Corpus of Spoken Yiddish in Europe

When I started writing my CSYE Glosses articles about ניט nit and נישט nisht ‘not’ (Part 1 and Part 2), I initially gathered the data by hand, searching each testimony for the words in question. Then I mapped the results. I enjoyed the process, but it was cumbersome. When I sent my first draft of the first article to CSYE Director Isaac L. Bleaman, he quickly figured out a way to automate the process. The result was the Word Maps tool, which has become a favorite of mine as a lover of maps and dialects. The tool allows users to create maps for one or two words or phrases, showing the birthplaces of the speakers that used these terms and, by color, which of the two prevails where. If a speaker uses both forms, the color indicates the proportion. In this article, I will go through a few examples that show interesting geographic patterns.¹

North/South Shibboleths

Some words only occur in speakers of Northeastern Yiddish. A good example of this is האָלט האָבן holt hobn ‘to love.’ A search for holt reveals a neat cluster stretching from near Lomzhe (Łomża) north to Rige (Riga) and east to Minsk. The particularly Northeastern participles געװעלט gevelt ‘wanted’ and געלאָזן gelozn ‘allowed’ show a similar distribution, albeit with an outlier in Warsaw in the case of the latter representing one Varshever survivor who was trying to speak Northeastern Yiddish. Conversely, the past participle געטון getun ‘did’ is exclusively found in the two southern dialects, Central and Southeastern Yiddish, where it is pronounced /gətin/.

Comparing the distribution of זײַנען zaynen and זענען zenen ‘are,’ we see that this pretty much exactly follows the boundary between Northeastern Yiddish and the southern dialects. Similarly, the alternative pronunciations פֿינף finf and פֿינעף finef ‘five’ break down along this north/south border. Furthermore, while אומעטום umetum ‘everywhere’ is found, well, everywhere, the alternate form אומעדום umedum is found throughout the southern dialects but never in Northeastern Yiddish.

There are many instances of vocabulary patterns that follow this north/south division, showing what a fundamental linguistic divide this is within Eastern Yiddish. And it is all the more noteworthy to see lexical patterns following this dialect division, because it is usually defined by differences in vowels, not words. But this is not the only such pattern of lexical distribution, as we will see.

East and West

It is not just northern and southern distinctions that show up in the corpus. Some are east/west. For instance, the participle געבראַכט gebrakht ‘brought’ prevails in the east, while געברענגט gebrengt is the western form, predominating in the Central Yiddish areas of Poland, Galicia, and Hungary. Another east/west division is between דערמאַנען dermanen and דערמאָנען dermonen ‘remember.’ Here, though, the southern part of the boundary between the two doesn’t neatly fit that between Central and Southeastern Yiddish, and indeed to call it a boundary is misleading; there are areas of overlap. Furthermore, the distinction runs through Northeastern Yiddish, with dermanen turning up in a cluster in western Lithuania. A similar distribution occurs with מסתּמ(א) mistam(e) and mistom(e) ‘probably,’² again with the former showing up not just in Central Yiddish and the western parts of Southeastern Yiddish but also in western Lithuania. The terms שטיגן shtign and טרעפּ trep ‘stairs’ also have an east/west split, with the former found in most of Poland and Hungary — though this is not simply a matter of Central versus Southeastern Yiddish, as one Galician survivor uses trep while a Bukoviner survivor uses shtign. גיך gikh and שנעל shnel ‘fast’ also have an east-west distribution, the former more eastern and the latter more western — but again, not simply Central versus Southeastern and Northeastern, as shnel crops up in parts of Southeastern Yiddish and gikh in the southern part of Central Yiddish. Overall then, there are numerous examples of east/west lexical divisions, but unlike the north/south ones, they don’t follow the classical dialect divisions, and they don’t necessarily coincide with one another.

You Say Potato

The famously rich Yiddish vocabulary for ‘potatoes’ is only partially represented in our corpus. קאַרטאָפֿל kartofl is clearly the predominant term, found everywhere. בולבע bulbe is restricted to Northeastern Yiddish, which makes sense, given its apparent origin in Lithuanian bulvė. In fact, in western Lithuania, we find the regional form בולװע bulve. Another regional variant that shows up is קאַרטאָפֿליע kartoflye, found in Moldova, as well as קרומפּיר krumpir in the region known as the Unterland, which is the area around where Hungary, Romania, Ukraine, and Slovakia meet. One more exotic term is בראַמבער bramber, which one survivor from Loytsk (Lutsk, Ukraine) uses in the plural: bramberes. Dovid Braun pointed out to me that the Czech word for potato is brambor; this would seem like an unlikely thing to show up in northern Ukraine, but Lutsk was surrounded by several villages settled by Czech colonists in the eighteenth and nineteenth centuries. In fact, this survivor speaks some Czech.

More in Store

It is fairly well known that each of the three Eastern Yiddish dialects has a different word for ‘store’: קראָם krom in Northeastern, געװעלב gevelb in Central, and קלײט kleyt in Southeastern. Our data reflects this three-way division nicely, with the added detail that kleyt is used by four speakers from western Lithuania, far from Southeastern Yiddish. But on closer inspection, two of these unmistakably mean ‘granary,’ not ‘store’. Again, while the traditional division of the three Eastern Yiddish dialects is based on their vowels, there are other linguistic patterns that reflect the same threefold division.

Relative Confusion

Not every lexical choice I found showed an obvious geographic component when I mapped it. I couldn’t see any pattern to the family terms שװעסטערקינד shvesterkind versus קוזינ(ע) kuzin(e) ‘cousin,’ מומע mume versus טאַנטע tante ‘aunt’, and פֿעטער feter versus אָנקל onkl ‘uncle.’ There may in fact be a regional pattern here, but it would require a more sophisticated analysis to uncover. Similarly, the Word Maps tool cannot easily reveal whether those who said shvesterkind were likelier to say mume and feter or those who say kuzin(e) likelier to say tante and onkl, although I suspect this is the case. װײַב/פֿרױ vayb/froy ‘woman’ or ‘wife’ also doesn’t immediately seem geographic in its distribution. 

Overall

One last example seems worth exploring. The word איבעראַל iberal ‘everywhere’ is found, appropriately enough, throughout Eastern Yiddish, but the variant form איבעראָל iberol, pronounced /ibərul/, is only found in the Unterland. Overall, the Unterland stands out for the number of words found only here, of which this is just one instance. This is one of several spatial patterns that the Word Maps tool, using the CSYE data, reveals. Aside from the distinctiveness of the Unterland, there are two overarching geographical tendencies in Eastern Yiddish. One is a stark north/south divide, already familiar in Yiddish dialectology. The other is a continuum marked by a fan-like distribution of divisions radiating across Southeastern Yiddish, revealing a clinal change of decreasing similarity to Central Yiddish as one moves eastward. Perhaps more patterns will emerge as the dataset grows and more users start exploring it through tools like Word Maps.

Notes