Tablet magazine, an excellent source of discussion of all things Jewish, has reprinted a 2010 article by the late Cherie Woodworth (I wrote about her sudden death last year, and I still miss her and find it hard to believe she’s gone) on the titular subject; she begins with the great scholar Max Weinreich and the new edition of his magnum opus, History of the Yiddish Language (over 750 pages of footnotes!), and his very influential theory that
…Jews from Rhineland France, presumably through contact with Jewish settlements in southern Germany, converted from old Judeo-French to western Yiddish, which was more purely German with some elements of Latin or early French. In subsequent centuries—when, exactly, is a source of considerable debate—this language moved east with Jewish emigrants, settlers, and refugees, either in the 12th century (after the Crusades and persecutions) or in the 14th or 15th. There it picked up a significant cargo of Slavic vocabulary and expressions and became the Yiddish more familiar today: eastern Yiddish.
She then moves on to Paul Wexler and his very controversial book The Ashkenazic Jews: A SlavoTurkic People in Search of a Jewish Identity; you can tell from the title why it’s controversial. Her discussion of all this is fascinating and (like all her work) well written; I’ll just quote a bit on my own former specialty and leave you to read the whole thing when you have the time and attention span:
Comparative linguistics poses two genuine, and interconnected, problems when its methods are used to make arguments about history. The first is that in its most specialized details, the evidence and arguments are inaccessible to outsiders; Wexler will not be able to persuade historians about the origins of the Jews by discussing lexical inventories and phonemic shifts (especially as long as other linguists return fire with equally arcane and scientific-sounding counterarguments about other phonemic shifts). Second, despite its stress on precision and details, comparative historical linguistics is not as scientific or as purely historical as it seems; lost forms must be reconstructed, development must be interpolated, and thus no argument is definitive. The majority view among Yiddish linguists—a very small but committed cadre of scholars—is that Wexler’s argument is untenable.
Tablet followed up Cherie’s piece with an equally long and very lively one by their staff writer Batya Ungar-Sargon, “The Mystery of the Origins of Yiddish Will Never Be Solved,” in which she discusses the, shall we say, vivid personalities involved in the arguments. She starts out with an amazing story about a 1987 book, Origins of the Yiddish Language, which got a scathing review in Language that turned out to have been almost certainly written by Wexler under a pseudonym (though he denies it). She goes on to describe the various contending theories. Dovid Katz’s is that “the Jews arrived in what Katz calls ‘the cradle of Yiddish,’ the city of Regensburg, speaking Aramaic. It is this spoken language that provided the source material for the Semitic component of Yiddish, spreading both further east as well as west, to the Rhineland, replacing whatever language the western Jews were speaking.” Then there’s Alexandre Beider, who has a doctorate in applied mathematics from the Moscow Institute of Physics and Technology and “believes that western and eastern Yiddish are simply too different to have a common origin. Rather, Jews spoke German dialects until the 14th century, when gradually their dialects became different from those of their co-territorialists.” And of course Wexler, who “holds the controversial position that Yiddish is neither German nor Jewish but a Slavic language with German and Hebrew words slotted into Slavic grammar in a process called ‘relexification’:
Other linguists have not taken kindly to the Slavic hypothesis, nor to its author. “I have no impression that Paul Wexler is searching for the truth,” said Beider. “Sometimes I even wonder if he himself believes in what he writes. If he is not believing, but making a provocation, his writings of the last 20 years are oriented just to prove that Jews are not Jews. In this case, there is nothing to discuss.” Indeed, though he has a following amongst non-specialists, most linguists disagree with Wexler. “I respect him as a linguist, but I don’t agree with him,” said Steffen Krogh. Simon Neuberg called the relexification theory “very adventurous” but said ultimately it “seems more of a marketing trick.”
And finally, there’s Manaster Ramer, who “believes that Roman Jews speaking a Romance dialect, not French but related to it, lived in both western and southern areas of what would become Germany. When the German tribes invaded the Roman territories, these Jews learned perfect German”:
So, if the Jews who started speaking Yiddish originally spoke German, how and when did the Semitic component enter the language? The question itself is unscientific, said Manaster Ramer, ignoring as it does the historical context in which Yiddish came to be, which incidentally was a time in which German too underwent a similar process, incorporating loan words from Latin. Indeed, said Manaster Ramer, the influence of Latin on German was far greater than the Hebrew and Aramaic influence on Yiddish, “and yet no Yiddishist seems to asks, why does the massive Latin influence on German not mean that German is not German, if the much smaller Hebrew influence on Yiddish is supposed to mean that Yiddish is not German,” he wrote in an email. Any language spoken by polyglot people can incorporate words from their second or third language at any time. Furthermore, he said, almost all of the Hebrew and Aramaic words in Yiddish are accretions added to the language after the 13th century.
We’ll probably never know the truth, but I sure enjoy the back-and-forth. Thanks for the links, bulbul and Kobi!
I know very little about Yiddish, even if I heard it daily for a few months when I was a student as I mentioned recently, but I remember reading an article in a linguistics journal some years ago which named Ratisbonne (alias Regensburg) as a major centre in the history of the Jews in Europe before they migrated further east. I don’t remember reading anything about the language which gave reasons to consider it as other than a dialectal (taken in a large sense) variant of German.
One thing that struck me in those two articles is that the authors obviously know very little about linguistics, and like many non-linguists (and some linguists too!) take vocabulary as a major determinant of language classification. Wexler does not come through as a trustworthy character either as a historian or a linguist (I don’t know anything about him, but it is possible that he has done good work on questions of detail but is not credible about his approach to the larger problem – something not limited to the discipline of linguistics). (Some commenters on the article also offer their own reactions). Although Wexler (as cited in the articles) mentions something about syntax (especially one detail of word order), most of his arguments for a Slavic origin or substrate in Yiddish seem to be based on resemblances of vocabulary, which because of the inevitable occurrence of borrowing under conditions of bi- or multi-lingualism, is the least reliable source of classificational evidence.
As for Cherie Woodworth’s opinion of comparative linguistics as not very scientific, again it shows how superficial her acquaintance with the topic is. It is true that the discipline is opaque to the uninitiated, but that is characteristic of most sciences beyond the most basic level. It is also true that comparative linguistics uses reconstruction and interpolation, but these are not conducted in a vacuum or by “going out on a limb” on the basis of isolated examples: they must be based on attested changes in a number of words, through which it is possible to fill the gaps in documentation, and since most words consist of several sounds, the individual changes in sounds must agree throughout the words (ie, each sound or sound sequence in one language must have a correspondence in the other language). The chains of reasoning supporting this or that interpretation of which changes must have occurred to transform, for instance, a Latin word into a French one, or to trace a French word back to its most likely Latin original, must be justified in terms of what is known from numerous other instances. An example was provided a few days ago in the discussion of the origins of route (a French word) in Latin rupta. Based on many known examples, linguists can say with assurance that the single t of the French word has to come from a sequence of Latin consonants where at least the second one must be t, and conversely, that the Latin sequence pt while evolving into French through natural processes could only result into the single t of the French word. As for interpolation, it is a matter of filling gaps, and again that is based on analogy with patterns of change observed in many examples, not on a single linguist’s whim.
I was just about to write that both articles are basically informed by the “big bag of words” folk theory of languages, but m-l has already done it much better!
In terms of historical-comparative linguistics, there’s no plausible debate to be had: Yiddish is German. However, that isn’t really the question being debated. The debate is about what the Ashkenazi Jews’ ancestors spoke prior to adopting Yiddish, which is not a question to which comparative linguistics can provide firm or easy answers.
It is very unfortunate that Wexler should have chosen to muddy the waters with definition-twisting talk of Yiddish being “relexified” Slavic (as I discussed in my post). But if his arguments are taken as defenses of a Slavic substratum for Yiddish, they’re debatable but not immediately absurd. Notably, he claims that Yiddish consistently interprets particle+verb combinations using Slavic rather than German semantics – although this claim is only based on comparison with standard German – and that Slavic loanwords include basic Jewish religious terminology, notably treibern “to remove the nerves/sinews from meat”, which he thinks unlikely to have been borrowed from Gentile neighbours. I still wouldn’t trust them, though – the genetic evidence rather suggests an Italian substratum, and he seems far too ready to cherry-pick facts that can be made to support his claim and ignore ones that don’t.
Some years ago, I read an article, or maybe a book, that claimed that Modern Israeli Hebrew was in fact a Germanic language with Hebrew lexicon and morphology. I feel like Wexler may have been the author, but I don’t recall who it was. As I recall, this was a highly controversial claim.
That would presumably be Ghil`ad Zuckermann. Wexler claims that Modern Hebrew, like Yiddish, is Slavic.
I have found that between my ability to speak German and my limited knowledge of liturgical Hebrew, I can handle Yiddish fine, even though until I was an adult, I probably had no more exposure to Yiddish than a typical American gentile. (My grandparents, who certainly learned quite a bit of Yiddish when they were young, went out of their way not to use it with their own children and grandchildren.) To me, Yiddish just feels like a low German dialect, with a fair number of semitic words tossed in.
“The debate is about what the Ashkenazi Jews’ ancestors spoke prior to adopting Yiddish.”
Correct, and also what sort of historic German(s) Yiddish is formed from. Historical linguistics can say a bit more about that, but at a certain point simple cladistic branching breaks down. This is one way phylogenetics in linguistics is unlike that in biology.
“That would presumably be Ghil`ad Zuckermann. Wexler claims that Modern Hebrew, like Yiddish, is Slavic.”
It’s interesting to juxtapose the two. They both know their stuff and (over?)state their claims in a provocative and attention-getting way; the difference is that in terms of the origin of Yiddish the history is not fully known, so more is at stake, whereas the origin of Modern Hebrew (or, as Zuckermann would have it, “Israeli”) is no mystery. To my mind, that makes his sensationalizing more tolerable, and also reins him in from making the radical historical claims Wexler makes.
“Yiddish just feels like a low German dialect.”
The picky part of me is compelled to insist that “Low German” not be used as a synonym for “nonstandard German,” even though Germans call Standard German “Hochdeutsch.” I remember the first time I heard someone call Bavarian “low German”; since I only knew the term from dialectology, I was pretty shocked, since Standard German, dialectologically, is lower than Bavarian.
Hmm, I can’t say I’ve heard anyone call Bavarian/Swabian etc. “low” but I guess I can see where people would be confused by high/low as sociolinguistic categories reflecting standard/non-standard or prestige/non-prestige versus what I think is the right analysis here of high/low being geographical terms that basically reflect altitude above sea level / distance from the North Sea (the same way Upper Saxony isn’t necessarily posher than Lower Saxony, just further upstream and uphill).
“Second, despite its stress on precision and details, comparative historical linguistics is not as scientific or as purely historical as it seems; lost forms must be reconstructed, development must be interpolated, ”
Well the same criticism can be made of paleontology.
“Rather, Jews spoke German dialects until the 14th century, when gradually their dialects became different from those of their co-territorialists.””
Yeah, no. The dialects were different from the Standard German he is erroneously comparing them to. The Jews’ “co-territorialists” – the indigenous population, presumably – weren’t speaking Standard German, either, and basically only adopted it toward the end of the 19th century, and a lot of those places in the Rhineland are still kind of diglossic.
The Rhineland dialects are Franconian from about Mannheim north. Nuernbergisch is wide-open Franconian. I don’t know about Regensburg, but it is in the Oberpfalz rather than core Bavaria, so there’s a better chance that its dialect is Franconian than Bavarian. So why didn’t the author look at Franconian dialects rather than Standard German, which in any case didn’t exist in its present form in the period he is looking at?
““believes that Roman Jews speaking a Romance dialect, not French but related to it, lived in both western and southern areas of what would become Germany. When the German tribes invaded the Roman territories, these Jews learned perfect German”:
Couple of problems here:
“”“believes that Roman Jews speaking a Romance dialect, not French but related to it, ”
Well duh. The main Jewish communities in Gaul had been in the south. No one spoke French there, not only because the chronology is off, again, but because the geography is off too.
“When the German tribes invaded the Roman territories, these Jews learned perfect German”:
Two problems here. One is that it presumes that German came into southern Germany during the migrations as the Empire crumbled. This is a guess. of course the migrating tribes spoke German, but that doesn’t mean there wasn’t already a substantial German-speaking population in place, which may even have predated the Roman presence. It’s the same situation as in Britain, where we cannot know what the peasantry was speaking in anywhere. The Romans record the tribes as Celtic – well yes, the people they had a military and political interest in, the elites were Celts – but they had zero interest in the peasantry as long as it was docile. They were not ethnographers.
The second problem with is “Perfect German” – which variety of perfect German? Certainly not Hochdeutsch. SO the immigrants coming in from Gaul could very easily have settled in with German-speaking people who had been there for centuries, speaking a Germanic language that is directly ancestral to Yiddish. American Jews do not speak some kind of Judeo-English.
“Dovid Katz’s is that “the Jews arrived in what Katz calls ‘the cradle of Yiddish,’ the city of Regensburg, speaking Aramaic.”
This one takes some believing. Specifically, you have to believe that after more than half a millennium at the very least of Jews living in southern France and retaining Aramaic as their daily language, they suddenly flipped to German on contact. Hmmmmmm.
And finally:
“And of course Wexler, who “holds the controversial position that Yiddish is neither German nor Jewish but a Slavic language with German and Hebrew words slotted into Slavic grammar in a process called ‘relexification’:
During this period there were lots of actual Slavs and Balts who adopted German language and culture. Even without political dominance, the German cities formed a trade and cultural network that was hegemonic. How Slavic or Baltic are the varieties of German that these people speak? So how would the situation be any different with Jews, and why?
I smell an agenda.
Manaster Ramer (A linguist I greatly admire, by the way, and whose work deserves to be more widely known) is the one who hits the nail on the head: the Hebrew-Aramaic component of Yiddish is an element which is NOT due to a substratum and which differs little in nature from the borrowed Latin component of German.
I also agree with him that it is likely that at first Jews must have spoken perfect German, i.e. German indistinguishable from their Christian neighbors’. I think this is likely because in the case of Judezmo (Judeo-Spanish) it has been shown quite conclusively that, before the expulsion of Jews from the Iberian peninsula Jews were in no way linguistically distinct (leaving aside perhaps knowledge of some lexical items relating to religion).
Jim: you’re quite right to smell an agenda. Much if not most of the “scholarship” on Yiddish aims to highlight its uniqueness, its nature as a “fusion language”: I am certain that the continuing trend to compare Yiddish to Modern Standard German (a methodological blunder which might have been forgivable in the nineteenth century), rather than to Middle High German and Modern German dialects (both of which have been described in detail, incidentally) persists solely because it makes Yiddish look for more unlike German than it really is. This cherry-picking of the data to highlight the special nature of the language being studied is a common feature of scholarship on minority languages, unfortunately.
If I ever get tenure somewhere I would love to do a comparative study of scholarship on Jewish languages and scholarship on West Indian Creole languages. The similarities are striking. Just as scholarship on Jewish languages is obsessed with the Hebrew-Aramaic loanwords (as I myself found out when trying to find information on other diachronic aspects of Jewish languages: it is very difficult, and in many cases there does not appear to exist any relevant scholarship for aspects which in other languages are among the first to be studied), scholarship on West Indian creole languages is obsessed with the “African substrate”.
(Indeed, a colleague of mine once wrote a parody of such “scholarship” involving the discovery that West Indian Creoles, just like the languages of West Africa, have vowels and consonants, and how surely this cannot be coincidental and proves that the former have been profoundly influenced by the latter. Sadly, this parody is much closer to existing “serious” scholarship than most people realize).
Furthermore, both are equally oblivious to the fact that Jewish languages and Creoles of the West Indies are on average less mixed, lexically or otherwise, than most of the major languages of Europe (including English).
This West African substrate is of course imagined as being almost Platonically uniform, which is again strikingly reminiscent of how Jews before they acquired the Germanic ancestor of Yiddish are assumed to have been uniformly Slavic or Aramaic or Judeo-French (a language which probably never existed, by the way) or whatever-speaking. How anyone could believe that a diaspora population in the Dark Ages could have been linguistically uniform is beyond me.
(As I’m sure you’ll realize, I’d need a tenured position to write it, as such a study would offend a *lot* of people).
I must also second Marie-Lucie’s point about the author’s understanding of Comparative Linguistics being superficial. The author’s point that because “lost forms must be reconstructed” Comparative Linguistics is somehow less than scientific is especially egregious: many such forms reconstructed by linguists have subsequently been shown to exist (Saussure’s Laryngeal consonants in Hittite being the best-known example), which if anything highlights that Comparative Linguistics is far more scientific than other linguistic subfields, inasmuch as it makes falsifiable predictions which more often than not are shown by later discoveries to be correct.
Ben, JW Brewer: when I was in the Canadian West I met some Hutterites who insisted on calling their dialect German “Low German”, even though after eliciting some words and forms from them it was clear that it was a Bavarian variety. I agree that “Low” here must have been used in a sociolinguistic sense.
“American Jews do not speak some kind of Judeo-English.”
Some do:
http://forward.com/articles/10974/judeo-english-/
http://forward.com/articles/11026/judeo-english-part-ii-/
I agree that “Low” here must have been used in a sociolinguistic sense.
Germans use Platt for local dialect, even in Hochdeutschland, so the ambiguity could have been brought over from the old country.
I fully support Etienne’s comments, not just the one about me!
Etienne, I especially like your remark: Comparative Linguistics is far more scientific than other linguistic subfields, inasmuch as it makes falsifiable predictions which more often than not are shown by later discoveries to be correct.
Unfortunately, historical-comparative linguistics is not currently a fashionable field, and those people (such as biologists) who claim to be able to supplement or replace it usually have no training in it, or even in more basic elements of linguistics. For instance, pointing out that the “tree” model is not fully adequate to representing the relationships between languages is hardly anything new, and other models have been proposed (eg the “wave” model). It represents major divisions, not all the details at the micro level, such as the differentiation of local dialects.
I was surprised when I first encountered Bavarian German, having learnt only standard German in school, by not a few points where it resembles Yiddish much more obviously than the standard does, and which did indeed show me that I’d been overestimating the difference of Yiddish from German due to simple ignorance of gentile German dialectology. Things like “mir” for “wir” and “eppes” for “etwas” and not using the old preterite forms; some other things I’ve subsequently discovered are pretty usual in non-standard German pretty much all over, like the dropping of the genitive and simplification of the horrible adjective ending system of the standard. I’ve seen the “fact” that German devoices word final stops and Yiddish doesn’t adduced to “prove” that Yiddish is radically distinct linguistically from German; this becomes (shall we say) somewhat less persuasive when you discover that there are gentile German dialects that don’t devoice final stops and Yiddish dialects that do …
Once you know all this it becomes pretty difficult to subscribe to the more extreme views of the peculiarity of Yiddish, at least at a purely linguistic level. I don’t think there’s really any doubt that an awful lot of special pleading goes on over this issue, including, sadly, on the part of scholars with quite enough linguistic sophistication to know better.
Agree with everyone else that it seems exceedingly unlikely that the Hebrew and Aramaic components of Yiddish are substratal. They are nearly all obviously regularly related to the source language forms in a way which implies borrowing from the Ashkenazi readings of those languages into an existing Yiddish language rather than a pre-Yiddish linguistic evolution, and generally belong to the same sort of strata within Yiddish as the Romance and Latinate elements in English or the Chinese layer in Japanese – not like the Frankish element in French or even the Norse in English.
There is another way of looking at this, which is to assert that Yiddish is not German because of the sociological context, as witnessed to by the fact that it is written in Hebrew letters etc by people regarding themselves as culturally separate. Birnbaum’s book “Yiddish” is pretty upfront about this, but he goes on to name as a distinct “language” practically every language spoken by Jews on the same grounds (“Javanic” for Judaeo-Greek, etc.) This seems pretty circular reasoning, if not a positive reductio ad absurdum.
This is a fascinating discussion, and I’m glad I posted the articles that inspired it.
Throughout history it’s a rule that migration leads to assimilation, linguistically as well as culturally. This is valid either people move into an area dominated by others or they themselves get to be dominating. As far as I know there are, at least in the western world, only two known exceptions over the centuries (or even millenia): the Jewish and the Roma peoples.
This means, that there have to be exceptional explanations for this, not following the general ‘laws’. Which are those exceptions? I sincerely don’t know but it seems plausible, that the reason is to be searched in their way of life, their culture, religion, traditions and/or persecution of them – including a strong sense of ‘sticking together’ within the community.
The Roma spoke an Indo-Iranian dialect and have kept it until now, although it has been split up into subdialects all over Europe and western Asia, often mutually incomprehensible ones.
The Jews however probably had no common dialect. If I am correctly informed Aramaic was the language in the streets of Palestine in Jesus’ days, while Hebrew was restricted to lithurgical contexts. Consequently the Jews like others adapted the languages of the countries they arrived in. Since dialects of German not only today is, but for centuries have been both geographically and by number of speakers the largest and most central in Europe it’s no wonder if they became dominating in the European Jewish diaspora.
Neither is it any wonder, if they, trying to keep their identity, developed a variety – a shibboleth, if you like – of German, mixed with whatever from their culture or the environment. The Roma needed not to do that, since they already had a language of their own, marking inherency to the community.
So Yiddish, clearly a West Gmc dialect, is not much of a linguistic mystery but together with Romani a sociological.
This is a fascinating discussion, and I’m glad I posted the articles that inspired it.
I’ll say. I am in awe of the scholarship and erudition on display. Though I’m sure I know more Yiddish than most Hatterites, I can add nothing scholarly to the discussion. Perhaps an opportunity for an anecdote will present itself . . .
It’s important to note that Eastern Yiddish was spoken in a belt of mixed populations, where there was no Dachsprache to assimilate to. Throughout the area, villages of every conceivable locally spoken language were adjacent to each other, and indeed some villages spoke more than one language. There was no particular reason for the Jews there to lose their language. The only thing that made them different from anyone else was that there was no substantial core region where their villages were in the majority.
Matters were quite otherwise in the West, where the Jews did assimilate to German or French or what not.
We’ve been talking Wexler in the historical genetics community for some while, after he was cited by Elhaik in his (very faulty and presumably agenda-driven) piece of genetic research which purported to prove Khazar origin of the Ashkenazim. (Elhaik captalized on the genetic affinities of Ashkenazi Jews and Armenians, and “since the Armenians live in the Caucasus and so did the Khazars, the Jews must be derived from the Khazars”. Nah, really. Of course Armenians are of Levantine / North Mesopotamian Fertile Crescent origin themselves, and genetically a part of Levant / Palestine continuum to which the historical Jews belonged, while Khazars were Turkic, but if one’s got an agenda, then bigger problems get overlooked).
Anyway this brought the Wexler “relexification” book in the forefront of the discussion. For example Asya Pereltsvayg spent much effort demolishing it.
The Turko-Slavic hypothesis contradicted several linguistic facts. For starters, there is no Turkic or Uralic layers in Yiddish. But more importantly, elements of Slavic grammar are conspicuously absent in Yiddish. For example, Slavic languages have “a distinction between the instrumental or absolute use of the instrumental case, which does not occur with a preposition, and the instrumental of accompaniment, which is used with the preposition ‘with’”. In Upper Sorbian, heavily influenced by German, this distinction collapsed, but Sorbian still keeps 7 cases and those are marked on the nouns. Yiddish, on the contrary, has 4 cases, marked on determiners / articles rather than nouns themselves, like German. In Sorbian adjectives are modified by case; in Yiddish, like in German, they aren’t. However Yiddish, like German or Norwegian, and unlike Slavic languages, changes adjectives according to their attributive vs. predicative role. There is also vestigial dual number present, to a varying but unmistakable degree, in all Slavic languages, but not in German languages. Yiddish doesn’t have any vestiges of the dual either.
Genetics may give a hint to the question raised earlier in this thread, “could ancestors of Yiddish speakers have all come from one and the same language background”. Apparently the ancestors of Ashkenazim were a very small group, on the order of 400 founders, which makes it more plausible that they may have been linguistically fairly homogeneous.
BTW my Yiddish-speaking, Swiss-based ancestors used to explain that “Yiddish was a lot closer to Sweitzerdeutsch than to Hochdeutch” which made their lives there a bit easier.
But the preservation of German in eastern Europe was not limited to the Jewish population, and I think the relationship between dialects of Yiddish and the speech of the German diaspora may be underexplored. At times (and places) there must have been continuous contact approaching on full unity, at other times disruption caused by expulsion and convergence between dialects brought together by forced movement.
Don’t start a comment with “but” without quoting the referent. I started writing after Stefan’s comment, but got distracted.
Yiddish adjectives *do* inflect for case; and some nouns do, notably names.
Don’t Slavonic languages use the short form of adjectives as predicative? Admittedly it’s not like the totally flexionless predicative use in German and Yiddish, but there’s still a distinction there. Could be wrong – my school Russian has degenerated from vestigial to virtual over the years.
That Asya Pereltsvaig article claims that no traces of Turkic or Romance influence have ever been identified in Yiddish. In reality, Wexler claims several Turkic loans, and Romance loans are uncontroversial at least since Weinreich. Wexler also identifies what he claims to be traces of the dual in early Yiddish, which that post makes no attempt to address. It looks very much as though she’s responding to some one-line summary of Wexler’s claims, instead of actually reading them.
“after he was cited by Elhaik in his (very faulty and presumably agenda-driven) piece of genetic research which purported to prove Khazar origin of the Ashkenazim. ”
Which by the way has been pretty definitely refuted, I believe.
Michael – that’s not very convincing. The examples in those articles are on the level of Army jargon. And not just Army jargon – here’s what the Border Patrol sounds like “We wetted them down right after they EWIed – two OTMs and a coyote. VRed two of them and sent the rest to camp.” etc.
Arthur Koestler published a book way back in 1976 called “The Thirteenth Tribe” pushing the Ashkenazim-are-Khazars thing. Don’t know if this antedates Wexlerism or was inspired by it, or it all has a common source. I vaguely recall it as a seriously annoying book. (OK, Arthur, when did all these Khazars learn German?) Though the Khazars are actually pretty interesting in their own right, converting to Judaism apparently so as to maintain their cultural independence both from the Muslims and from the Byzantines. There seems to a whole political subculture devoted to pushing this thesis as a justification of anti-Zionism.
Once Western Yiddish died out, most Yiddish speakers farther east lived in places where the dominant local goyische tongue was something Slavic or otherwise non-Germanic (Lithuanian/Magyar/Romanian). There were pockets of ethnic Germans scattered through the same areas of course, but I’m not sure how many places there were where some gentile variety of German was the primary language of the local gentiles the Yiddish-speakers would deal with. That said, it would be quite interesting to know what similarities there might be between e.g. the dialects of the old German-speaking communities in Transylvania and the local variety of Yiddish.
I believe there are some words of direct or indirect Turkic origin in many Eastern European languages as a long-term result of the rather multicultural political history of the area (less politely known as the Tartar Yoke and/or Ottoman Yoke), some of which could then have found their way into Yiddish. Although perhaps for any given word in Yiddish there’s a more scientific way (in terms of phonology etc.) to assess the chances of it having come directly from a Turkic language rather than via such an intermediary.
Lameen: I’m pretty sure “Romance” was an error for “Romanian” there, given the context. I posted a comment to that effect, which is in the moderation queue.
David E: Koestler definitely predates Wexler; indeed, Koestler wrote on the Khazars in 1976 and died in 1983, whereas Wexler didn’t publish his Judaeo-Sorbian theory until 1991.
AFAIK it’s hardly controversial that there is Slavonic infleunce on Eastern Yiddish, obviously with loanwords, but also in more subtle areas like semantics and to a small extent syntax, though nobody without some weird a priori axe to grind could seriously maintain that the language is some sort of relexified Slavonic. How would that even happen? When were Eastern Ashkenazim suddenly surrounded by enough gentile German speakers to make them give up almost all their nice Slavonic lexemes?
Mind you, Litvak Yiddish has only got two genders, and it’s hard to believe it’s pure coincidence that Lithuanian only has two genders as well.
(On the other hand, plenty of hard-to-believe things are, in fact, true …)
Grammatic inter-infleunces between languages sharing an area aren’t that unusual BTW, isn’t it what the whole concept of Sprachbund is about?
JWB, good point. The presence of a handful of words from this or that language is not necessarily proof of direct contact or substrate. English is a Germanic language with a large proportion of Romance vocabulary, but browsing through a dictionary of English will also occasionally reveal words from a huge number of languages, most of which are unlikely to result from direct contact of English speakers with those languages. An obvious example is coffee, a word of Ethiopian origin (I think), which two or three centuries ago spread through a vast territory containing numerous languages which adopted the word as well as the substance even though there was no migration of Ethiopian speakers into the region, nor a wave of Westerners overwhelming Ethiopia and sending coffee beans back to their homelands. Instead the word spread through various trading intermediaries who tried to pronounce the word as best they could, rarely if ever actually hearing it in the original language.
@Dmitry:
Very true. So long as there were substantial numbers of Jews who spoke Lithuanian as well as Yiddish it’s no great wonder.
Mind you, I was reading somewhere not long ago that prior to the war Vilna was actually largely Polish and Yiddish speaking and only got Lithuanianised after the Soviet takeover. No idea to what extent smaller towns would have been Polish rather than LIthuanian speaking, or whether Vilna was exceptional.
@marie-lucie:
I think “coffee” is ultimately from Arabic qahwatun, which apparently actually meant “wine” in pre-Islamic times.
Doesn’t undermine your point, of course.
Dmitry: Grammatic inter-infleunces between languages sharing an area aren’t that unusual BTW, isn’t it what the whole concept of Sprachbund is about?
Inter-influences and borrowing of various elements of language have been studied within the context of different types of “contact”. A Sprachbund is a group of languages which may not be related but exist in an area with multiple ethnic groups who are in frequent contact and share many elements of culture, without a single language dominating the others culturally and socially. In such an environment, many people speak or at least understand two or more languages on a regular basis and adapt their speech to that of the people they are speaking to. This means also that since few of these people totally master the other languages, they tend to favour structures of phrases and sentences which exist in their own languages, or imitate structures found in other languages they hear a lot. The result is that even if speakers of each language are not borrowing many words from each other (since each language already has its own words for elements of the common culture), sentence structures tend to resemble each other across those languages, since many speakers are using a form of “translationese” which becomes the norm after a few generations.
With respect to Slavic or other influences on Yiddish (which I am not competent to pronounce on), I don’t think it has been suggested that a group consisting of Yiddish and some other languages spoken in the same area existed to form a Sprachbund. The socio-cultural historical contexts would seem to preclude such a formation.
I was reading somewhere not long ago that prior to the war Vilna was actually largely Polish and Yiddish speaking and only got Lithuanianised after the Soviet takeover.
You might have read it here.
@Hat:
indeed yes.
You see, I *do* pay attention sometimes …
DE: I think “coffee” is ultimately from Arabic qahwatun
Thank you, I had not run into this origin theory. As you say, it doesn’t matter to my point: just replace Ethiopia by Arabia, etc. One could say the same thing about tea, tobacco and other such substances which spread outside of their area of origin and rapidly became transnational though trade rather than population contact.
“There seems to a whole political subculture devoted to pushing this thesis as a justification of anti-Zionism.”
I don’t know much about this subject (the alleged Khazar origin of Ashkenazi Jews), but when I have encountered it, it has always been in an anti-Semitic context. My attitude has been, so what? What if they were converts a thousand years ago and not direct descendants of Moses? This didn’t stop the Nazis from persecuting them.
and not direct descendants of Moses?
the narrower anti-Zionist line of thought has always been the same, “give us a proof that they don’t have ancestral roots in Israel”. The more generic anti-Semitic sentiment also fed on description of Khazars as a some sort of an empire of evil.
“Coffee” is bunna in Amharic.
Oddly, my father, who was probably a native speaker of Yiddish (as well as of English), managed to use Yiddish to communicate successfully in German-speaking Switzerland when we were there in the late 1950s.
Coffee is definitely not < the Ethiopian word for coffee: it may be derived from Kaffa province in Ethiopia, or from the Arabic word, or both.
For the record, “low German” was indeed a slip on my part; however, I didn’t mean to suggest a nonstandard dialect, but rather a southern German dialect.
All in all, this has been a fascinating discussion.
OK, perhaps “derived from an Ethiopian word” or “a word used in Ethiopia”, or something similarly vague and non-committal. I thought I was following common knowledge, but common knowledge is often wrong. I had an Amharic-speaking Ethiopian student some years ago who insisted on telling our class the traditional story (which I had heard of) of goats becoming frisky after eating coffee berries off a particular bush. I am not sure if she told us the actual word she used for °coffee°. In any case, my point was about the rapid spread of a word to distant lands and unrelated languages without actual contact of spekaers.
@GeorgeW:
Anti-Zionism is what I meant, which is not by any means necessarily the same as antisemitism though it’s of couse all too often a cover for it. The value of the Khazar/Ashkenazi idea to anti-Zionists is straightforwardly that it undermines the historical claim of Jews to Palestine/Israel by claiming that the group of Jews most prominent in the European context do not actually have any historical connexion by descent with Israel at all. You could logically believe this without being an antisemite (presumably Koestler did) but the appeal to such people is obvious. [Come to that, you could be a perfectly consistent antisemite and still think that the Khazar/Ashkenazi thing is utter nonsense. I doubt whether it ranks high in the “rationale” of most of these people.]
Actual Jewish anti-Zionism is far from dead and was once actually fairly mainstream among the orthodox. But the eager espousal of anti-Zionism by antisemites and the history of Israel since the war have made it an increasingly difficult proposition.
And the word “tea” is a better example – where else an unaffricated pronunciation of te (otherwise known as cha) became the West European word for tea even though they traded in Canton, where 茶 is pronounced as cha in Cantonese. 😛
To complete the picture, antisemitic Zionism has historically been a significant thing, and has led to some very strange alliances. I think.Koestler’s motive in pushing the Khazar hypothesis was probably that as an assimilationist he was
looking for ammunition against the sort of people who felt the the best thing for Jews would be for them all to go “home.”
David Eddyshaw: actually, if you’re interested, I am willing to make a case that Litvak Yiddish does NOT owe its two-gender (masculine-feminine) system to Lithuanian influence.
Marie-Lucie: what is it about Central/Eastern Europe which precludes its forming a Sprachbund? Indeed Yiddish speakers, with their geographical mobility and multilingualism, may have played a major role in the diffusion of vocabulary, and perhaps grammatical structures as well. Some such features remain to be studied…
For example, here’s a feature I haven’t seen ever discussed as a product of contact: Yiddish and Southern German dialects have MIR instead of WIR as the first person plural pronoun. Now, this innovation is supposed to be due to inversion involving verb and subject pronoun, i.e. in instances such as HABEN WIR the final /n/ was assimilated to /m/ by the following /v/ (or was it still /w/ at the time? Ah well, that is immaterial to the scenario I am sketching), and subsequently re-analyzed as the initial consonant of the pronoun. Hence MIR. Okay.
But is it really a coincidence that, West of the Elba, this innovation affects most if not all German dialects (and Yiddish, of course) geographically contiguous to Slavic and Hungarian, both of which have a first person plural pronoun with an initial /m/, followed by a front vowel in most cases? It seems to me the above re-analysis of (HABEN) WIR to (HABEN) MIR could have been favored if not caused by Slavic-German (and Hungarian-German?) bilingualism. In this context it is all too easy to forget that, at the height of the expansion of the Slavs in Europe, Slavic was a continuum spoken everywhere East of a line stretching from Hamburg to Triesta, and that the eastward expansion of German must have involved large-scale Slavic-German bilingualism.
Sorry, next to last line: EASTward expansion. Thank goodnes this is Language Hat, not Geography Hat…
East/west confusion is downright traditional here.
It’s me or mi (rather than vi) in the southern and western dialects of Norwegian too, including in some otherwise very conservative inland varieties.
Re the word ‘tea’ as opposed to ‘cha(i)’, it’s from Fujianese(Hokkien)/Taiwanese dialects:
http://zh-min-nan.wikipedia.org/wiki/Tê
and also used in Malay/Indonesian:
http://ms.wikipedia.org/wiki/Teh
“I can see where people would be confused by high/low … reflecting … prestige/non-prestige versus … altitude above sea level”
Similar confusion prompted the renaming of French départements at the mouths of big rivers from [River]-Inférieure to -Maritime or -Atlantique.
Apparently, Ashkenazim have a little over 50% European ancestry, according to The Ashkenazi Genetic Consortium: http://slideplayer.us/slide/1719549/ (slide 15). That hardly fits the Khazar hypothesis, and makes it all the less likely that they arrived in Regensburg speaking Aramaic, but works fine for the other proposals.
You could logically believe this without being an antisemite (presumably Koestler did) but the appeal to such people is obvious.
Apparently Koestler thought if he could prove Jews weren’t Semites, it would end antisemitism. Which is charmingly naive, I guess.
Sorry, next to last line: EASTward expansion.
Fixed!
That hardly fits the Khazar hypothesis, and makes it all the less likely that they arrived in Regensburg speaking Aramaic,
I don’t see a connection. When a population is partly derived from the locals, how does this knowledge help you figure out where its other, migrant ancestors may have come from and what languages they might have spoken?
Once the Khazar-genetic-origin theory is taken off the table, I must confess that I am baffled by the apparent intensity of investment in the other rival theories on linguistic evolution, assuming (as the articles suggest) that it’s different from ordinary types of scholarly dispute in which people get emotionally invested for non-political reasons. I.e., it is not clear to me what the “natural” political/ideological implications of any of them ought to be, or why such-and-such theory (Rhineland v. Regensburg v. whatever) ought to be deprecated because of fears of its political/ideological consequences.
I assume the standard account is that the early diaspora communities in the Western parts of the Roman empire whose descendants might have centuries later made their way up from Italy / Southern France to the Rhineland did not preserve Aramaic as an intra-communal living language, but had some mix of koine Greek and later Vulgar Latin / Proto-Romance. Although I wonder how complete the evidence is on that or whether there’s a lot of inferences from silence from historical periods where the documentary evidence is very patchy.
[River]-Inférieure
Similarly, anglophones usually think of Lake Superior as being named for its size: it’s the largest lake in the world by area, and the largest in North America and third-largest in the world (after Baykal and Tanganyika) by volume. But it’s just ‘the upper lake’ in French.
(The Caspian Sea is larger in all respects, but it is brackish water and geologically speaking is an ocean basin.)
WIR and MIR
I am not happy with …en wir as the origin of Yiddish mir. I think that Slavic influence (which is documented in other cases, as several commenters have pointed out) is probably enough to explain it. Almost all Slavic languages have a form like my for ‘we’, and initial m also goes along with the occurrence of m as a formant in 1st person singular words such as German mein, Russian moy, etc.
It is common for 1st person singular and plural words and endings to share some sounds, since the meaning ‘we’ includes ‘me, along with one or more others’: for instance the -m and -mus endings in Latin sum ‘I am’ (note the vestigial English ending!) and sumus ‘we are’, and in the imperfect and several other regular tenses of most verbs, as in cantabam ‘I was singing’, cantabamus ‘we were singing. These endings are also reconstructed almost identically for PIE (which means that Latin is not the only source).
I don’t mean to suggest that the Slavic and Yiddish forms can be explained through the Latin ones, only to point out that there is a natural semantic relationship between the singular and plural person forms, which is often formally reinforced in the morphology (as is still perceivable in Latin). In the Yiddish case, the initial m of the Slavic pronoun might have “sounded” more appropriate to the meaning than the Germanic w (whether still [w] or already [v]), the meaning of which is not obviously supported by other semantically related forms.
As for dialectal Norwegian mi for vi, one would have to know when the change occurred, but that is often difficult to determine for long unwritten dialects.
I think that Slavic influence (which is documented in other cases, as several commenters have pointed out) is probably enough to explain [Yiddish mir].
I don’t have my dialect materials on me, but I distinctly recollect a book on the dialect of Vienna (probably this one) making the point that mir is common in a number Upper German dialects, including Viennese.
an old dialectal variation vs. an influences of the neighbors (in case of mir/wir) vividly reminds me of some discussions flaring in the genetic anthropology field today, where the problem is framed as “ancient substructure vs. admixture”. It’s relatively hard to spot, and harder still to date, “ancient substructures” (~~ the population group has already been heterogeneous before it contributed to a derived population)
In the case of Yiddish, and even Viennese, influence from Slavic neighbours is possible as I wrote above, but it would hardly explain the Norwegian case. However, interchange (= change in either direction) between [w] and [m], although not very common, is attested in some languages. At a time when the ancestral [w] was becoming [v] in most Germanic languages and dialects (not in English), some instances of [w] could have become [m] instead, especially if there was a reason to analogize with other forms, such as mein etc, and/or with influences from neighbouring languages.
Bach’s “Peasant Cantata” begins “Mer hahn en neue Oberkeet,” presumably in the dialect around Leipzig
I don’t think it’s implausible that “mir” for “wir” is a language-internal development due to reanalysis of -m wir, owing nothing to Slav influence.
Analogy: Reanalysis of the consonant cluster at the junction of verb + subject pronoun is the usual explanation given for the -t in the 2sg of verb forms like “thou dost”
And in fact there are Highest Alemannic dialects where the *reverse* process has actually happened in the 1pl: the Bosco Guerin dialect has -v as the ending for 1pl in monosyllabic verb forms like “viar tiav” = “wir tun”
Seen as a graffito in a lavatory in Regensburg years ago (from memory, happy to be corrected by passing Bavarians)
Mir saan mir, mir saan mir
Mir saan stiarka as a Stiar
Mir saan stiarka as a Bahn
Weil mir Niederbayern saan!
We are us, we are us
We are stronger than a bull
We are stronger than a train
Because we’re Lower Bavarians!
Etienne,
“and that the eastward expansion of German must have involved large-scale Slavic-German bilingualism”
Yes, but probably only on the part of Slavs. German was culturally and probably economically dominant.
M-L,
“In the case of Yiddish, and even Viennese, influence from Slavic neighbours is possible as I wrote above,”
The flow of borrowing is usually (if not always) from the high prestige language to the low prestige language, and where the low prestige language is the source of borrowing, the usage is usually high prestige. So when Hindi words like “pajamas” or “chukkar” or “mufti” entered English, it was because they marked elite status.
The obvious exception is terminology for exotic items that exists only in a low status language, and even then the relative status of the languages can’t be taken for granted. A lot of English borrowing form Irish in the Middle Ages occurred when English was a low status language in Ireland – no literature of its own, elites intermarrying with Irish elites, that kind of thing. And in the case of a lot of borrowings in the American colonies in the early years, English language and culture were far from dominant.
The cultural dominance of German in Central Europe was pretty entrenched, and it was unassailable in Vienna. I wonder if the pattern of borrowing from Slavic sources reflects this.
Jim,
“Pajamas” marked elite status? wasn’t it rather an “exotic item” that existed only in a local language? like “kimono” in Japan, for instance. Both pajamas and kimono were adopted as loose, light, comfortable clothing suitable for relaxed moments at home, they did not make it into business attire or even social occasions.
The cultural dominance of German in Central Europe was pretty entrenched, and it was unassailable in Vienna : what period(s) are we talking about?
It just now occurred to me that the consonant cluster assimilation seen in “etwas” > “eppes”, seen in Yiddish and various German dialects, is actually very closely parallel phonetically to eg “gebenwir” > “gebemmir”, differing only as oral vs nasal.
Dmitry: I should have mentioned that the other half was identified as Middle Eastern, so not much room for Khazar ancestry there either,
Re wir/mir again: It occurred to me to look at the specimens of Western Yiddish in Solomon Birnbaum’s “Yiddish”. They show “mir”, same as Eastern Yiddish. So if there is any question of this being a Slavonic influence, it would have to be pushed back to a period before the Eastward migrations of the Ashkenazim and their sojourn among speakers of Slavonic languages. But given that the form is not confined to Yiddish anyhow, I suppose even that isn’t a clincher: you could maintain that it was a Slavonic or Hungarian influence on Upper German generally. As far as I know there isn’t much other evidence for such influence, though.
@Etienne:
“actually, if you’re interested, I am willing to make a case that Litvak Yiddish does NOT owe its two-gender (masculine-feminine) system to Lithuanian influence.”
How could I *not* be interested? This is Language Hat! Tell all …
One ought to keep in mind that for most of the last thousand years the whole area of central-eastern Europe has been inhabited by people speaking Germanic, Slavic, Baltic, Hungarian etc. tounges. It therefore seems likely that a lot of local pidgins, lingua francas etc. existed (Yiddish being one, and perhaps the only survivor). Furthermore it should be taken into account that even German wasn’t ‘German’ prior to Johann Wolfgang von Goethe and Friedrich der Große but a patchwork of dialects, presumably more in number than the German principalities prior to Otto von Bismarck.
Such an environment probably makes it a hopeless task to sort out wherefrom every local (spoken) mir as an alternative to wir comes. Germany as we know it is a modern phenomenon, named by the language(s) Deutsch originally simply meaning ‘folkish’.
David: Litvak dialects with their two-gender system cover more than just Lithuania: they were found in most of Poland as well, so if Lithuanian influence is the reason why Litvak dialects only have two genders this begs the question: why didn’t Polish (A language with a vigorous three-gender system, plus the animate/inanimate, masculine human/masculine non-human subdivisions) preserve the inherited Middle High German gender system? Especially since historically Polish was the more prestigious language: far more Polish loanwords are found in Lithuanian than Lithuanian loanwords in Polish, for example.
It turns out that Litvak dialects share another innovation which I think holds the key to the loss of the neuter: this is, in nominal case-marking, the complete collapse of the accusative/dative distinctions: in pronouns the former dative case was generalized to mark the accusative. So where other dialects of Yiddish had IM (masculine accusative) versus ES (neuter accusative), with IM as the dative for both genders, Litvak Yiddish had IM as the accusative as well as the dative for masculine and neuters alike. Now, add to this the fact that this IM form is also the one surfacing after prepositions and you can see that this spread of the dative at the expense of the accusative must have sharply increased masculine-neuter syncretism, so much so that I suspect that the number of contexts where distinct masculine and neuter forms surfaced had become so few in number that a merger of the two genders was a natural outcome.
As natural, indeed, as the loss of the neuter in other Indo-European languages, such as Lithuanian…or Gaelic or Hindi, for that matter, neither of which may be reasonably claimed to have influenced Litvak Yiddish. My point is that losing the neuter is not rare in Indo-European, so that it is reasonable to assume the Litvak Yiddish/Lithuanian similarity to be coincidental. Especially since dative/accusative case syncretism, as far as I know, is quite alien to any variety of Lithuanian or Polish.
Stefan Holm: German has plenty of dialect diversity, granted, but I think you are overstating the case: Middle and Old High German are well-attested, and as a rule most German dialects (and Yiddish!) are later, changed forms of these earlier languages. The lack of a single dominant capital for most of the history of Central Europe/Germany means that there were a lot of nuclei/prestige centers which innovations could radiate from, making the history of German murkier than that of English or French (where London and Paris (respectively) quickly became the dominant prestige center), granted. I certainly would not call Yiddish a lingua franca, even less a pidgin: indeed Yiddish is an excellent example of a language heavily influenced by contact without thereby becoming pidgin-like.
On MIR as due to language contact with Slavic: I freely admit the Norwegian dialect forms with initial /m/ weaken my case somewhat.
David: the Bosco Guerin form “tiav” corresponding to Standard German “tun”, with /v/ from the initial consonant of WIR, is quite fascinating, because a similar type of person-marking endings, deriving from post-verbal subject pronouns, is typical of Lombard dialects in Italy, i.e. of those dialects geographically closest to Southern Swiss German varieties such as the Bosco Guerin dialect. I smell a local Sprachbund.
M-L,
““Pajamas” marked elite status? wasn’t it rather an “exotic item” that existed only in a local language? like “kimono” in Japan, for instance.”
That was the point. Only elites had exposure to those exotic places and cultures and the ability to return to Britain on a regular basis. It was a form of snobbery. It’s the same impulse as Orwell denounced when he talked about (English) people who hated suet puddings and a whole long list of other iconic cultural items, with the implicit addendum that they probably just loved clafoutis.
“The cultural dominance of German in Central Europe was pretty entrenched, and it was unassailable in Vienna : what period(s) are we talking about?”
The Hapsburg period? There were substantial German-speaking minorities everywhere and they controlled trade and academic life. (And anyone who posed the slightest threat to that hegemony was going to be the object of jealousy.)
Stefan,
“Furthermore it should be taken into account that even German wasn’t ‘German’ prior to Johann Wolfgang von Goethe and Friedrich der Große but a patchwork of dialects, presumably more in number than the German principalities prior to Otto von Bismarck.”
This is part of the reason English its dominance in the American colonies/US despite slight majorities of Germans for long periods.
@Etienne:
You may very well be right about “tiav”; unsurprisingly there is huge Italian influence in the Bosco Guerin dialect.
It’s strikingly divergent in other ways, not least in that it preserves three different vowel qualities in unstressed syllables where nearly all other German dialects have collapsed the distinctions to -e- as far back as Middle High German. It’s as if there were still English dialects in which “son” was “sunu.” Interesting language(s), German …
m.-l., what other languages do you know of with /m/~/w/?
@Etienne:
many thanks too for the very interesting thoughts about NE Yiddish and the loss of the neuter. Glad I took you up on it …
Not sure about the loss of the neuter in NE Yiddish arising from a collapse of the accusative and dative, though.
I grant you “mir” for “mich” and “dir” for “dich” as NE-isms, but I thought the masculine sg personal pronoun was the same in both cases in all Yiddish, varying in/im depending basically on stress rather than case. And adjectives also end identically in masc acc/dat, with the variation of -m/-n being conditioned by the preceding consonant rather than case. Similarly those masculine nouns that inflect at all have the same form for acc and dative across Yiddish.
Is the *feminine* accusative replaced by the dative in NE Yiddish? (Not a rhetorical question – I don’t know, and you presumably do.) If not, isn’t it equally possible that the replacement of neuter “es” by “im” is simply the *result* of the collapse of the masculine/neuter distinction?
I think you’re pretty certainly right that it’s just coincidence about Lithuanian having lost its own neuter, though. The more I think about it, the less likely a causal connexion seems, especially in view of the stuff Hat linked to, which tends to suggest that there wouldn’t have been all that many Jews who spoke Lithuanian even in what’s now Lithuania, let alone (as you point out) over all the rest of the area where Yiddish lost the neuter.
As I said, many things which are hard to believe are, as a matter of fact, true. And this seems to be one of them!
@Y:
M/W:
Hittite has 1pl verbal forms in -w- where non-Anatolian Indoeuropean has -m-, but with -um- for expected -uw-, a change which takes place elsewhere in this phonological context and is not morphologically determined; the -w- is the basic form.
tarnumeni “we leave” but edweni “we eat” contrasting with Latin -mus, Greek -men etc etc
I don’t know much about IE but I think there are thought to be other doublets among things like derivational suffixes where one item has -w- and the other -m-. I don’t know if anyone has linked that with the odd initial in “we” and its cognates vs “me” but then the whole personal pronoun system is so suppletive anyway I suppose that would be supererogatory …
In Crow, the reflex of Siouan m appears as b/m in free variation initially, b after an obstruent, w between vowels, and m elsewhere.
In later forms of Akkadian, original intervocalic -w- becomes -m-: awilum “man” > amilu
Lenited forms of m turning up as nasalised w are common, but it’s probably cheating to count those.
Etienne: You wrote that Middle and Old High German are well-attested, and as a rule most German dialects (and Yiddish!) are later, changed forms of these earlier languages.
This is of course in a sense correct but remember that we are dealing with the written language. How it was spoken we know next to nothing about. Until recently writing was restricted to very few people, scholars educated at monasteries or a handful of universities. Eg the regional medieval laws of Sweden show little diversity (with the exception of the oldest one, the Older Law of the West Geats, but mostly just due to lack of influence from Middle Low German). But when written descriptions of our dialects appeared (around 1700) the diversity turned up to be huge.
Think about it – it’s not until the very few last decades that we through ‘social media’ like Facebook etc. really have got direct access to the ‘people’s language’. Earlier it was almost always given by the educated elite or passed through proofreaders. Even when people today get a microphone stuck under their nose they try to apply themselves to ‘proper language’.
After all, spoken language is the real stuff while the written one is just a fossilized sublimation of it. In other words, few commoners probably spoke the dialects of Hildebrandslied or Heliand. Written works like those may have contributed to the levelling of dialects but the diversity is due to (not documented) speech.
“This is of course in a sense correct but remember that we are dealing with the written language. How it was spoken we know next to nothing about.”
Loosely related to this, medieval Arabic texts written by Jews in Cairo were written in Arabic with Hebrew script. Because, unlike Muslim scholars and scribes, they weren’t socially constrained by standard Arabic, they wrote in the language they spoke. So, this is a great source for the language that was actually spoken at the time.
they wrote in the language they spoke
The great 11th century Jewish scholar Rashi, who lived in Troyes, in northern France, wrote his biblical and talmudic commentaries in Hebrew. I’m told that he occasionally inserted a word or two of French, written in Hebrew script, to clarify a point. This too would provide an indication of the pronunciation of French 1,000 years ago. Interestingly, the Wiki entry on Rashi says he had family in Mainz and Worms, which suggests he also knew at least one German dialect of the day or a precursor to Yiddish. I’m not aware of German words showing up in his work. If there are, they may illuminate the development of Yiddish orthography.
Apparently Koestler thought if he could prove Jews weren’t Semites, it would end antisemitism. Which is charmingly naive, I guess.
Naive, certainly, but I wouldn’t apply “charmingly” to such a thoroughly nasty person as Koestler was (something I didn’t know when I read The Thirteenth Tribe). I find nauseating the way his friends — and even people who were not his friends — fall over themselves trying to excuse his violence towards and rape of different women, as if being a reasonably good writer (though far inferior, in my opinion, to many others) and having the “right” attitude to communism justified everything else. Worse — if that were possible — his murder of his last wife is barely even mentioned, least of all talked about. You may say that “murder” is too strong a term, but I don’t know what else you can call bossing a woman much younger than himself into entering into a suicide pact? If he had survived, and she hadn’t, the law would certainly have treated it as murder.
M/W
DE, thank you for all those examples. At one point I assmbled examples of similarly “odd” or “sporadic” changes, mostly from Romance, Germanic and Celtic, but that was long ago and I can’t find the list anymore.
However, I know that there were a few examples in Spanish, from which I remember the following:
– Latin uimin- stem for ‘wicker’, Spanish usually vimbre, occasionally mimbre
Among non-European languages, Tlingit does not now have the sound [m], but older transcriptions and borrowings from or into neighbouring languages which have [m] in words where Modern Tlingit has [w] show that the Tlingit change is quite recent, as in:
– Older Tlingit ts’aam ‘crab species’, Modern Tlingit ts’aaw, Southern Tsimshian ts’aam ([m] and [w] contrast word-finally in ST)
– “Tsimshian” ts’∂m-(t)sxan, Tlingit ts’u:-tsXan (where ts’u:- is from *ts’∂w-, from Tsim *ts’∂m-)
– English James, Tlingit and Tlingitized English Dzéiws (courtesy of Tlingit linguist James Crippen)
English James, Tlingit and Tlingitized English Dzéiws
I love that kind of example; it makes the change apparent and memorable even if you don’t know the language in question.
You get an intrusive m occasionally after w in the Algerian Arabic of Dellys, eg dyawlu > dyawmlu / dyamlu “theirs”. In Koyra Chiini, nasalised w became m in mow > mom “hear”.
I agree with Athel about the hateful Koestler and the extraordinarily passive attitude of his left-wing friends, including Richard Crossman and many others. I don’t know if Orwell knew about the raping, but it seems like everyone else did. From Jill Craigie’s obit in the Indie (she was an actress, filmmaker & the wife of Labour Party leader Michael Foot):
In the comments to Ungar-Sargon’s article, Dovid Katz links to a reply he’s written to both articles, here. It’s long and it’s in Yiddish, which I don’t read, but I managed to decipher some of it. His main point is that both articles are sensationalizing things way too much. He takes great exception with the claim that Weinreich’s works were colored with ideology, and adds more bits to the Wexler/Slobodanski story. His most pertinent points, which I think are spot-on, are that the articles concentrate more on the juice and less on the science, and that what they make out to be a battle royal between disparate viewpoints is actually a normal case of scholarly quibbling. Certainly, if we leave out the unsupported extremes of Wexler and Beider, we are left with the variants of Weinreich, M-R and Katz, which are not that far apart geographically, and they don’t seem to be disagreeing with each other all that vehemently, either. Certainly the differences are not unusual for discussing the homeland of an 800-year-old language family.
It’s worth struggling with Katz’s note if one can, because his humor and reasonableness come through, sauced with Yiddish liveliness.
@David E., thanks! I’ve see m~w sporadically in Northern California languages, but I’m not sure if it’s regular anywhere.
…and thanks, Lameen and m.-l.!
I agree with Athel about the hateful Koestler and the extraordinarily passive attitude of his left-wing friends, including Richard Crossman and many others.
Thanks, AJP. How can one explain or understand the extraordinary passivity? Is it really the case that we can ignore the fact that someone was a rapist thug if we agree with his politics? Erwin Schrödinger was in some ways a contemptible person whose major interest in life apart from physics was to bed as many women as he could, but, as far as I know he achieved this by charm rather than violence, and, moreover, he was a very great physicist, whereas even we forget his personal behaviour I’m not sure what was great about Koestler. (The Case of the Midwife Toad was typical of the sort of rubbish non-scientists write when trying to write about science. The Sleepwalkers was perhaps better, as it was as much history as science. The Act of Creation I found so boring as to be almost unreadable.)
The google n-gram viewer suggests that mentions of Koestler have been in steep and continuous decline since his death (obviously the fate of all but a handful of writers who are famous while alive) – it would not surprise me if many/most even reasonably well-read people under 40 have never heard of him. I myself don’t have a sense of whether his later writings (more, um, eclectic or exotic in their subject matter) would have attracted any sizable audience but for the early fame built on Darkness at Noon and The God That Failed.
The passing of the Soviet Union would you think have made the high percentage of Communist sympathizers in the Western intelligentsia look even worse in hindsight but I think what happened instead was the ability to change the subject and treat that shameful history as all irrelevant water under the bridge, with the side effect that those who were anti-Communist when that was rather noteworthy seem likewise less noteworthy in hindsight despite (or perhaps because?) of having been vindicated.
@J.W. Brewer: I’m still a few years under 40, and I’d like believe I am pretty well read. It’s not quite true that I’d never heard of Koestler, but I had to look him up to refresh my memory of who he was as I was reading this comment thread. I don’t think I’d ever heard of his writings except for those you mentioned—Darkness at Noon and his contribution to The God that Failed.
Moreover, I don’t think that many people my age and younger read much of the material that was written about Communism before the fall of the Soviet Union. I am personally very interested in Soviet political history, but much of the material about the written subject prior to the 1990s is full of gaps and guesswork. I feel like I’m much more likely to get an accurate picture of what was going on in Kiev in 1936 from a book written in 2000 than one written in 1940. I think Orwell may be the only mid-twentieth-century writer who dealt with the subject of Communism who is still widely known among my peer group, and the staying power of his work is probably enhanced by the fact that his two most famous anti-totalitarian novels were not historical by allegorical.
The term awwoxiwant is a jocular example of m/w interchange.
(Sorry in advance, if my attempts to link end up in a haywire).
I agree with Athel about the hateful Koestler and the extraordinarily passive attitude of his left-wing friends
Funny world. In Sweden Koestler (like, to some extent, Orwell) is a favourite of the right wing, mainly because of his Darkness at Noon and the anthology The God that Failed. I imagine, this could be explained by the Swedish establishment’s inherited hatred, not primarly against communism, but against Russia. A pendant, maybe, to what appears to have governed US foreign policy since WWII: ‘My enemy’s enemies are my friends’.
Even The Case of the Midwife Toad is highly estimated in Sweden. But never ever otherwise than as an introduction to a litany against Trofim Lysenko. ie individual western scientists might be frauds but among those who stole Finland from us back in 1809 it is part of an evil system or even an evil race.
And, Athel, what has Erwin Schrödinger’s amorous life to do with a rapist like Koestler? I remember my school days struggling with his wave equation and the satisfaction when I finally understood (some of) it. Even more I after nearly fifty years (together with most modern physicists) feel the dissatisfaction of not yet having learned what happened to his poor cat.
JWB: It’s also the case that essentially all the anti-Communists were ex-Communists, which meant that among those in the know they were already somewhat tainted. This doesn’t so much apply to those who left the Party after 1956. (Disclaimer: My parents were Communists lato sensu but were always opposed to Stalin and were never CPUSA members. My father voted the Socialist Workers ticket to the end of his days, however.)
Brett: Orwell is also studied in schools, which certainly creates familiarity if not popularity.
Athel, Crown: Reading Koestler’s WP article, it seems to me that he had his nuts (pun intended) pulled out of the fire by women on many occasions. One got him out of Spanish Nationalist detention, and another out of a French internment camp. This suggests to me that he looked for, and found, a huge amount of heroic devotion from the women in his life, which gave him a huge sense of entitlement to do whatever he wanted to with them, and all other women, in “exchange”. Unfortunately, not an unfamiliar pattern.
John Cowan: surely you mean “the only anti-Communists whose books had a reasonable chance of being read by the leftist Western intelligentsia were ex-Communists,” which is not quite the same thing. (And of course there were other ex-Communists whose readership was more right-of-center – the whole book is rather a slog to recommend to Brett but the little set piece where Whitaker Chambers rejects the intellectual adequacy of scientific materialism while watching his young daughter in her highchair smear oatmeal all over her face is a minor classic of 20th century literature.) It is true, however, that this led to a genre of literature that tries to explain how extraordinarily deceptively seductive and noble-seeming Communism was to the naive young idealist, in order to portray the author as having always been a good and thoughtful person who was cunningly seduced rather than a not-so-good person whose motives for getting mixed up with the Reds in the first place had been not-so-noble, or even an uncritical sucker who could be taken in by something whose level of deceptive seductiveness was not particularly superhuman or awesome.
Paul Ogden: Rashi did include a few handfuls of Germanic-language glosses, which might not otherwise be discernible without his preface “in the language of Ashkenaz.” Most I find indecipherable.
Two more thoughts on other comments:
1. “Mir” is found even in the very earliest Yiddish texts, and it’s always been pretty much exclusive in Yiddish texts, unless the authors were trying to make it pseudo-German; in fact, the use of “wir” and “und” are probably the two clearest signs that a text is in (or is attempting to be in) Hebrew-letter German, not Yiddish. In any case, it’s clear that it predates Slavic influence and reflects instead the particular German dialects that Yiddish drew from.
2. As for Polish Yiddish having two genders, that’s not really true if by “Polish” one means the Central Yiddish dialect found in Poland, which has a robust neuter. There are times when loss of postvocalic r in much of that dialect can blur the distinction between “di” and “der,” but the result is nothing like the common gender of Dutch, etc. It is true, however, that two-gender “Lithuanian” Yiddish was spoken in a sizable chunk of what is now Poland, and an even bigger chunk of interwar Poland.
JWB: Yes, certainly. I was speaking of the anti-Communists in question, viz. those like Koestler, rather than all or any anti-Communists. I do think, however, that anti-Communist fervor was highest, at least during the relevant period, among the ex-Communists. (This was rather before Nixon and Joe McCarthy and such-like opportunists.)
@John Cowan:
I, too, like vowol harmono the best.
Rashi did include a few handfuls of Germanic-language glosses
Time was generous today, enabling me to read the article by Batya Ungar-Sargon: “Rashi’s biblical and Talmudic commentaries (which, in addition to 3,000 Judeo-French glosses, contain, astoundingly, 24 glosses in Yiddish) . . .”
I have to wonder how one could hold that a few snatches written in Hebrew script 1,000 years ago
are Yiddish, as opposed to, say, a German dialect. Curious too as to whether Rashi dedicated certain letters to vowel functions, as is common today.
According to the Hebrew Wiki article about him, he inserted what we would call quotation marks before the last letter of the French words to indicate their non-Semitic provenance. This is similar to current Hebrew practice — and I think much older too — of inserting quotation marks to indicate an acronym.
The Wiki article says he transliterated 1,000 Old French terms. A glossary of these words is available here.
Re: [m]–[w] alternation: According to chapter 18 of the World Atlas of Language Structures Online, another language with this alternation is Eyak, a Na-Dene language of Alaska: in that case it is allophonic variation between two phones in complementary distribution, with /w/ being realized as [m] before a nasalized vowel and as [w] elsewhere.
Re: verb inflectional endings accruing the first sound of nominative pronouns: Another example is Spanish, where a number of first-person singular present-tense verb forms end in -oy rather than the usual -o — soy “I am”, estoy “I am”, voy “I go”, doy “I give”, etc. — apparently due to a formerly common tendency to follow them with the pronoun yo “I”. (Or so I’ve read. I haven’t encountered an explicit explanation for why this only affected a handful of verbs, but I notice that these forms are all stressed on the last syllable, by contrast with the regular -o ending, which is unstressed. Most of the relevant forms are monosyllabic, so obviously they have no choice but word-final stress; the exception is estoy, where the word-final stress is odd, but consistent with the other forms (est´s, está, están rather than the regular *estas, *esta, *estan). So I assume that the word-final stress somehow enabled the y to be picked up.)
Norwegian me has a much wider distribution than mi. I don’t know for sure, but I think me is the original form and mi a sociolinguistic compromise in a transitional zone. If so, it would seem to go against simple phonetic change.
But it’s interesting that both Norwegian me and German mir are identical to the first person singular dative. How can that be? A phonetic change in German leading to a conflation of forms that in turn spread to western Norwegian? But then, also the WNo. 2p plural de is identical to the singular dative.
No, the vowel quality is just an east/west thing, I think, with western dialects keeping ON e: and eastern dialects (like Danish) changing it to i:.
The 2pp change of ér to þér was on its way already in ON. Maybe the resulting conflation of the 2pp nominative and the 2ps dative led to an analogical change in the 1p paradigm.
I came across a description of Judeo-French that may be of interest to readers of this thread.
Caveat: Paul Wexler is one of the researchers. This would be the same Paul Wexler who, according to Batya Ungar-Sargon’s article, denied being the Pavlo Slobodjans’kyj who in a review utterly panned the book ‘Origins of the Yiddish Language’, and whose email address at Tel Aviv University, I just discovered, begins with pavlo@.
Ha!
Yiddish, as opposed to, say, a German dialect
In the context of Rashi, I’d say he was using Jews’ German (I), as opposed to the Jews’ German (II) that has replaced Western Yiddish. The latter is not a descendant of the former, though used by the same population at different times; both are variants of ordinary German.
1-Paul Ogden: here is to my mind the most interesting part of the description you linked to:
“So far, no linguistic features have been shown to differentiate Judeo-French as a whole from the Christian varieties of Old French”.
This seems to confirm that there indeed didn’t exist a “Judeo-French”.
2-Ran: Spanish verbs with first person singular in /oj/ have nothing to do with a post-verbal YO: it’s an innovation which began in Vulgar Latin where the first person of the verb “to be”, SUM, which phonologically was turing into SO (A form found in some Romance varieties) was re-shaped to a form */sujo/, thereby keeping a clear difference between a stem /su(j)/ and the regular first-person ending /o/ (This is also the etymon of French SUIS, which received its -S by analogy with verbs such as (JE) FINIS at a later date).
Subsequently, in Spanish this /oj/ was extended to first-person singulars stressed on the /o/ ending: hence DO, VO and ESTO became DOY, VOY and ESTOY (The former forms survived in at least some varieties of Judeo-Spanish, incidentally).
ESTOY owes its unusual (final) stress to the fact it was monosyllabic in Latin (STO), with /e/ appearing before the s + consonant cluster in order to eliminate a word-initial combination which Vulgar Latin disallowed (and indeed Spanish and Portuguese still do).
3- Trond Engen: I very much doubt the Norwegian and German initial /m/ first plural pronouns are “related”, in the sense that the innovation spread from one language to the other: in German the innovation didn’t even reach the Low German area, which would be the point where I would have expected it to spread to Norwegian.
4-Ran again: Eyak is very reminescent of the situation in the history of Cherokee. This Iroquoian language has a phoneme /m/ which arose as an allophone of Proto-Iroquoian */w/ followed by a nasal vowel (Proto-Iroquoian otherwise lacked an */m/ phoneme, and indeed the non-Cherokee (=Northern) Iroquoian languages still lack one, leaving aside in some languages loanwords from European languages): when some of these nasal vowels were de-nasalized in the transition from Proto-Iroquoian to Cherokee /m/ became a phoneme.
Etienne: Granted. I convinced myself that it was an internal development in Norwegian. Parallel changes in the plural nominative leading to homonymy with the singular dative in both languages is perfectly natural, since the paradigms are cognate.
@Etienne: Thanks for both explanations. I find your explanation of -oy more convincing than the one I’d read; and your explanation of estar‘s stress is so obvious that I don’t know why I hadn’t already realized it. (I don’t know Latin in general, but I did know that estar came from stare, ultimately cognate with English “stand”, so I really should have made that connection on my own.)
It’s interesting to see the fate of Proto-Romance epenthetic i > e before s-clusters:
French: absorbed and forgotten except as a circumflex where the s was (être), no longer used in loanwords.
Spanish, Portuguese: still required in both native words and loanwords (estop, estrés, e.g.)
Italian: still surviving in formal language after /n/ (in iSvizzera > in Svizzera change is nearly complete).
Romanian: either never existed, or no trace of it now (sta).
Jim: This is part of the reason English [maintained] its dominance in the American colonies/US despite slight majorities of Germans for long periods. [with regard to Stefan’s: Furthermore it should be taken into account that even German wasn’t ‘German’ prior to Johann Wolfgang von Goethe and Friedrich der Große but a patchwork of dialects, presumably more in number than the German principalities prior to Otto von Bismarck.]
Surely the English spoken in the early American colonies was itself just such a patchwork of dialects ?
I don’t know how that word “early” crept in. I mean “in the American colonies”.
Re: w ~ m alternations
Is English ‘with’ and German ‘mit’ a coincidence? The OED gives a Common Germanic *wider- in the etymology of ‘with,’ but I don’t read German and cannot determine what the etymology of ‘mit’ is.
Is English ‘with’ and German ‘mit’ a coincidence?
Yup. The German word is related to Greek μετά [meta].
Old English had “mid”, the expected match for German “mit.” The original of “with” meant “against”, as still in “withstand.” Just guessing, but I wonder if the semantic shift is due to Norse influence (as with “dream” and “bread”)? Icelandic has með for “with” but “við” seems to overlap a bit.
Mid/mit is old IE; cf Greek “meta”
English with, German mit
In spite of their common meaning and partial resemblance, these two words have a different origin.
With has more than one meaning: a positive one (in the company of someone, helped by a person or instrument) and an adversarial one (eg fighting against someone). The latter meaning is the original one, still present in withstand, literally ‘to stand against’. Note the ambiguity of with in the context of fighting: you can fight with (= against) an adversary but also fight with (= together with) a group, army, etc. Eventually the ‘together with’ meaning became the most prominent one in both social and instrumental contexts.
Before with had extended and eventually changed its original meaning, the word which meant ‘together with’ was mid, which is still present in midwife, the helpful woman assisting another woman while she gives birth. This is the cognate of German mit. The German cognate of with is the wid- in wider, a word which still means ‘against’.
I see that others have already provided the right cognates, while I was writing my own comment.
I don’t think my suggestion of Norse influence looks very plausible, on reflection. Just looking through examples of the use of Icelandic við, the core meaning looks very much like “against” shading into “alongside.” The cases where it’s translatable with “with” look more like cases where it’s English that happens to have kept an older sense of the word, as in marie-lucie’s “fight with”.
LH’s numerous Scandiphones will know best.
marie-lucie: what a neat connection between German wider and English “with” in the older meaning (that I did not know as such, as in “withstand”) ! The clue is there, but I had never thought about it: widerstehen = “resist, withstand”.
I can think of only a single situation in which German mitkämpfen means “fight with” in the sense of “fight against” – er kämpfte mit sich selbst. That’s merely because there is only one person involved, so that person play both roles ! Otherwise it’s always kämpfen gegen.
Are there any cognates of “with” outside Germanic? None immediately leaps to mind (not that that means a lot.)
I just knew it. No sooner do I make that claim about uniqueness, than countless further examples occur to me: Jakob kämpfte mit dem Engel, sie kämpfte mit einer Eßstörung (bulimia).
I would now say that kämpfen mit means “fight together [against …]” when referring to an actual physical fight against physical opponents. When the fight is metaphorical – not against a person, but against something harmful, such as an illness – then kämpfen mit means “struggle against” (“struggle with” !). It’s often used with reference to an individual person trying to come to terms with something adverse to him/her.
Is that a legit locution: “no sooner do I …, than ..” ??
David: The OED is silent on non-Germanic cognates of with, but Etymonline says “from PIE *wi-tero-, literally ‘more apart,’ suffixed form of root *wi- ‘separation’ (cognates: Sanskrit vi, Avestan vi- ‘asunder,’ Sanskrit vitaram ‘further, farther,’ Old Church Slavonic vutoru ‘other, second’).”
Here’s what the OED says anent semantics:
Such semantic loans, as they are called, from Norse are quite common in English: modern dream is a descendant, as far as its phonology is concerned, from Old English dréam ‘joy’, but its meaning is from Old Norse draumr. Similarly, the English earl owes his specially high rank (in pre-Conquest times, effectively a royal governorship) to the jarl, for eorl in Old English meant merely ‘noble’. (After the Conquest, each earl was associated with a county, but no longer in the king’s name; rather the king’s official was the directly appointed sheriff. The connection between the two is shown in Latin, where the earl was comes ‘companion’ > French comte > Eng. count, whereas the sheriff was vice-comes. Dukes and marquesses (= march earls) are later growths with French names.)
m-l: Most synonyms of fight also take with in the sense of opposition, such as quarrel with, struggle with, and even go to law with. Two more verbs with oppositional with- are withhold and withdraw lit, ‘drag away from’. There were once many others for which other prefixes have taken over, such as withsake ‘forsake’ (cf. Middle German widersachen) and withcall ‘recall’. For lagniappe, Etymonline gives the etymology of avec < avoc < Latin apud hoc ‘with that’.
@John Cowan:
Thanks. Thought you might know …
Looks like my first guess about the sense of “with” being partly influenced by Norse was right after all, then.
Doesn’t look like there are unequivocal cognates of “with” outside Germanic really. Why do the English and Norse forms show no trace of the -r-? And “further removed” is not really all that close to “against” semantically.
It occurred to me that if that were the case that there are no obvious candidates for the origin of “with” in Indoeuropean, “with/mit” might after all go back to the sort of spontaneous variation in Indoeuropean that presumably underlies the w/m variation in the 1pl verb forms. [Come to think of it, the Old Church Slavonic 1dual forms have -v-, like the Sanskrit, so there’s a w/m alternation in first person forms there too without bringing in Hittite, even.]
The two forms might represent different specialisations from an original central sense encompassing “alongside, in relation to, in comparison with, against”.
But the forms also differ in the vowel and presumably in original accent judging by the different working of Verner’s law in the two. *wéti versus *metá/metí? (Why is the vowel ‘i’ in English ‘mid’ and German ‘mit’ vs Norse ‘með’? Wouldn’t you expect -e- with an original following -a, as in the Norse?)
“The OED is silent on non-Germanic cognates of with . . .”
As is “The American Heritage Dictionary of Indo-European Roots.”
John: There were once many others for which other prefixes have taken over, such as withsake ‘forsake’ (cf. Middle German widersachen) and withcall ‘recall’.
This formulation was a little misleading, at least for me. Perhaps it’s only the punctuation ? In the lemmas in the OED that you seem to have checked, I failed to find a claim that “withsake” meant what the word “forsake” means today. That is how I understood your “such as withsake ‘forsake’ …”.
The OED lists the obsolete word “withsake” meaning “to contend, dispute, deny; cf. MHG. widersachen”, and indicates (“=”). that “withsay” was synonymous The meanings of the obsolete word “withsay” were “renounce (rare), contradict, resist”.
However, the lemma for “forsake” gives as an obsolete meaning “decline, refuse”. So the obsolete word “withsake” had a meaning similar to one which the non-obsolete word “forsake” once had, but has no more.
Are there any *other* IE cognates of mid/mit besides Greek ‘meta’?
To confuse the issue still further, Etymonline attributes the mid/mit etymon to IE *medh, as in middle, medium.
Ah. Just looked at an actual book (Hoffmann’s somewhat ancient [1950, my copy, inherited from grandfather] Etymologisches Wörterbuch des Griechischen) which duly links Greek ‘meta’ with OHG mit(i) and happily posits either IE metí or médhi as possibilities for the origin of the Germanic forms, saying it’s possibly (‘wohl’) connected with the ‘middle’ root of Latin medius, Greek mesos etc. Apart from this it just comes up with a few vague possibilities based on speculative analyses of Illyrian place names.
JC, thanks for the extra examples.
It looks like the many old words with prefixed with have replaced the prefix by another one, more semantically obvious to more modern ears.. I note withsay which seems to have been replaced with gainsay, in which the element gain is the stem of against.
Which reminds me that there are Aeolic and Doric forms ‘peda’ and ‘peta’ corresponding to Attic/Ionic ‘meta’. Just in case the matter wasn’t quite confusing enough already. Hoffmann’s book makes a valiant effort to ascribe them to *ped “foot” …
If Germanic mid/mit really does go with *medh(y) “middle”, it’s hard to shake the feeling that the close resemblance to Greek “meta” is then just coincidence, especially as the Greek forms would then have no obvious cognates elsewhere in IE and also show such bizarre dialectal differences within Greek itself.
It’s just occurred to me that mit/meta might be one of the things at the back of the strange theory of a Germanic prehellenic substratum that came up on LH the other day.
David: Anticipatory lowering, or a-umlaut (not to be confused with “a-umlaut” in the sense of the ordinary umlaut (i-umlaut) of /a/) is a sporadic process at best in Germanic. Some instances of it are wer, ON verr ‘man’ (now only in werewolf) cognate with Latin vir, and horn < PGmc hurnaz, ultimately < zero-grade *kr- of the IE root *ker¹-. This root is prolific in English, and underlies cervine, cornea, carrot (from its shape), carat (why a weight in Greek should be called a ‘little horn’ is a mystery to me, maybe also from its shape?), (rhino-)ceros, cerebrum, migraine/megrims < hemicrania, cheer < OF chiere ‘face’ ult. < καρα ‘head’, reindeer, rinderpest in various grades and derivations. As you can see, the meaning spread from ‘horn’ to ‘head’ generally, and to other parts of the head as well.
Stu: Quite right about forsake; post in haste, repent at leisure, as I am always saying to and about myself. “No sooner do I * than” is standard, as googling that quotation will show (Google interprets the word “*” as meaning “a reasonable number of intervening words”)
@John Cowan:
Thanks again.
In fact the Old High German variant form miti (which I just discovered) makes it look like anticipatory lowering wouldn’t have been an issue as the W Germanic form originally ended in -i not -a anyhow. Don’t know what to make of Norse með, except it confirms me in a longstanding view that Norse umlaut transcends human understanding.
A final -i does add plausibility to the idea that mit/mid goes back to *medh(y) “middle”, which seems strong evidence against a connexion with ‘meta.’ (And also utterly torpedoes my idea that with/mid could ever go back to a w/m doublet, as the -th of “with” cannot possibly go back to IE *dh, any more than the Greek t could.)
Reclaiming my Old Norse grammar from my daughter, who had half-inched it, I discover I was quite wrong about there being no trace of -r- in the “with” form: viðr occurs beside við, in fact. For good measure, meðr occurs beside með too …
The vowel of með, moreover, doesn’t rule out a stem in meðj-, judging by the form beðr “bed”, which is a ja-stem, plural beðjar etc. (“Middle” *is* miðr, though.)
So “with” is indeed an apocopated version of the form which turns up in German “wider”, and mid/mit looks very likely to be related to “middle” etc, and hence probably *not* to Greek ‘meta’ after all despite the seductive similarity of sound (and meaning.) Which also means they are undoubtedly unrelated, even if one were prepared to posit a w/m initial variation.
Also means that my scepticism of a possible connexion of “with” and (say) второй was premature. The phonology works, anyhow, though I still think the semantics are a bit of a stretch.
Nonsense about beðr, sorry. The -e- in that word will be umlauted from -a-.
Still much less happy with the idea that mit/mid is related to Greek ‘meta’ than I was. For it to work, the peculiar Greek dialect forms have to be set aside as irrelevant; and then, assumingly the only plausible IE cognates of ‘meta’ are Germanic mid/mit/með you’ve got the difficulty that those have a very plausible looking altermative origin in *medh which is unfortunately incompatible with them being related regularly to any of the Greek forms (which would have to have been *metha)
Etymonline’s entry for “meta-” posits PIE *me- “in the middle”, which I suppose has the advantage of unifying me(dhy) and me(ta) but seems rather ad hoc. Given that the vowel -e- is just the basic IE default vowel this also means the for essentially boils down to just m-, which looks fishy to me. I’m very far from expert in these matters, though.
Many thanks for pointing me towards Etymonline, by the way, John Cowan. Hours of pleasure lie ahead …
There are cases in biology that get about this complicated and can still be unraveled. In human evolution, first there’s our split from the ancestors of the chimpanzees + bonobos that took about a million years; in much more recent times, Homo s. sapiens has interbred with Neandertalers, Denisovans and an otherwise altogether unknown lineage, and Neandertalers and Denisovans have interbred with each other as well.
That word only extends to the northern parts of Central German, not to Bavaria.
Not true for German either.
The article talks about Germanic languages, so it’s making a claim about Icelandic… which retains the dual. Equally irrelevantly, the 2nd person plural pronouns in Bavarian dialects are generally derived from the poorly attested Old (not Middle) High German 2nd person dual pronouns; and the 2nd person plural verb ending is usually explained as the usual one fused to a clitic, but it’s identical to the Gothic 2nd person dual verb ending as well… Otherwise, most kinds of German retain the word “both”, the dual of “all”. (Some, like my dialect, have lost it and resort to “all two” just like French does.)
Almost all of them; I think the only exceptions are Walser dialects.
In any case not Upper German, because of -ee- instead of -ei-.
Pretty much any after the Frankish immigration of the 9th century that has left traces in today’s dialect.
*lightbulb moment*
I’m not aware of any evidence that that was actually the case.
It is, however, universal in Low German and was a famous feature of the dialect of Berlin (just barely Central German).
Possible; and indeed some or all of the other Walser dialects have such endings as well. But in at least one Lower Bavarian dialect, the 1st person plural clitic -/mɐ/ has become a full-fledged ending, too.
Could this */sujo/ be composed of */so/ and */jo/?
In one of them, Marie was borrowed as Onwari (with a nasal vowel and probably a velar rather than labiovelar approximant).
…German Traum means “dream”, though.
Unstressed */e/ became */i/, followed by various applications of analogy. Old Norse had both stressed ek and unstressed ik for “I”, while West Germanic has settled on the latter.
That’s “probably” rather than “possibly”, though the degree of confidence expressed this way can be quite low.
Coincidentally, German wider is obsolete as an independent word and has been replaced by gegen… “contradict” is still widersprechen, though.
…Oh. Could be.
*lightbulb moment* So German Rind “cattle” (singular) is cognate with horned? That makes sense!
Oh, actually…
Worse: PIE roots are not supposed to end in a vowel or to contain only one consonant. And you can’t simply speculate that there was *h1 at the end, because that would have left a long vowel in later descendants.
@David: German Rind, English rother, OE. hryðer. So yeah.
Oh! So that’s what Rotherwood means! 🙂
And Rothermere.
Philologos at the Forward has written three columns on these articles. Much is a restatement of the original authors’ writing, but he contributes plenty of his own thoughts. The comments are worth looking at too.
Origins of Yiddish Are Anything But Understood
The Origins of Yiddish: Part Tsvey
The Origins of Yiddish, Part Dray
Meh, hardly anything new.
Meh but not feh.
Another interesting interchange is /m/ ~ /bh/. The dative-ablative plural ending in IE began life ending in /m/, whereas the instrumental ended in /bh/, but when the forms merged, the Western IE languages specialized in /m/, whereas the Eastern ones (broadly construed, and including Italic) specialized in /bh/.
Are you sure Western-Eastern is the distinction? Celtic has /bh/ and Slavic has /m/.
Yes, I was wrong to make it a Western-Eastern split. It’s /m/ in Balto-Slavic and Germanic, /bh/ in Italic, Celtic, and Indo-Iranian.
A good explanation, but not the best one. (Annoyingly, this book chapter is scanned sideways.)
Yeah, I read that article, and I think it’s special pleading.
Most PDF readers have options to rotate the pages, fortunately.
Thanks for introducing me to the concept of relexification.
Now it’s all quite clear.
French is relexified Gaulish. Russian is relexified Finnish. And English, I suppose, is relexified Welsh.
Why?
The difficulty with the case-merger account is that not all the languages did merge inst. pl. with dat./abl. pl. If the latter ending had *-m-, why does it have -bh- in Sanskrit, where it remains distinct from the former? Similarly in Balto-Slavic, where the endings remain distinct but both have -m-. It’s hard to see why the endings would contaminate each other in this way.
I like the Hill article. For the price of two sound changes (loss of intervocalic *-m- in inflectional endings, and the dissimilation *nm > *nbh) it accounts for this -m-/-bh- divergence as well as the 1sg. ending -o:, which has always stuck out like a sore thumb from the rest of the verb endings. The first change would be morphologically conditioned in being confined to inflectional endings, but such things happen. There must be more evidence relevant to the second change than he adduces in the article, though.
John C.: Most PDF readers have options to rotate the pages, fortunately.
I put my laptop computer sideways on the table and scrolled from left to right through the paper. Much better than the usual way. I’m not even joking.
TR: I like the Hill article.
Me too. Also, might even Greek ὰμφί < IE *h₂ṇt-bʰi- be a fossilized instrumental meaning “with an opposing side”? Same for Latin, I think. Not so sure about Germanic.
See how i took the effort to pick out symbols there? So don’t come saying it’s the wrong accented alpha or something.
OK, I won’t. But it’s the wrong accented alpha.
Isn’t that what the Hill article says about ἀμφί?
Though come to think of it, it seems like a weakness of Hill’s theory that the putatively secondary forms (those with -bh-) should have been the ones to be generalized in most of the languages, and out of an originally quite limited distribution, at that (after n-stems).
I do the same in such cases. I almost never bother to log in (why should Google get to track all this?!?) and download papers from academia.edu; after all, I read the vast majority just once.
I’ve also encountered this etymology elsewhere.
But it’s the wrong accented alpha
Sad as it is. I saw it too late.
Isn’t that what the Hill article says about ἀμφί?
Not that I noticed. When ἀμφί first occurs, it has the traditional etymology “of two fronts”. The fossilized instrumentals are listed on page 197, and I can’t see that he goes back to readdress ἀμφί. But it seems like a natural thing to do after identifying the morpheme φί (I mean, I did, didn’t I?). Maybe he’s suggested it in a later paper.
Oh, you’re right. He discusses it on 183-4 but doesn’t derive it from an instrumental *h₂ṇt-bʰi because the semantics are difficult. Which seems cogent — how do you get from “with/by means of a face/surface” to “on both sides”? In any case the *h₂ṇt- part seems uncontroversial, whatever the suffix may be.
Well. I think I’d have to suggest that the semantics of *h₂ṇt- was “opposing (face)” rather than “face”. This would at least fit the etymology of the Germanic “other” and “and” words. “With an opposing face” seems quite similar in its loose instrumentality to his example “outside” (<- “by the door”).
It still strikes me as a bit of a leap from “with an opposing face” to “on both sides (of)”, but who knows. The competing theory he outlines, based on a dual form, does seem rather clunky.
Not “face”: “forehead”. Apparently preserved as such in Hittite.
Sure, but it must have gone through a sense like “front” to give the meanings of ἀντί etc.
Also, aṇti- “against, before. opposite”. The “opposing face” semantics of *h₂ṇt- seems pretty well established without my help. A straightforward semantic path would be “with an opposing face” -> “with its opposite” -> “from both (or all) sides” -> “on both (or all) sides”.
Also, why isn’t *h₂ṇt- regularly derived from *h₂eṇ- of ana-, something like “on/toward-ness”?
.Also, The AHD IE Roots Appendix is back. If that’s old news, I haven’t noticed.
Also, I forgot to replace the dotted n,
Also, why isn’t *h₂ṇt- regularly derived from *h₂eṇ- of ana-, something like “on/toward-ness”?
Interesting idea… The semantics seem plausible enough, but are there analogues for a derivational nominal suffix *-t-?
I meant to keep spinning on this.
There’s definitely a *-t- used to form nominal suffixes. Wiktionary lists the state-of-being suffixes *-teh₂ and *-teh₂ts, the action noun suffixes *-tus and *-tis, the agent suffixes *-tér and *-to:r, and hence the instrumental suffix *-trom. There’s also a *-t- used for adjectives in *-tós and (rather unsystematically, I think) the comparative suffix *-tero-.
I don’t think it’s controversial to suggest that these t’s are all one and the same. I’d guess that the basic adjectival derivation came first, with the nominalizations following from the adjective. *-teh₂ would simply be a *-t- adjective + the abstract suffix *-eh₂ “-y-ness”. I think the different reflexes of *h₂eṇt- point to the suffix *-tis.
(I’ll not even add a caveat. I see it’s become eponymous.)
Sure, it’s plausible enough that all those *-t-‘s are related, but if you’re positing a bare nominal suffix *-t it would be good to be able to show it elsewhere on its own. But maybe you’re not:
I think the different reflexes of *h₂eṇt- point to the suffix *-tis.
What makes you say that? The Hittite noun is hant-; the -i in forms like Gk. ἀντί is originally the locative ending, not part of the stem.
Also, The AHD IE Roots Appendix is back
And now with laryngeals!
I don’t think a pure nominal suffix *-t- is necessary. Is there a meaningful difference between noun and adjective in PIE?
What makes you say that?
Ignorance. That and surprise. I expected to see some evidence of independent derivations in different languages and was surprised to find similar vowels in the suffix.
A root upgrade.
Not much, but the Greek pattern of deriving nouns from adjectives by moving the stress to the first syllable does seem to go back to PIE or thereabouts.
And yes, anti (initial stress in Sanskrit, final stress in Greek) is “on the forehead”, quite literally “in front of”.
deriving nouns from adjectives by moving the stress to the first syllable.
Right. Of course. and it seems to have been a very productive derivational pattern — if not a regular or even paradigmatic relationship. If you had an adjective you had the noun available.
And yes, anti (initial stress in Sanskrit, final stress in Greek) is “on the forehead”, quite literally “in front of”.
One problem with that is the diverse literal meanings of the reflexes. It seems simpler to assume that anti always meant “in front of”, and that *ant- (or derivations thereof) was nominalized to “chest”, “forehead” and “outer edge” independently.
final stress in Greek
Orthographically yes, but this may not represent anything real; almost all prepositions are conventionally written with final accent in Greek (except in anastrophe), but I suspect this might be a meaningless convention and that they were simply proclitic.
What reflex of *h2ent- means “chest”?
Lith. antis, if I am to believe Bjorvand & Lindeman.
Lith. añtis is from a derivative *h2ent-io- (thus Lubotsky, Etym. Dict. of Proto-Germanic s.v. *anþja-), which in turn looks like it’s formed from the locatival adverb in *-i — so it would represent a secondary development from “in front” to “(part) in front, chest”.
Ah, good to know…
One problem with that is the diverse literal meanings of the reflexes. It seems simpler to assume that anti always meant “in front of”, and that *ant- (or derivations thereof) was nominalized to “chest”, “forehead” and “outer edge” independently.
It may have meant simply “front side of the body”. The pathway “body part” + local case ending > preposition is attested many times over.
True. My beef was strictly with precisely “on the forehead” as the source of the Greek preposition.
I may be pushing this beyond what’s entertaining, but whether it meant “forehead” or “front side of the body”, it’s an etymon in search of a further etymology. Like “the towardy”. (Though, admittedly, ‘front’ itself is from a meaning “forehead” with no certain etymology.)
Aren’t all etymons (apart from such things as onomatopoeias) always in search of a further etymology? I don’t see what’s particularly problematic about this word.
In favor of Trond’s *h₂en > *h₂ent- idea, it turns out there are some examples of a nominal derivational suffix -t-: e.g. Latin dōs, dōtis “dowry” (*deh₃- “give”), nox, noctis “night” (cf. Hitt. nekuzzi “get dark”). (Also a larger set of examples in compounds, which are less relevant.) So the semantics and form both look like a good fit. I wonder if the connection has been suggested anywhere.
What is PIE *nokʷts “night” supposed to be derived from? Wiktionary says “Pooth (2015:1, 29) argues that the word derives from *negʷ- (“bare, naked”)” but that seems a bit of a stretch semantically. It says Latin dōs comes from PIE *déh₃tis, which just has the -tis suffix that was already mentioned earlier.
What is PIE *nokʷts “night” supposed to be derived from?
Well, Hittite nekuzzi “be night” could represent PIE *negʷ- or *negʷʰ-. There are some Greek cognates which lack the -t- formant and seem to point to an aspirate, e.g. νύχιος “nightly, by night”.
It says Latin dōs comes from PIE *déh₃tis
This could be true as the -i- of -ti-stems was often syncopated in Latin. But the same word appears in Greek as δώς, which if it’s an exact cognate rules out -ti-; unfortunately the Greek word is only attested in the nominative, so we can’t be totally sure it really was a t-stem.
Thanks for that. Couldn’t both *déh₃tis and *negʷtis reasonably be nominalizations of adjectives?
I’m not put off by the semantic leap from “nakedness” to “night”. The enemy came upon them in the middle of their nakedness.
They had wagons and chariots with wheels and axles, and they slept naked?
Can’t argue with the evidence.
I have an automobile with wheels and axles, and I sleep naked.
When I sleep behind the wheel, I’m rarely naked.
I have neither wheels nor axles, and I sleep in my underwear.
Couldn’t both *déh₃tis and *negʷtis reasonably be nominalizations of adjectives?
That should be *nogʷʰts and (probably) *deh₃ts, I think. Why would you want them to be nominalizations of adjectives rather than simply nouns formed from a verbal root?
Yes, thanks. But aspiration?
The adjectival *t is there, *deh₃ts is said to have followed this path, and ‘naked’ is an adjective, so it seemed like a reasonable assumption.
Some of the Greek cognates, as I said above, contain an aspirate. That would seem to rule out the “naked” connection, but I haven’t read the Pooth article Eli cited.
I don’t think *t is specifically adjectival. *deh₃ts comes straight from the verb root, “give” > “thing given”.
Belatedly:
1) mit and Mitte can be explained in every phonetic detail. First, the former is from the latter, which is why it has a short vowel instead of having undergone the lengthening of vowels in monosyllabic words that don’t end in too many consonants. In other words, etymologically it would be spelled mitt. Second, the tt is from West Germanic consonant lengthening: PGmc. *ðj > PWGmc. *ddj > High German tt. Third, before the *j was lost (medius, mesos etc. etc.), it triggered umlaut of *e to *i.
2) There are Walser dialects (possibly all of them, I don’t know) where the 2pl nominative pronoun is /dir/ just like the 2sg dative. This is explained as the 2pl verb ending followed by ihr. Stranger things have happened, like the 2sg nominative clitic /st/ in Bavarian…
3)
My point there was to ask (more or less jokingly) if we can add that to the stereotypical list of things that reconstructed PIE tells us about the lifestyle of its speakers.
Ah. I took it to be an ironic contrast, on the order of “They can send spaceships to Mars, and they don’t understand simple mathematics?”
So now you’re telling me the Indo-Europeans sent spaceships to Mars? The field has changed so much since I was studying it…
List of Indo-European spaceships which landed on Mars:
Mars-3, Viking-1, Viking-2, Mars Pathfinder, Beagle 2, Spirit, Opportunity, Phoenix and Curiosity.
Bullshit. I misremembered my sources, possibly two chapters of the same Google book. 2pl dir is Luxemburgish, as found in several examples in this paper (in German); I think it’s spelled out as a footnote in another work by the same author.
Here’s a precis masquerading as a review by Alec (Leyzer) Burko of Alexander Beider’s 600-page work The Origins of Yiddish Dialects (2015). Beider thinks the direct ancestor of Yiddish is Bohemian German, which was a koine of at least four German varieties. The exception is westermost Yiddish (he draws the line west of the standard Western-Eastern line), which he traces to East Franconian. He definitely disposes of Wexler’s ideas of the origins of Yiddish, but agrees that there may be a variety of substrates underlying it. Well worth reading.
But what really caught my eye was footnote 1: “I thank the editors of In geveb [the journal] for this rare opportunity to publish in my native tongue.” I thought, what? What language does he normally publish in? German? Yiddish? Seems unlikely. But then my eye fell on this sentence in the main text, “To be sure, though this argument dismisses much of Weinreich’s Rhineland theory, there is what to be salvaged from it”, and I realized that by his native tongue he means Yiddish-flavored English! It’s not heavy-handed, but it’s definitely there.
Nice!
!!!
The truncation of etwas to was is very widespread in German, my dialect has it, and so do Viennese dia- and mesolect. But I thought Yiddish specifically lacked that, instead fusing */tw/ to /p/? Does that vary within Yiddish, like so many other things?
From there:
I thought that was Beider’s idea all along – that Jews in Prague shifted from Judeo-Czech to Prague German? Why would EY descend from any of the dialects on Bohemia’s mountain rim?
Komín is evidently from German Kamin (final stress), itself clearly a cognate of chimney through Romance.
But that does not reduce how diagnostic the word is. All the words in this list are regionally specific in German – in this case, the form I’m used to is Rauchfangkehrer!
The dialect sample from Landskron, BTW, is the same kind of confusing mixture to me as EY is. Some words sound exactly like my dialect, others exactly like Standard German (though that may be exaggerated because the speaker appears to be reading example sentences from the Deutscher Sprachatlas), non-final /b d g/ (especially /g/, it seems) are usually voiced like in EY but not always, then there’s the merger of /eː oː/ into /iː uː/ which I just learned was a feature of Bohemia’s north rim…
@David Marjanović: In places where Jews were in regular contact with a German-speaking majority, their Yiddish naturally picked up features of the local characteristic German (particularly toward the south, where the Jewish and gentile dialects were already more similar). So while it was eppes in the Russian Empire, where most of my Jewish ancestors came from, was was fine further west. There was a Viennese bakery run by an elderly Jewish woman who got out in the 1930s that we used to go to when I was a kid, and as I recall, she said was and/or what like that. Of course, I don’t know whether she would have considered her native language to have been Yiddish or Austrian German. Among highly assimilated German/Austrian/Bohemian Jews (like Hannah Arendt), it was common to claim that they truly spoke German, not Yiddish.
Overall, today’s well documented Yiddish is probably significantly less diverse than it was historically. Personally, I regret that I didn’t start learning German until after all my direct ancestors who grew up in Yiddish-speaking households had died. It would have been interesting to see what my grandfather or great grandfather remembered of real pre-Second World War Yiddish.
Googling for “we had what to eat”, I find ~15 ghits, one from a Romanian holocaust survivor and one from a survivor of early-20C pogroms in Lithuania (the part that is modern Belarus). Both of these areas are deep in Eastern Yiddish territory. So even people who said epes apparently learned to say what when speaking English. Note that the phrase is habitual/consuetudinal in aspect, whereas “we had something to eat” can also be punctual; I would translate it into my own English as “we (almost) always had something to eat”.
Overall, today’s well documented Yiddish is probably significantly less diverse than it was historically.
No doubt about that! The western varieties in particular were moribund by the 1880s as Jews assimilated into France, Germany, and the Netherlands and learned to speak (probably) only slightly Jewish-flavored varieties of those languages. It also turns out that far to the east of the Pale of Settlement there existed from the 1830s to the 1980s a substantial group of Yiddish-speaking Jews in Udmurtia. Unfortunately, I know nothing about their Yiddish except for the unsurprisingly large number of Udmurt and Tatar loans.
In places where Jews were in regular contact with a German-speaking majority, their Yiddish naturally picked up features of the local characteristic German
Not just German. Occasionally, Slavic as well? Alexander Beider explained that the phonetics of the Litvak Yiddish are lifted from Polish dialects of Mazovia (where the Jews were especially few in numbers when they moved, or were expelled to, Lithuania in the 1500s). Since the DNA tells us that the early founders disproportionately took local wives, then the hypothesis may be that the outsize influence of the Mazovian Polish may have been mediated by intermarriage…
The strongly reduced vowel system of Generic Eastern Yiddish, with syllable-final [j] instead of diphthongs, and the consistent voicing of non-final /b d g/ are obviously Slavic.
List of Indo-European spaceships which landed on Mars:
There is now a Trans-Himalayan spaceship on the Moon.