (and in other scholarly traditions too, I imagine).
For hatters who know French, this article on the topic may prove illuminating (although, annoyingly, the non-Latin symbols in the article did not get transmitted in this on-line version):
https://www.cairn.info/revue-langage-et-societe-2002-1-page-9.htm#
I take your point (indeed I recall Dante – no less – has something to say about the topic.)
I picked Vlach, specifically, with malice aforethought though; I’d have thought its Romance affinity to be by no means obvious to premoderns. Admittedly even Romanian would probably have diverged a lot less obviously from the western Romance languages a millennium ago, though.
I see that your linked article does indeed support the view that the “Welsch” etymon had got broadened to mean “Roman”, which explains the application to my forebears, at any rate, though also suggesting that its primary significance was actually not linguistic. Unfortunately there don’t seem to be a lot of other cases for comparison where Germanic speakers would have encountered ex-Western Romans who didn’t speak Romance. I wonder what they called Albanians? Did the Basques get lumped in with the rest of the Romans?
David Eddyshaw: Don’t forget that when the Angles, the Saxons and the Jutes began colonizing Eastern England and Scotland Brythonic Celtic was not the only language they encountered on the ground: there is strong evidence that Latin/Romance (choose whichever label you prefer) was a spoken vernacular in Roman Britain, especially in its Lowlands/Eastern parts. Part of the evidence comes from Welsh itself, which (like Cornish and Breton) contains a huge number of Latin loanwords: their existence suggests large-scale (i.e. non-elite) bilingualism among Brythonic speakers.
(Incidentally, I am certain that it would be possible to write a whole paragraph, if not a whole page, of impeccably standard Welsh using nothing but borrowed Romance nouns, adjectives and verbs. Would you happen to know whether anyone has attempted this?)
So: Perhaps the term “Welsh” in Britain was originally linguistic, referring to L1 speakers of British Romance as well as to L2 speakers of British Romance (whose L1 was Brythonic). Naturally the term must have lost this linguistic meaning with time, as British Romance went extinct and Brythonic speakers (no longer fluent in British Romance or indeed any Romance variety) outside of Cornwall became the only “Welsh” individuals English speakers were in contact with.
In answer to your question on Albanian and Basque: Considering how little direct contact Germanic speakers ever had with either group, I suspect your question will forever remain unanswerable. There exist a few suspected loanwords of Gothic origin in both Albanian and Romanian, but there is no way to tell whether the words first entered Romanian (and thence spread to Albanian) or vice-versa: for geographical reasons the former scenario does seem somewhat more likely.
I am certain that it would be possible to write a whole paragraph, if not a whole page, of impeccably standard Welsh using nothing but Romance loanwords in the lexemes
I very much doubt this*, even if by “lexeme” you are implying “non-function word” in the most generous sense of “function word.” Still, it has sometimes seemed to me that as a general rule, if you can’t see how a Welsh word could conceivably be a Latin loanword, then it’s a Latin loanword (eisiau was the most recent word that tripped me up like that.)
Good thought about British Latin in general though; and as you imply, it’s really only the use of the “Welsch” etymon to mean “Cymric” that needs to be explained away if you are trying to make the term primarily linguistic. It’s also the case (now I think of it) that the previous politically-Roman character of the Vlachs would hardly have been very obvious to mediaeval Germans – perhaps even less so than the Latin origin of their language. So that might even be interpreted as supporting your point.
*Actually, having given the matter five minutes’ thought, yes, you probably could do that (in Perec mode) so long as you gave yourself a free pass when it came to all the function words. Don’t know of anyone who’s actually tried …
A professor of English who translates Reihenhaus as “rowhouse” (Collins has “terraced house” (British) or “townhouse” (American)).?
When I grew up in the 1960s/1970s I also heard (or read) those same stories about Rotwelsch – except that they were supposed to have last happened about the time when my parents were kids – in the 1920s. Nobody talked or wrote about Rotwelsch as contemporary phenomenon. There must have been something peculiar about the area where Puchner grew up.
@Etienne:
We had some discussion about British Latin before, as I recall; I remember raising Kenneth Jackson’s point about how Latin loanwords in Welsh don’t show the vowel mergers you’d expect to see on the basis of Romance in general (which he interpreted as evidence that these could not be loans from any spoken mother-tongue British Romance.) I remember you said this had been debunked, but I can’t remember what you said the modern explanation actually was.
ulr: “Rowhouse” (or “row house”) is very common AmEng in the vicinity of Philadelphia and Baltimore even if “townhouse” is more common in many other parts of the country. Searching google books finds it at least occasionally used in reference to Europe so maybe it also has an ESL usage.
David Eddyshaw: the source I referred you to in our 2017 discussion was this one:
Gratwick, A. S., 1982. “Latinitas Britannica: Was British Latin Archaic?”, in N. Brooks (ed.), Latin and the Vernacular Languages in Early Medieval Britain, (Leicester 1982), pages 1–79.
Because no on-line version was available you asked for a brief summary. Here it is, more detailed than what I offered back in 2017:
1-Jackson basically assumed that the loanwords must have entered Brythonic comparatively late: the assumption is unjustified, and the phonological “archaisms” could be explained as being indicative of the loanwords having entered at an earlier date than Jackson supposed. A point Gratwick stresses is that such loanwords could, *in Celtic*, have maintained older phonological features even as Celtic speakers spoke L2 Latin/Romance with a more “advanced” (=contemporary) phonology. An analogy he offers: Many older English loanwords in Welsh maintain older English phonological features, but no native Welsh speaker today would use this form when speaking English, nor modify the form of the loanword (making it more Modern English-like) when speaking Welsh today.
2-Jackson trusted established opinion among Romance scholars a little too much. One of the “archaic” features found in Latin loanwords in Brythonic, namely, the failure of intervocalic /b/ and /w/ to merge, might seem archaic, since the merger (in intervocalic position!) is pan-Romance. But as Chadwick points out, in Vulgar Latin inscriptions of the fifth century AD and earlier, the merger is conspicuously absent anywhere North of the Alps, suggesting here again that the spoken Romance of Roman Britain was not especially conservative when compared to neighboring forms of spoken Romance.
3-In the matter of vowel quantities, the problem is simple: based on the phonological adaptation strategies available to Brythonic speakers borrowing Latin/Romance vocabulary, a huge number of the Brythonic words which allegedly exhibit preservation of inherited Classical Latin quantity have a form which is identical to the form which would have been taken by a loanword from a Romance stage of British Romance where inherited vowel quantities were being/had been lost.
4-Finally, the reconstructed forms and the word choices of a large number of British Romance loanwords are quite Vulgar, and not Classical.
I recommend you (indeed all interested hatters) read the article: it is quite thorough and methodologically exemplary, and it is also an EXTREMELY entertaining read. Oh, and an extra hook: the author is of (at least) partially Welsh origin (he states explicitly that he never became fluent in Welsh, despite having been heavily exposed to it at home as a child).
Thanks!
@ulr, @JWB:
i think “rowhouse” is the standard throughout the u.s. northeast (and i think in other places as well) for non-detached, generally single-unit (or 2-3 unit) city housing. “townhouse” is pretty much a realtors-only term in my experience (except in mid-20thC novels where characters also have country houses). in nyc, we call a lot of them “brownstones”, but that has other architectural connotations as well. my house gets called both a rowhouse and a “brick brownstone” (which to me is a contradiction in terms, however cromulent), and no one is confused by either…
Not only does Rotwelsch sound like it had functional overlap with American hobo sign, the “cross inside a circle” that “signified that the house’s occupants would give you food” sounds identical with the hobo sign — an x inside a circle — commonly glossed as “good place for a handout”.
… ah, I see Puchner called out the connection (essay linked).
rozele, COCA has 1254 instances of townhouse, 88 of rowhouse, and 267 of row house. It’s not (just) a real estate speak.
I feel like in Boston “townhouse” is the normal term. You wouldn’t say “she lives in a rowhouse on Beacon Hill”, for example. My wife’s response was “well, most people just live in condominiums these days”.
Is there a class distinction, i.e., do students and low income people live in rowhouses/brownstones and higher income people live in townhouses? Here I understand the townhouse was the squire’s residence in London or Dublin, when he had to attend Parliament and (together with his family) participate in social activities. His main residence was of course his countryhouse.
Medieval Slavs apparently figured out the Romance thing pretty easily.
For example, in Polish wołoski means Wallachian and włoski means Italian.
Spot the difference.
Italians are shorter.
@LH: on your original question, what you post at least agrees with the standard summary descriptions of Rotwelsch I have read, so at least it looks reliable at a first glance. I’ve never read deeply about that topic, so I can’t judge any details.
@PlasticPaddy: I think “brownstone” is the most up-market of the descriptors for a row house.
The link to Researchgate didn’t have the original text, but it is available on the publishers website. However, it is in Dutch (with an English-translation abstract).
Hans den Besten. Jiddisch Hebreeuws in Nederlands en Bargoens. Tijdschrift voor Nederlandse Taal- en Letterkunde. 2006.
I wondered if there was any connection to the cant language of Polari, but WikiP makes no explicit mention of it, so.
The specific sort of housing stock and how it matches up with price and class distinctions will not necessarily be consistent across dialect regions. I think of the paradigmatic Philadelphia rowhouse as being a non-posh domicile in a non-posh neighborhood simply because there are quite a lot of them like that, but the ones in posher neighborhoods (Society Hill etc) are still called the same thing. Brownstone-heavy NYC neighborhoods tend to be posher, but pre-gentrification Harlem was an exception, so there was nothing incongruous about a junkie going uptown to score smack and
Up to a brownstone, up three flights of stairs
Everybody’s pinned you, but nobody cares
He’s got the works, gives you sweet taste
Ah, then you gotta split because you got no time to waste
No, but of course it contains all the Romani words the Ganoven* could grab.
* annoying professional petty criminals
Same here.
I think Tolkien identifies the piece of evidence that refutes the ‘Romanized foreigner’ explanation of wealh, the OE form of this word. I’ll let him tell it.
If you are back now, the key point from the present perspective is that a Gaelic-speaker was thought of as a wealh and in need of a wealhstod to speak to monolingual anglophones. JRRT’s footnote 15, which I didn’t transcribe last time, is “We see here the word [wealhstod] applied to a tongue which though Celtic was not British [and its speakers were certainly never Romanized]. Wealhstod became the ordinary word in Old English for either an interpreter or a translator; but that was at a much later date. It seems never, however, to have been applied to communications with the ‘Danes’.”
And if all this is correct (apart from the question of whether the “Danes” needed translators), then wealh was what Tolkien said it was, a “word of primarily linguistic import.”
If Etienne is right, the application of wealh to speakers of Brythonic was already an extension from an application to speakers of British Latin, so it would be natural enough to extend the term yet further, to cover yet another group of pesky aborigines who couldn’t understand Ænglisc no matter how loud or slowly it was spoken.
I should think we were quite happy without translators. What part of the point of a sword is hard to understand? It’s the rest of you lot who wanted to talk talk talk all the time, who needs that?
Eheu fugaces.
so it would be natural enough to extend the term yet further, to cover yet another group of pesky aborigines
And yet.
Tolkien again: “It did not occur to anyone to call a Goth a walh even if he was long settled in Italy or Gaul. Though ‘foreigner’ is often given as the first gloss on wealh in Anglo-Saxon dictionaries this is misleading. The word was not applied to foreigners of Germanic speech, nor [crucially] to those of alien tongues, Lapps, Finns, Esthonians, Lithuanians, Slavs, or Huns, with whom the Germanic-speaking peoples came into contact in early times. It was, therefore, basically a word of linguistic import, and in itself implied in its users more linguistic curiosity and sophistication than the simple stupidity of the Greek barbaros.”
What part of the point of a sword is hard to understand?
Nothing. But the point of a plow needs some cooperation, and after 876 (per the Anglo-Saxon Chronicle) that is just what the Danes in the Danelaw were doing.
True. But as you implied before, talking LOUDLY and s l o w l y may have sufficed. Also, marriage may have obviated the need for professionals.