Unity of the German component of Yiddish: myth or reality

DOI 10.1515/ijsl-2013-0077 IJSL 2014; 226: 101 – 136 Alexander Beider Unity of the German component of Yiddish: myth or reality? To Erika Timm1 Abstract: The article deals with the question of the unity of the German component of Yiddish. Before the 16th century, the languages used in works compiled by various Jewish authors from western German-speaking provinces show close similarities to dialects spoken by local Christians and do not represent texts written in a single speciically Jewish language. Texts dating from the 16th to 17th centuries demonstrate the existence of two separate Jewish idioms: western and eastern. The former covers western Germany and northern Italy and is mainly based on East Franconian and Swabian. The latter characterizes works written in Bohemia and Poland. It is closely related to Bohemian colonial dialect of German. It is inappropriate to consider all varieties of modern Yiddish to be dialects of one single language. Indeed, the analysis shows that in many aspects, Southwestern Yiddish inherits features of East Franconian, while Eastern Yiddish is primary based on Bohemian. Its consonantal system was later adapted to the Silesian dialect spoken by German Christian urban population in Polish towns. These two Yiddish idioms inherit numerous features from the two languages, western and eastern, respectively, that existed during the 16th to 17th centuries. As a result, as a whole, Yiddish is not descending from any hypothetical ProtoYiddish. Keywords: Yiddish, German dialects, phonology, Ashkenazic Jews, historical linguistics Alexander Beider: Paris. E-mail: [email protected] 1 This article could not be written without Erika Timm. Firstly, her works represent a basis for numerous elements of the analysis presented here. Secondly, during the preparation of this text, she shared with me copies of numerous quite helpful sources and answered to many questions of mine. 102 Alexander Beider 1 Introduction Given the wide range, in time and space, of various forms of speech and writing in Hebrew letters fundamentally derived from medieval German, how sure are we that all these diferent kinds of Jewish German or Judeo-German, from the 13th to the 21st century, from Alsace to Eastern Ukraine, are really one and the same Yiddish? This is what we would normally assume when speaking of Yiddish dialects, rather than Yiddish languages. Without this assumption, it makes no sense to talk of all these diferent linguistic systems as Yiddish or even to talk of a single ield of Yiddish linguistics. The question of the unity of Yiddish has two dimensions: chronological and geographical.2 The irst dimension deals with the question of whether the various earlier forms of the Jewish speech based on German are related to modern Western Yiddish (WY) and Eastern Yiddish (EY).3 The chronological dimension of the question of the unity of Yiddish was marginal for both Weinreich and Birnbaum, the classical authors in the domain of Yiddish studies. Due to their rather broad, and in a large extent extra-linguistic, deinition of the term Yiddish, the language of various texts written in the past by Jews using Hebrew letters in their vernacular German-based language automatically receives for them the label Yiddish.4 The second, geographical, dimension deals with the question of whether all the dialects of modern (that is, late 18th to 20th centuries) Yiddish, spread over much of Europe, form a single language, derived from a single source. The geographical dimension of the question of the unity of Yiddish thus concerns the origin of WY and EY and their sub-dialects. Several works show the unity of the Hebrew-Aramaic component of WY and EY. For the German component, in all aspects by far the most important in Yiddish, the situation is much more complicated and far from being consensual. Quite logically, the question of unity here is related to several other questions. Among them are: 1. To what German dialects are WY and EY akin? 2. How to distinguish Yiddish traits that resulted from internal innovations from those inherited from some German dialects? 3. How to distinguish features that one of the two major subdivisions of modern Yiddish borrowed from another from those that both of them inherited from their ancestor(s)? 2 This formulation is due to Alexis Manaster Ramer (personal communication). 3 For terms basic for German and Yiddish studies, this paper also uses the following abbreviations: MHG (Middle High German), NHG (New High German), OHG (Old High German), StY (Standard Yiddish). 4 See the discussion of their views in Beider (2004: 209–213). Unity of the German component of Yiddish 103 4. How to distinguish elements present in EY or WY since their proto-period(s) from those that were acquired later under the inluence of German dialects spoken by coterritorial Christians? Globally speaking, any attempt to provide an answer to any of these questions faces serious methodological issues. Maybe, it is for these reasons that no one has ever tried to address all of them. Generally, Yiddish scholars either avoid/ ignore all of them, or try to suggest a very general answer to some of these questions, principally the irst one. For example, Landau, who is oten considered to have written the irst scholarly studies of Yiddish, considered Yiddish to result from a mixture of various German dialects (Landau 1895: 58). The global position by Weinreich and Birnbaum in this domain is somewhat similar to that of Landau. Speaking about the genesis of Yiddish, Weinreich also refuses to give preference to any of German dialects considering it to be a total mixture of them. When considering them separately from each other, he only points to a few phonetic EY forms that he links to dialects of northwestern Germany, relates one major vocalic feature of WY to East Franconian, links a few words to Bavarian – some of them being limited to Central Yiddish – and a few others to East Central German. At the same time, he severely criticizes the opinion of Gerzon (see below) who suggests the existence of a genetic link between EY and East Central German.5 Weinreich (1973: Vol. 2, 94–110) also makes emphasis on the internal Yiddish innovation. He considers – without providing any strong argument – that this process was more important for the development of Yiddish than the inluence of German dialects. Moreover, Weinreich stresses that numerous German regional items could have come into EY through the mediation of WY rather than directly from German dialects. Birnbaum (1979: 72–76) provides a list of a dozen characteristics of EY (and, more precisely, only its western part: Central Yiddish), mainly phonological or lexical, that – according to him – are related either to Upper or Central German features. He postulates that Jews principally came to Poland from the areas of Bavarian (sub-dialect of Upper German), Thuringian and Upper Saxonian (subdialects of East Central German) and, as a result, the German component of EY is to a large extent a mixture of these dialects. This idea about the dialectal mixture that took place in Poland contradicts to his own consideration of Yiddish being about one thousand years old. For Bin-Nun (1973: 33–37), globally speaking, (Eastern) Yiddish realized a unique synthesis of diferent German dialects, taking various items from 5 Inaccuracy of all four linguistic features proposed by Weinreich as distancing ECG from Yiddish is shown in Beider (2010: 68, footnote 43). 104 Alexander Beider independent German sources, retaining some of their archaic features, and making a number of innovations. According to him (1973: 48), the fact that a Germanbased language became the vernacular tongue for East European Jews can be explained by three principal reasons: (1) massive migrations of German Jews to Poland; (2) the linguistic dominance in Polish towns of numerous German Christian colonists (speaking a dialect related to East Central German) that lasted during several centuries; (3) regular trade contacts and exchanges of students and teachers between Polish Jews and their coreligionists from German-speaking countries. Bin-Nun also addresses the question of where stressed protodiphthongs of WY and EY could appear. He states that according to the relexes of MHG î, ei, û and ou, WY is related to Bavarian (understood in his text as including Northwestern Bohemian and East Franconian, both heavily inluenced by Bavarian), while Central Yiddish, a sub-dialect of EY, is mainly related to East Central German. To the latter, he also links consonants of proto-EY saying that, generally speaking, the consonantism of Yiddish is similar to that of surrounding German dialects, while the vocalism stays apart: it evolved following internal Jewish innovations (Bin-Nun 1973: 79, 183, 209, 229, 255, 323, 326). Several other comparative studies of the German component of WY and EY, also end up with the idea of the existence of independent sources for WY and EY, and hence the polygenesis of Yiddish as a whole. Blosen (1986) bases his analysis on the comparison of the geographic distribution in, on the one side, modern German dialects, and, on the other side, WY and EY, of one morphological element (diminutive suix) and several phonological features: monophthongization of phonemes related to MHG ie and uo; diphthongization of those related to MHG î and û; unrounding of formerly rounded vowels; relexes for MHG ei; the consonants [p], [f] or the africate [pf] in various word environments. Taking into account some additional extra-linguistic data (such as the historical importance of Jewish communities in the past in certain geographic areas), Blosen comes to the following conclusions: WY appeared in the Hessian area, while EY is related to Silesian colonial German dialect. In comparison to studies by Bin-Nun and Blosen, Beider (2010) introduces two major additional elements. Firstly, the analysis does not take into consideration modern dialects only. The comparison between German and Jewish dialects is made using the historical data obtained by German linguists. Secondly, the paper deals not with separate phonemes but constructs entire systems of stressed proto-vowels. It comes to the following conclusion: for WY and EY, these systems could not have had one common ancestor. Stressed vowels of proto-EY are likely to have originated in Central Europe, in territories where the Gentile population was mixed and consisted of both German- and Slavic-speaking people. They are primarily related to the inluence of Bohemian and, to a lesser extent, Silesian colonial dialects of German. Those Unity of the German component of Yiddish 105 of proto-WY appear to be closely related, irst of all, to East Franconian. Though these conclusions concern all components of these two major subdivisions of modern Yiddish, the analysis that allowed obtaining these results was focused on the vowels in words from the German component of WY and EY. Moreover, the stressed vowels whose relexes in reconstructed schemas of proto-WY and protoEY are diferent (namely, proto-diphthongs and front rounded vowels /y/ and /⊘/ appearing in proto-WY but absent from proto-EY), are present only in words from the German component. Timm (1987: 457) suggests an idea about numerous features of WY originally appearing in the Jewish speech in areas where Christian population spoke Rhine Franconian or, to a lesser extent, East Franconian, before being spread out to other western German-speaking territories during the 15th and early 16th century. For drawing such conclusion, she considers several phonological features: (1) merging of [e]- and [ε]-qualities for short vowels, but diferent development for long or lengthened; (2) relexes for MHG ie and uo became equal to lengthened MHG i and u, respectively; (3) [a:]-realization for both MHG ei and ou; (4) diphthongal relexes of MHG ê and ô and lengthened MHG e and o; (5) lowering /i/ > /e/ and /u/ > /o/ before /r/. The genesis of EY was out of her scope. On the contrary, there exist a number of studies whose authors ignore WY, but try to establish links between EY and some German dialects. Gerzon (1902: 131) takes into account only one phonological feature (the distribution of [p]-[pf]-[f]) and makes a remark about the possible close relationship between EY and ECG. Several authors suggest the derivation of the German component of EY from Bavarian.6 Two of them, correspond to the irst quarter of the 20th century. Sainéan (1902) starts by postulating this link (1902: 94) and later adds a small list of peculiar items from Central Yiddish vocabulary that he generally relates to Bavarian (1902: 132– 137). The Bavarian origin of Yiddish represents one of the main ideas of Mieses (1924) whose linguistic arguments are rather poor. During the last quarter of the 20th century, these theories were revived, principally thanks to eforts by King (compare Faber and King 1984; King 1992, 1993). His argumentation includes a list of four major phonological features of EY, two of its morphological characteristics (namely, the diminutive and plural suixes), and a group of peculiar 6 Bavarian is mainly understood by them as covering not only Bavaria and Austria, as in standard works on German dialectology, but also the speech of German colonists in Bohemia and Moravia. Moreover, examples are generally checked by these authors in Schmeller (1827–1837), a book that covers the territory of the state of Bavaria at the moment when the book was compiled. This area incorporates territories where Christians spoke not only Bavarian, but also (eastern) Swabian, East Franconian, and (partly) Palatinate German. 106 Alexander Beider pronouns (actually limited to Central Yiddish and not to all of EY) that he all links to Bavarian, the German dialect he declares to be the basis for the German component of EY. To this, he adds a list of three phonological features that Yiddish received due to its contacts with East Central German. One part of his arguments does not represent King’s original ideas. Certain of them were already quoted in Weinreich (1973), Bin-Nun (1973) or Birnbaum (1979). Contrary to King, these scholars never generalized these similarities to advocate the Bavarian origins of Yiddish. Gold (1986) and Katz (1993: 12–14) provided the irst criticism of his ideas. Manaster Ramer and Wolf (1996) clearly demonstrated striking methodological drawbacks of King’s approach. Eggers (1998) is another proponent of the theory of the Bavarian origin of Yiddish. To support his ideas, he suggests a list of a few grammatical, morphological, and phonological features considered by him as “markers” of Bavarian. Many of them are taken from Mieses (1924), some are similar to King’s arguments, and others are original. Manaster Ramer (1997) deals with common origins of varieties of modern Yiddish. He provides a wide list of lexical, semantic, phonological, morphological, and phraseological features, proper to both EY and WY and unknown in other languages. A similar approach also characterizes Timm (1987: 375–385) who, moreover, presents the earliest references to certain words. Globally speaking, ater almost one hundred years since the irst scholarly publications about Yiddish appeared, both the geographical and chronological dimensions of the question of unity of the German component of that language remain in many aspects controversial. It was the study by Timm (2005) that created the irst real breakthrough in this domain. She shows that a large number of semantic and morphological peculiarities of the German component of both WY and EY have common origins that can oten be traced to medieval Jewish texts from West Germany. A large pan-Yiddish layer considered in detail by her is due to the tradition of the translation of the biblical text into the vernacular language of Ashkenazic Jews. This continuous tradition – whose distinct traces can be found in western Germany by the end of the 14th century and whose stabilization is clearly observable in sources compiled during the 16th to 17th centuries by Jews in such distant areas as western Germany, northern Italy, Bohemia and Poland – had a huge inluence on the formation of the German component of Yiddish. This is principally because all Ashkenazic boys in diferent European countries were being educated in Jewish elementary schools using translated biblical texts. Timm’s opus magnum provides a profound insight for both dimensions of the question of the unity of Yiddish. In some major aspects, Timm joins general, rather intuitive, ideas about the existence of the unity exposed in the past by Weinreich and Birnbaum, providing irst strong evidence that corroborate the ideas in question. Unity of the German component of Yiddish 107 Timm’s work mainly focuses on linguistic innovations that occurred within Ashkenazic communities and not features that are shared by Yiddish and German. Links that could exist between Yiddish and various German dialects are outside of the scope of her book. The aim of the present paper consists in providing additional insight into the controversial question of the unity of the German component of Yiddish. It attempts to complement the results obtained by Timm by a systematic study of elements that do not represent internal Jewish innovations but likely were inherited by Yiddish varieties from German dialects. The paper primarily addresses phonological diferences: isoglosses from this domain generally serve to identify various German dialects. Moreover, phonological isoglosses are oten more useful for establishing genetic links than morphosyntactic. To avoid paying attention only to a few characteristics whose choice could, in principle, be considered as random, the selection of phonological elements in this study is primarily based on the synthesis made by Žirmunskij (1956), one of the most authoritative in German dialectology. Comparisons are made between the following entities: – High German dialects (considered in the historical perspective) – Modern Yiddish varieties – Languages of a sample of available early Ashkenazic sources (13th to 17th centuries) The irst comparison (Section 3) is done between German dialects and the language(s) of early Ashkenazic sources in order to see whether it is appropriate or not to consider these vernacular Jewish idioms to be diferent from the dialects spoken by coterritorial Christian population. Since their links to modern Yiddish remains to be proven, it is on purpose that the word “Yiddish” is never used in this paper when speaking about these early idioms. The second comparison (Section 4) is done between modern Yiddish varieties and German dialects to see whether any distinctive links between them can be discerned. It sheds more light on the geographic dimension of the question of unity of Yiddish. Modern Yiddish varieties are also compared to the languages of early Jewish sources in order to establish genetic links between them. This process – dealing with the chronological dimension of the question of unity of Yiddish – allows separating some features that were really inherited by modern Yiddish varieties from their ancestors, from those features that are due to rather recent phenomena of interdialectal borrowing internal to Yiddish or to the inluence of coterritorial dialects spoken by German Christians. In Section 5, modern Yiddish varieties are compared between themselves and the non-adequacy of their classiication currently used in linguistics is established. For this reason, the expression “Yiddish dialects” is avoided in 108 Alexander Beider this article when speaking about WY versus EY, or about various Jewish idioms generally considered being subdivisions of WY. Yet, during the preparation of this article no information was found contradicting the idea of unity of EY. Subdivisions of EY clearly descend from the same ancestor and for them the expression “EY dialects” is totally adequate. For establishing links between linguistic entities, the importance of various elements is not the same. It is clear that of smallest importance are items that concern only individual words. The interest of this layer is particularly marginal when dealing with old sources because (a) we can be dealing with a scribe error; and (b) we never know the exact geographic distribution of any particular word in various German dialects during the period under consideration. To this category – marginal for the analysis present in this article – belong all lexical elements except for a few basic vocabulary terms, as well as numerous phonological and certain morphological elements too. On the other hand, of particular signiicance are elements that are structural for the development of a dialect and concern a large number of words. For these elements, crucial for our analysis, the dialectal geography either was already totally established before the early Ashkenazic sources known to us were compiled (this is particularly true for consonants), or the chronology of changes in various German dialects is well documented. In this article, when discussing links between various dialects, attention is paid to characteristics that make dialects diferent and not to those according to which the dialects look similar. The main advantage of this method consists in avoiding the construction of logically incorrect inferences.7 Its main drawback: when measuring a diference between two dialects, the existence of correlation between them according to some feature is not favored in comparison to the situation when the information about this feature in one of these dialects is lacking. Nevertheless, this drawback is less signiicant than the advantage formulated above. The following abbreviations are used in this paper for High German: CB (Central Bavarian), CF (Central Franconian), CG (Central German), ECG (East Central German), EF (East Franconian), HA (High Alemannic), LA (Low Alemannic), MF (Moselle Franconian), NB (North Bavarian), PG (Palatinate German), RF (Rhine Franconian), SB (South Bavarian), UG (Upper German), UpS (Upper Saxonian), 7 An example of inappropriate logic: Since (Yiddish variety Y exhibits – according to a feature F – a behavior similar to that of German dialect G1), one concludes that (Y is descending from / is related to G1). Actually, Y can, in principle, be descending from another German dialect G2 that also shares the same behavior. Unity of the German component of Yiddish 109 WCG (West Central German). Unless speciied explicitly, the information about realizations found in diferent German dialects was directly taken from their corresponding dictionaries.8 Table 1 lists varieties of modern Yiddish.9 Table 1: Modern Yiddish varieties English name Abbreviation 1. Eastern Yiddish 1.1 Central Yiddish 1.2 Northeastern Yiddish EY CY NEY 1.3 Southeastern Yiddish 2. Southwestern Yiddish 3. Dutch Yiddish 4. Czech Yiddish 5. East German Yiddish SEY SWY DuY CzY EGY Modern country (historical province) Poland (main part) Lithuania, Belarus, northern Ukraine, northeastern Poland, Latvia southern Ukraine, Moldova (Bessarabia) France (Alsace), Switzerland Netherlands Czech Republic (Bohemia, Moravia) eastern Germany, Russia (East Prussia) 2 List of features taken into account A considerable number of dialectal variants concern the realizations of consonants. See Table 2.10 8 Schmeller 1827–1837; Heilig 1898; Jelinek 1911; Martin and Lienhart 1899–1907; Müller 1928– 1971; Müller-Fraureuth 1911–1914; Krämer 1965–1997; Mitzka 1963–1965; Fischer 1904–1936; Staub and Tobler 1881–2000; Grimm and Grimm 1854–1960. Facts from historical German dialectology were taken from Moser (1929, 1951), Paul (1998), Sauerbeck (1970) and Stopp (1973, 1978). 9 Only the names of the EY dialects appearing in Table 1 are more or less standard in Yiddish linguistics. The designations of other Yiddish varieties used in this paper are purely conventional. The information was mainly taken from the following sources: (1) for EY, from classical dictionaries of StY that is to a great extent based on NEY; (2) for EGY, from Friedrich (1784); (3) for DuY, from Beem (1959); (4) for SWY, from Weill (1921), Zivy (1966), Zuckerman (1969) and Guggenheim-Grünberg (1976); (5) for CzY, from Schnitzler (1966) and Beranek (1965). 10 All features taken into account in this paper are conventionally designated by a code that starts with a letter that identiies the category concerned (C = consonants, V = stressed vowels, U = unstressed vowels, M = morphology and grammar) and ends with a sequence number. These codes are introduced to simplify making references to these elements in Sections 3–5. 110 Alexander Beider Table 2: Consonantal features Feature Description Possible phonetic realizations {C1} {C2} {C3} {C4} {C5} {C6} {C7} {C8} {C9} {C10} {C11} {C12} {C13} {C14} {C15} {C16} {C17} {C18} {C19} {C20} {C21} {C22} {C23} Old German p in various positions Old German k in various positions Initial MHG tw Initial and internal Old German d Initial Old German b German neutralization of consonants (stops) MHG b (intervocalic, between a liquid and a vowel) Final MHG p ater a vowel or a liquid Final MHG we ater a liquid MHG we ater u or û Intervocalic MHG v Internal st MHG rs Intervocalic and initial prevocalic MHG s Final consonant of MHG –em MHG mp and mb Verbal preix MHG erMHG ht MHG hs Introduction of /n/ between /j/ and /s/ or /š/ Shits /g/ > /j/, /j/ > /g/, and /xt/ > /št/; elision of inal /n/ Verbs with /j/ in ininitives and /g/ in past participles Phonetic shit [rš] > [rž] p, pf, f k, kx, x tv, tsv, kv, kw t, d b, p Yes11 or No b, w/v p/b, f b(e)/p, f zero, v/b v, f, b st, št rs, rš s, z m, n (or zero) mb/mp, m er, her, der ht, t, xt hs, s, xs, ks Yes or No Yes or No Yes or No Yes or No A signiicant number of dialectal variants concern the realizations of stressed vowels. See Table 3. Several major dialectal distinctions can be also observed for unstressed vowels. See Table 4. An important dialectal variation is found in several morphological and grammatical features. See Table 5. 11 With a special distribution in Bavarian (depending on the length of the preceding vowel). Unity of the German component of Yiddish 111 Table 3: Features related to stressed vowels Feature Description {V1} {V2} {V3} {V4} {V5} {V6} {V7} {V8} {V9} {V10} {V11} {V12} {V13} {V14} {V15} MHG î and û MHG ie and uo MHG ei and ou Merger of MHG â and ô Merger of MHG ô and uo Raising of MHG a MHG e, ë and ä/æ in open and closed syllables Lowering of MHG i and u Raising of MHG o Unrounding of MHG rounded vowels /u/ in place of MHG ü Shit /a/ > /e/ before /š/ Shortening of MHG â and ô before f and ᵹ in polysyllabic words Shortening of long vowels before f, ᵹ, and ch Shortening of MHG î, û, and iu before a cluster formed by a fricative consonant and /t/ or /ts/ General lengthening of short vowels in open syllables Lengthening of short vowels in open syllables before /t/ {V16} {V17} Table 4: Features related to unstressed vowels Feature Description Possible realizations {U1} {U2} {U3} {U4} General apocope of the unstressed -e Apocope in adjectives with strong declension MHG verbal preix zerMHG verbal preix ver- Yes or No Yes or No tse(r), tsu(r) far/var, fer/ver, for/vor 3 Early Jewish sources and German dialects The order of the presentation of early Ashkenazic sources in the text below is conventional, but mainly chronological:12 12 In this article, the information from R8, Be, Ox, H96, Kr and Pr, was taken from Röll (2002) where all glosses for the Book of Job appearing in these sources are indexed and some are discussed. For R9 and Le, also covered by the same book, some additional elements were found in other sources that are made explicit below. 112 Alexander Beider Table 5: Morphological and grammatical features Feature Description {M1} {M2} {M3} {M4} {M5} {M6} {M7} {M8} {M9} {M10} Main diminutive singular suix: -(e)l, -le, -li, -xe(n) Existence of diminutive suixes of second degree (-ele) Existence of the diminutive plural suix –lex Loss of preterite Present tense forms of the verb ‘to have’ Participle of the verb ‘to have’ First singular of the verb ‘to be’ Second singular of the verb ‘to be’ Plural forms of the verb ‘to be’ Pronouns meaning ‘you’ (nominative, dative, accusative) and ‘your’ (1) SAB: The Short Arukh, a manuscript from Bern, represents a HebrewAramaic dictionary (Frakes 2004: 5–6) whose 318 Germanic glosses appear in their totality in Timm (1977). This manuscript was compiled in the Cologne area in 1290. Only the following elements do not conform to local Ripuarian dialect: {V12} (one word), {C6} (two words), and {C4} (a few words). A comparison to other dialects reveals more important diferences. Their number is small only for other WCG dialects: (a) for MF and Hessian, it also includes {C1} but this is precisely the main criterion according to which Ripuarian is distinguished from other WCG dialects; (b) for PG, {V2} and {V8} should be added.13 (2) CC: This collection of eight documents – mainly poems – from the Cairo genizah kept in the Cambridge University Library (Frakes 2004: 8–43), is oten called Cambridge Codex. One of these texts holds the date of 1382, while the year according to the Jewish calendar noted in another text points to 1382/1383. The consideration of {C1, C10, C17, V1, V10, M7, M8} eliminates all dialects except for two Alemannic subdialects: Swabian and HA, with only {U3} that does not conform to both of them. Numerous other, less important, features also contradict CG, Bavarian, EF, Bohemian and LA. Swabian its better than HA according to {C2} (but see the explanation below concerning R8), {C4} and {C28}. Yet, HA is preferable to Swabian according to {C6, C9, C19}.14 13 The analysis by Timm (1977) also shows the direct link between SAB and Ripuarian. 14 Several attempts to assign CC to German dialects were already done in the past. The most detailed one is that by Althaus (1971) who took into consideration only one of the texts, the fable “An old lion”. The author makes a summary of the results of his predecessors (1971: 32) who wrote about Hessian, EF or NB “connections”. He also proposes iteen arguments to point to the linguistic origin in southern Thuringia circa 1300 (1971: 199–202). Of these arguments: one (no. 5) Unity of the German component of Yiddish 113 (3) R9: This biblical glossary from the collection by Johann Reuchlin (1455– 1502) is generally assigned to the end of the 14th century. The information contained in it is present and/or discussed in: Heide (1974) (referred to as R), Timm (1987) (as KJ), Heide (1977), Timm (2005) and Röll 2002 (in all, as R9). According to {C1, C4, C12, V2, V5, M1, U1}, this document does not conform to any dialect except for PG or Hessian, that is, the two subdialects of RF.15 However, the element {V1} clearly contradicts both of them, as well a small number of other, less important, elements too: {M7} to Hessian, while {C17} is only partly correlated with RF. Two alternative explanations can be valid to explain this situation. Firstly, R9 could appear in the area near the border of RF and EF where some features of EF were present: the diphthongization of MHG î and û {V1} did take place and the preix der-{C17} was used. This scenario looks plausible: note that (a) no diphthongization was present in the language of the pointer of the same document that presumably was from the same general area, and (b) R9 also includes the preixes er- and her-, a feature typical for a document from a transitional area. results from an incorrect quote from Moser (1929: 106); nos. 6, 9 and 10 are unreliable because the factual materials on which they are based can be interpreted in a diferent way; nos. 13–15 deal with individual words. The approach by Weinreich is totally diferent. For him, this manuscript unambiguously testiies that even during the 14th century there was already a very marked fusion of German dialectal elements from distant regions in the vernacular language of Ashkenazic Jews (Weinreich 1973: Vol. 2, 392–393). To illustrate his idea he proceeds in several steps. Firstly, he postulates the existence of a few typical CG features. Here Weinreich refers to the form he ‘he’ that “appears interchangeably” with a more standard er. According to German dialectology, he is unknown in UG and southern part of WCG (compare König 1978: 164). Weinreich also gives a few examples of /d/ instead of /t/ that, according to him, also point to the northern part of WCG situated in the vicinity of the Low German area. Secondly, he mentions a number of UG traits. Here he deals with two general features (that in this article are designated {C1} and {C19}), as well as with a few individual words. One of them, the equivalent of the NHG modal verb sollen ‘should’ regularly starts with /š/ in CC. This fact reveals, according to Weinreich, its Bavarian origin. As a result, Weinreich thinks to show the presence of traits that are reminiscent of diferent corners of the German dialectal space. However, his arguments do not appear to be cogent. Indeed, the “Bavarian” feature in question was actually also typical to medieval EF and well known in Alemannic (Moser 1951: 227). Medieval references to /d/ instead of /t/ in the speciic context ater /l/ to which all his examples correspond characterize all of WCG and were commonly found in EF too. It suices to look into the index prepared by Hakkarainen (1973: 174–181, 221) to see that the putative “interchangeability” of he and er in CC is an exaggeration. We ind there many dozens of references to ‫‘ ער‬he’ and only a few references to ‫ הי‬that oten correspond not to the meaning ‘he’, but to the MHG hie ‘here’. 15 This result conforms to the conclusion present in the most detailed and adequate analysis of R9 that was already published, namely that by Röll (2002: Vol. 1, 62) who says that this document shows features of southern WCG. 114 Alexander Beider Secondly, R9 can represent one of the earliest Jewish sources that show a blending of elements taken from diferent German dialects. (4) Le: A manuscript with a biblical glossary kept in Leipzig, it was most likely compiled in the turn of the 14th to 15th centuries. It attracted attention of several scholars: Heide (1974, 1977), Timm (2005) (in all of these works, referred to as L), Timm (1987) (as LJ), Banitt 2005 (Vol. 1), Röll (2002) (as Le, see [Vol. 1, 20–25, 77–87]). This document shows numerous features that are not correlated with any dialect except for LA.16 Only the element {C10} contradicts to (modern) LA. Since this document behaves here as MHG, while we have no direct references that would allow us to know the exact time when LA acquired its modern relex, in this particular case, we can easily deal with a testimony about the relex that was valid for LA at the time Le was compiled.17 (5) R8: Another biblical glossary from the collection of manuscripts by Reuchlin, it is considered to be compiled at the start of the 15th century (Röll 2002: Vol. 1, 15–20, 70–77, referred to as R8). Taking into consideration of {C1, C10, C17, V1, V2, V5, V10, M1} and numerous other elements show a close link to HA. The only exceptional element is {C2}: in HA, one would expect to ind traces of the africate /kx/. However, this contradiction is not strong. The manuscript is written using the Hebrew alphabet that has no equivalent for this africate, and therefore, being based on the spelling only, one can not judge about the exact pronunciation of the sound expressed in R8 via ‫ק‬. (6) Be: This manuscript containing a glossary for several biblical books is kept in Berlin. It is assigned to the 15th –16th centuries (see Röll 2002: Vol. 1, 26–30, 87–102, referred to as be). Taking into account {C1, C2, C7, V1, U1} and a number of other elements leaves only NB, CB, EF, and Bohemian as potential donor dialects. However, {C10, V13} contradict both Bavarian dialects. {M1} does not it EF. {V10} does not conform to Bohemian. Similarly to the way R9 was considered above, two alternative scenarios can be suggested here to explain these exceptions. Firstly, the document can correspond to an area on the border between NB and EF (that is, around Nürnberg). Secondly, we can face in Be an example of Jewish merging of elements taken from diferent German dialects. (7–10) BB: Bovo-bukh (1507, published in 1541), a chivalric verse romance by Elia Levita (1469–1549), the famous Jewish polymath. Born in Ipsheim (Middle Franconia), he spent a large part of his life in northern Italy. To the same author are also due: SD Shemot Devarim // Nomenclatura Hebraica (Isny, 1542), a dictio- 16 Same conclusion appears in Banitt (2005: Vol. 1, 430). 17 Even in books printed during the 16th century in the area of LA one still inds the spelling blauwer ‘blue’ (Moser 1951: 86). Unity of the German component of Yiddish 115 nary of his German-based vernacular language together with the Hebrew translations to which the editor Paul Fagius added the German and Latin equivalents; ST Sefer Tehilim (1545), a translation of Psalms, and most likely PuV Pariz un’ Viene (circa 1540, published anonymously in 1556), another chivalric verse romance.18 These four works have little correlation with any German dialect except for EF and Swabian.19 If we exclude a few elements that correspond to individual words, only {V12} contradicts to EF, while only {C7, V4} do not conform to Swabian. It is diicult to say whether the language of these works was due to a merging of EF and Swabian features realized by Jews or was simply due to a German dialect from an area intermediate between EF and Swabian. Note also the partial character of unrounding {V10} in both PuV and BB, while unrounding is typical to Swabian and does not characterize EF. (11) MM: Mirkevet ha-mishne by Rabbi Asher Anshel (Kraków, circa 1534; Frakes [2004: 174–176]) represents a dictionary of words from biblical Hebrew with their translation into the vernacular German-based language of its author. Numerous quotes from it are present in Timm (2005), some speciic aspects are discussed in Heide (1974, 1977) (in all three studies, this source is referred to as MM). Taking into account the elements {C1, C5, C10, C17, V1, M1, M3} eliminates WCG, Thuringian, Alemannic and EF. Among the remaining dialects: Silesian and UpS are problematic because of {V9, U1} and, less important, {C3}; Bohemian is clearly the closest dialect. (12) Mel: Sefer Melokhim, generally called Melokhim-bukh in Yiddish philology, is an epic poem irst published in 1543 in Augsburg (Frakes 2004: 194– 213).20 The application of such major criteria as {C1, C4, C5, C17, V1, V12, U1, M1} leaves only the following dialects with which Mel is correlated rather well: EF, Swabian and Bohemian, for which important exceptions are {C15, V12}, {V4, U3} and {V4, V12}, respectively. The relexes of {U4} make implausible a possibility of any link to Bohemian. As a result, exactly as for works by Elia Levita, Mel reveals a mixture of EF and Swabian. 18 Information is taken from: SD (directly), Timm (1996) (with index of peculiar words in PuV); Timm and Gehlen (1996) (detailed discussion of linguistic features of both PuV and BB). Numerous features of all four works appear in Timm (2005) where they are referred as BB, SD, El, and PuW, respectively. See also Weinreich (1993: 51–67) and Frakes (2004: 120–139, 189–192, 254–259, 393–414). 19 This follows from the consideration of criteria related to {C1, C5, C12, C17, V1, U1, M3} and some others too. 20 The information is mainly extracted from Fuks (1965) with a facsimile of the edition of 1543 and a glossary. A number of features are also discussed in Timm (2005) (referred to as Mel). 116 Alexander Beider (13) Ox: This biblical glossary held in Oxford University was written in northern Italy in 1546/1547 (Röll [2002: Vol. 1, 33–37, 104–108], referred to as ox). Taking into account of the elements {C1, C4, C5, C17, V1, V10, U1, M7} eliminates all dialects except for EF. (14) H96: This biblical glossary from a collection held in Hamburg was copied most likely during the second half of the 16th century (Röll [2002: Vol. 1, 41–45, 112–115] referred to as h96). Its language represents an example of doubtless mixture of elements taken from various German dialects. For example, {M1} is typically CG, while {C1} can not be CG. Still, some dialects are in a favorite position in comparison to others. For example, in the list of major characteristics {C4, C17, V1, V10, V12, U1, M7}, only EF and Swabian contradict only one characteristic ({V12} and {V10}, respectively), all others do not conform to at least two characteristics. If we also take into account {U3, M9} and other features, we can globally see that EF is the closest dialect, while various CG dialects difer from the language of H96 most substantially. (15–17) BZR, BZV, BZP: Three versions of Beria ve-Zimra, an anonymous Jewish tale (Frakes 2004: 355–367). The irst one is a manuscript written during the 1580s by Isaac bar Judah Reutlingen (whose place of residence and native town are both unknown). The two others are printed editions from Prague due to the Bak family (1597 and circa 1620, respectively). Their texts and detailed linguistic analysis are present in Timm (1987) where they are referred to as R, V, and P, respectively. Taking into account such fundamental criteria as {C1, C4, C5, C10, C17, V1, V3, V10, U1} shows that the language of BZR its to EF and does not conform to all other dialects. Only {C15, V12} are not compatible with EF. Note that all these exceptions are in line with Swabian, a neighboring dialect. On the other hand, if we apply the criteria related to {C1, C17, V1, V10, V12} to the two documents from Prague, all dialects are eliminated except for Bohemian.21 (18–19) Kr (Ayola Sheluḥa by Naphtali ben Asher Altschuller, Kraków, 1593– 1595) and Pr (Lekaḥ Tov by Moses ben Issachar Sertels, Prague, 1604; Frakes [2004: 496–499]) are two printed biblical glossaries (Röll [2002: Vol. 1, 45–50, 116–125] where they are referred to as kr and pr, respectively). The consideration of such features as {C1, C17, V1, V10, M1} shows that Kr can be closely related to Bohemian and, to a lesser extent, Silesian (with which {V9, U1} are not correlated). For Pr, numerous characteristics do not conform to CG22 and UG.23 21 See also Timm (1987: 459–463, 476–479). 22 {C1} is not CG, {C17} and many others are not WCG, {V12} is not ECG, {C6} is not Silesian, while {C3, C7} are neither UpS, nor Thuringian. 23 {C5, U3, M6, M9} are not Alemannic, {C17, V1} exclude LA and HA, and {C5, C10} are not Bavarian. Unity of the German component of Yiddish 117 Smaller but still visible diferences exist in comparison to EF: {C16, V12}. Only {V12} and partially {V3} are not compatible with Bohemian. (20) NH: Sefer Safa Berura [The book of pure speech] by Nathan Nota ben Moses Hanover (Prague, 1660) represents a quadrilingual Hebrew-“Ashkenazic” (the vernacular language of the author)-Italian-Latin glossary (Frakes 2004: 669– 670; Weinreich 1993: 114–121). The author came to Prague from Ukraine during the Cossack wars; he also lived in Italy. Only Bohemian “resists” the consideration of such elements as {C5, C12, C17, V1, V5, V9, V10, V12, M1, M2} and numerous others, less signiicant. However, this dialect does not it to such fundamental characteristics as {C1} for which the language of NH behaves as ECG. Summarizing the above discussion, one can see that the twenty early Jewish sources taken into consideration in this paper can be divided into three groups. The irst of them encompasses the earliest manuscripts. They are linguistically extremely heterogeneous: SAB related to Ripuarian, R9 to RF, CC correlated to Swabian, Le to LA, R8 to HA, and Be to (partly) NB, CB, EF and Bohemian. The second group includes western sources from the 16th century and all of them are primarily related to EF and, at to a lesser extent, Swabian: four works by Elia Levita, Mel, Ox, H96 and BZR. The third group deals with materials from the 16th and the irst half of the 17th century that were printed either in Prague or Kraków: MM, BZV, BZP, Kr, Pr and NH. All of them show close similarities with Bohemian. It is important to note that these results were obtained only via a “neutral” comparison of the behavior of various linguistic features found in these documents to the corresponding features of German dialects: no extra-linguistic factor was taken into account. Additional available information of diferent kinds can be used in order to test its compatibility with the above links constructed theoretically. If we take into account historico-geographical factors then the corroboration is good enough: (1) SAB was compiled in the area of Ripuarian; (2) Le includes not only Germanic glosses but also French Jewish glosses too, while LA is peculiar to Alsace, an area near the French border; (3) Elia Levita’s native town is situated in the area of EF close to the border of Swabia; (4) the editor of Mel was Paulus Aemilius, a converted Jew born in the area of EF; (5) we do not know anything about the life of Isaac Reutlingen, but his nickname is derived from the town of Reutlingen in Swabia; (6) all books from Prague show similarities with the Bohemian colonial dialect of German. Useful information can also be extracted by comparing diferent aspects of the Jewish sources in question not with German dialects but between themselves. For example, considering {C18}, one can immediately see the speciicity of all oldest documents (except for R9) in comparison to others. One can also see that among the oldest manuscripts Be stands apart if we take into account {C6}. This factor corroborates its geographic separation from other oldest 118 Alexander Beider sources suggested above without taking this criterion into account because for early times it does not appear to be reliable. A number of observations can be also made analyzing their Hebrew spelling. The irst criterion is related to the letter heth. In a number of early documents, it appears in place of MHG (word- or syllable-) initial or intervocalic h: SAB (Timm 1977: 32), CC (Timm 1987: 266), Le (Heide 1977: 6, Röll 2002: Vol. 2, 84), R9 (Heide 1977: 5–6), R8 (Röll 2002: Vol. 1, 74) and Be (Röll 2002: Vol. 1, 99). The /h/-pronunciation of heth by Elia Levita is well known (Weinreich [1958: 103–104], see also Timm and Gehlen [1996: 315–316] about PuV). This feature characterized Jews from western German-speaking provinces, but was not valid for Jews from the town of Regensburg, Austria and Slavic countries (Weinreich 1958). This information is well correlated with the above results. For Be, it allows to exclude Bohemian and therefore the region of Nürnberg becomes its most plausible source. The second argument comes from the use of alef to express /o/. This spelling characterizes, thoroughly or partially, only documents from Prague or Kraków. Other, western, early Jewish sources under consideration use only vav (compare Timm 1987: 113). For MHG v in the initial position, fe represents the main letter used in all documents from Prague and Kraków except for the earliest one, MM, that – exactly as the majority of western sources – uses vav. The use of silent alef expressing no sound at the end of a word represents a third factor. Only in the group of oldest sources (SAB, CC, R9, Le, R8, and Be), it regularly appears ater a consonant providing this way an additional argument to their spacio-temporal closeness in comparison to other documents in which the silent alef can be present only ater a vowel. The last three manuscripts are also uniied by the spelling of pronouns, articles, and prepositions together with a word they precede or follow, without any blank between them. The use of special diacritic signs over certain letters, unknown in other Ashkenazic sources, characterizes SAB (caron over shin), CC (caron over shin, ṣadi, and pe), R8 (two parallel dashes over shin, ṣadi, pe, and double vav) (Röll 2002: Vol. 1, 70). Le uses ḥolem-vav and shureq-vav for stressed vowels to distinguish not only /o/ from /u/ (as do other manuscripts), but vowels without umlaut (/o/, /u/ and diphthongs starting with them) from those with umlaut (/⊘/, /y/) (Röll 2002: Vol. 1, 77–78). The presence of several words of Romance origins can also be used to test closeness of various sources. Among them are: or(e)n ‘to pray’ (found in R8, Mel, BZR), pülts(e)l / pilts(e)l ‘maiden’ (R9, ST, PuV and Mel), and preyen/prayen/brayen ‘to invite’ (R9, PuV, Mel, BZR, BZP and Pr). Taking into account these references and those from other early Jewish sources (Timm 2005: 439–440, 445, 456–457), we can see that they are much more common to western documents than to those from Central and Eastern Europe. The information presented above shows that nothing suggests any linguistic unity of the Jewish sources written before the 16th century. Globally speaking, Unity of the German component of Yiddish 119 their language is likely to have had the same dialectal peculiarities as those present in the German dialects spoken by their Christian neighbors. Evidently, some diferences existed already too. Jews clearly used a number of words of Hebrew origin in their speech. For stylistic reasons, these words rarely appear in early Jewish manuscripts: authors were clearly aware about their provenance and tried to avoid their use in their German-based texts or glosses. A few words of Romance origin used by Jews were unknown to Christians. In the German component, a number of speciically Jewish semantic peculiarities oten related to the tradition of biblical translations. Among the earliest sources, their number is signiicant in R9 only. We also have a few examples of the creation by Jews of words using Hebrew or Romance roots and German aixes. Most important, however, is the fact that we do not ind during this period any reference to Jews retaining archaic German elements and not following processes of linguistic change that were taking place for their Christian neighbors. Moreover, we do not see any unambiguous evidence about Jews mixing various German dialects. For sources known from the 16th to 17th centuries, the situation is signiicantly diferent. They clearly fall into two groups: western (from northern Italy and western Germany) and eastern (from Prague and Kraków). Inside every group, sources show numerous similarities. However, linguistic diferences between the two groups are important: they clearly had diferent German dialects as primarily donors. 4 Modern Yiddish varieties and German dialects Table 6 lists elements of Yiddish varieties that are not compatible with various German dialects. Abbreviated codes of basic linguistic features given in this table all appear in Section 2 in braces in bold face (these braces are omitted in Table 6). The information present in Table 6 allows seeing that EY is close only to two German colonial dialects: Bohemian and Silesian. These dialects are complementary. Bohemian is compatible with the main vocalic features and morphology of EY. However, several major consonantal features – {C1, C6} and, less important, {C16, C23} – are typically Silesian. On the other hand, Silesian does not conform to EY for a number of morphological and grammatical features {M2, M3, M4} and contradicts to EY by {V5, V9, U1}. The only consonantal trait by which EY is distanced from Silesian is {C20}. However, it concerns only a few words and may be due, at least partly, to an internal Jewish innovation. It is generally admitted that in numerous Polish cities and towns a large part of the Christian population spoke before the 16th century (and in some places even later) a dialect of German German dialects 1.1.1 CF 1.1.1.1 Ripuarian 1.1.1.2 MF 1.1.2 RF 1.1.2.1 PG 1.1.2.2 Hessian 1.2 ECG 1.2.1 Thuringian 1.2.2 UpS 1.2.3 Silesian 2. UG 2.1 Alemannic 2.1.1 Swabian 2.1.2 LA 2.1.3 HA Yiddish varieties EY CzY EGY DuY M2, M3 C1, C4, C7, C17, C22, C23, V6, V12, M5 C8, C9, C18, C19, C20, V2, V7, M1 V1, V10, V16, M7, M8 C6 C6, C22, C14, V3 C12, C20, V17 V2, V14, M1, M7 C20, V5, U1, M4 M3, M10 C1, C4, C7, C16, C17, V6, V12 C8, C9, C18, V2, V3, V7, M1 V1, V10, V16, M8 C6 C6, C22 C12 V2, M1 V3, V5, U1, M4 C1, C4, C7, V6, M5 C1, C4, C7, V6, V12 C8, C9, C12, C18, C20, V2, V3, V7 V1, V10, V16 C6 C6, C22 C20 V2 C12, C20, V3, V5, U1 C6, C7 C6, C7 V9, M1 C1, C6, C11, V2, M1 V3, V6, V12 C20, V2, M7 V3, V5, U1 C6, C7, C15, C16, C23, V14, M1 C6, C7, C15, C23, V14 V9 C1, C6, C11, C14, C22, C23, V2, V17, U2 C12, C15, V6, V12, U3, M5, M6, M9 V11, M1 C7, C17, V1, V4, V15, M8 C2, C10, C13, C17, C19, V1, V10, V15, V16, M1, M3 C6, C7, M1 C6, C7 V9 C1, C6, C11, V2 C1, C6, C7, C15 C1, C6, C7, C15 V9, M1 C1, C6, C11, V2, M1 C12, C16, V6, V3, V12, M10 V11, M1 C7, C17, V1, V4, M8 C2, C17, V1, V10, V16, M1, M3 C15, V3, V6, V12, M5, M9 V11 C7, V1, V4, M8 C2, C10, C19, V1, V10, V16 V3 C7, V1, V4 C2, C10, V1, V10, V16 C7, V4 C17, V1, M8 C2, C6, C7, C10, C17, V1, V4, V16, M3 C8, C9, C18, C19, V2, V3, V7, V12 V1, V10, V16, M7, M8 C6 C6, C22 SWY C1, V4, M1, M3, M6 C4, C17 C8, C9, C12, C18, V2, V3, V7, U2 C6, V1, V16, M7, M8 C22 V2, U2, M7 C12, V3, V5, V12, U1, U2, M4 C16 V9, C6, C7 V2 Alexander Beider 1. CG 1.1 WCG 120 Table 6: German dialects and Yiddish varieties: incompatible features 2.2 Bavarian 2.2.1 NB 2.2.2 CB 2.2.3 SB 3. Intermediate 3.1 EF 3.2 Bohemian C5, C10, C16, C20, V3, V11, V13 C7, C15 C7, C15, V7 C2 C1, C6, C23 C7, C11, C14, C15, V3, V6, V10, V17, U2, M1, M5 C16 C5, V11, V13 C5, C10, V11, V13 C7 C7, V7 C2 C1, C6 C7, C11, C16, V6, V10, M1, M10 V3 C7, C15 C7, C15, V7 C2 C6, M1 C1, C7, C11, C15, C20, V6, V10, M5 V3 C5, C10, C16, C20, V13 C7, C12 C7, C12, V7 C2 C1, C6, M1 C7, C11, V6, V10 C5, C6, C10, C16, V4, V12, V13, M1, M6 C12 V7 C2, C7, C12 V12 C12, V3 C12, C16, V3, V4, M1 Unity of the German component of Yiddish 121 122 Alexander Beider related to Silesian.24 It is much more likely that the language of these urban Christians (and irst of all, its consonantism) was determinant for the development of EY rather than the dialect spoken by Germans in Silesia. For CzY, the closest dialects are also Bohemian and Silesian. Incompatible elements – non-Silesian {V5, V9, U1, M3, M4} and non-Bohemian {C1, C6} – represent a subset of those listed above for EY. For {V3} and {C16} – two features that distinguish CzY from EY – CzY simply follows local German dialect. If we compare early Jewish sources discussed in the previous section to modern Yiddish varieties, we can observe that all those printed in Prague or Kraków (and only they) are much closer to CzY than to any other modern Yiddish variety including EY, and, as discussed in the previous section, all of them are close to Bohemian dialect of German. This should not be interpreted to say that at the start of the 17th century EY was not yet existent. In principle, our sample of early Jewish sources can simply be non-representative for analyzing early stages of EY. Indeed, the only sources from the area that during the last centuries was a part of the EY territory are those from Kraków, that is, a city in the southwestern end of this territory whose Jewish community included a large number of Jews whose families came here from the Czech lands. The authors of the books in question could simply be of Czech origin. Since Kraków Jewish typography was wellknown, the authors of the books in question were not necessarily local Jews. In principle, they could also be, for example, from Bohemia or Moravia. In some works, we can also face an attempt made by their authors to follow certain norms of Jewish speech peculiar to Central Europe. For more insight into the genesis of features of EY, it would be important to ind early Jewish sources from other places in the Polish Kingdom or the Grand Duchy of Lithuania: Poznań, Lublin, Lwów and especially Brest and other northeastern communities. It is also worth noting that, as it could be expected, in comparison to other early Jewish sources, Kr, that is the most recent source from Kraków studied in this paper, shows the smallest number of features that distance it from modern EY25 and conforms to EY in a large number of other traits. The list of German dialects that are largely incompatible with both EGY and DuY encompasses CF, Bavarian, HA, and LA. EGY also shows important contrasts to EF, Swabian, and Hessian, and, to a lesser extent, PG, while the diferences between DuY and these four German dialects are less signiicant. For both Yiddish varieties in question, Bohemian and ECG appear as the least remote dialects. 24 Blosen (1986) shares the opinion about the important role that the Silesian dialect of German urban population in Polish towns played for the development of EY. 25 This list encompasses the /en/-relex of MHG –em {C15} and one innovation (preix uminstead of un-). Unity of the German component of Yiddish 123 Among their diferences in comparison to these German dialects one distinguishes several groups: (1) those enumerated above for CzY: non-Bohemian {C6, V3}, non-ECG {V3, V5, U1, M4}, and non-Silesian {V9}; (2) the diminutive suix –khen {M1} that contradicts to both Bohemian and Silesian, but is compatible with two other ECG dialects, Thuringian and UpS; (3) {C7} that for both DuY and EGY contrasts to both Thuringian and UpS; (4) {C1} that for EGY is compatible with Bohemian and Silesian only, while for DuY it is non-Bohemian; (5) {C15} that for EGY contrasts to both Thuringian and UpS; (6) {C12} that for DuY contrasts to both Bohemian and ECG being a feature peculiar to westernmost German dialects only. Globally speaking, one can observe that: (1) EGY is clearly related to eastern German dialects; and (2) DuY results from a fusion of elements that arose in distant areas, one part coming – as for EGY – from the East, but another part being of western origin. The information in the table presented at the beginning of this section shows that SWY is very close to EF, the only signiicant diference being /e/ before /š/ {V12}, a feature that is peculiar to all westernmost German dialects, that is, Alemannic and WCG. It could be due to the inluence of neighboring Swabian, the dialect to which SWY is also close enough but for a few important features (especially vocalic). Moreover, such fundamental family terms as SWY fra:le ‘grandmother’ and harle ‘grandfather’ are doubtlessly related to EF. On the other hand, SWY has signiicant diferences in comparison to LA (including basic vocalic characteristics {V1–V3}), Bohemian and PG (for both, in consonantism, vocalism, and morphology), and striking diferences with Bavarian, HA and all of CG dialects other than PG. Note that – as discussed in the previous section – the same link to EF and, to a lesser extent, Swabian was found in early western Jewish sources from the 16th century. Among them, the language of works by Elia Levita, Mel and BZR are particularly close to modern SWY.26 Yet, for Ox and H96 links to modern Yiddish varieties are less evident to be established. If we count only the most signiicant features, SWY appears to be closer to these two documents than other modern dialects. However, if we add to our comparison less important characteristics that still concern series of words and not individual words, the “distance” between H96 and DuY appears to be slightly smaller than it is between H96 and SWY, while for Ox the closest dialect becomes (maybe, fortuitously) CzY. The earliest Jewish sources – such as CC, R9, Le, R8 and Be – all have signiicant diferences in comparison to all modern Yiddish, still they are much closer to SWY than to other Yiddish varieties. SAB, the oldest and the northwestern most 26 If we exclude the unrounding {V10} that seems to be an innovation, the only important features that contradict to SWY are {C10} for Elia Levita’s works, {V7, U4} for Mel. 124 Alexander Beider of them, shows the larger number of elements contrasting to all dialects than other early sources. Its language is still closer to DuY than to any other modern Yiddish variety. 5 Classiication of Yiddish varieties Today the classiication generally used in works on Yiddish distinguishes two major subdivisions of modern Yiddish: EY and WY. The latter encompasses (among others) dialects for which in this paper the abbreviations SWY, DuY, EGY, and CzY were used. This classiication was initially introduced by the founder of modern Yiddish linguistics, Landau (1895) and later regularly used in Weinreich (1973). Landau based it on the relexes for MHG ei and ou that in what he deined as Western Yiddish appear as /a:/. In this paper, this feature is designated {V3}. Without any additional analysis about genetic links that can exist between Yiddish varieties, this classiication sounds purely conventional: it is not clear in which way the isogloss associated with these vowels is so signiicant that it can serve as an appropriate criterion for separating dialects. Table 7 lists linguistic elements that distance Yiddish varieties from each other. Table 7: Yiddish varieties: incompatible features EY CzY EGY DuY CzY EGY C16, V3 C1, V3, U4, M1 XXXXX XXXXX XXXXX XXXXX DuY C11, C12, V3, V8, U4, M1 C1, C6, C7, C11, C12, C21, V3, V4, V6, V12, U2, U4, M1, M6 XXXXX C1, C16, U4, M1, M10 C11, C12, C16, V8, U4, M1, M10 C6, C7, C11, C12, C16, C21, V4, V6, V12, U4, M1, M10 C1, C11, V8, XXXXX C1, C6, C7, C11, C21, V4, V6, V12, M1 C1, C6, C7, C11, V4, V6, V12, M1 SWY These data can be represented by the following linear chain: SWY – DuY – EGY – CzY – EY. This chain illustrates the links that exist between various dialects: the closer are elements within this chain, the closer are the corresponding dialects. It is worth noting that the “distance” between neighboring members of this chain varies. It Unity of the German component of Yiddish 125 is the shortest between CzY and EY. It is, by far, the longest between SWY and DuY: here the “distance” is similar to that between DuY and EY, the latter being situated on the other end of the chain! This information shows that the standard classiication is at least doubtful: CzY and EGY appear to be much closer to EY than to SWY, while DuY is signiicantly better correlated with EGY than with SWY. Moreover, as discussed in the previous section, SWY stays apart from other four dialects studied in this paper because of its close relationship to EF, while other dialects show kinship with Bohemian and ECG. A link between EGY and EY also follows from the earliest classiication of Yiddish varieties known to us, that by Friedrich (1784: 48–50). The author says that his own dialect (named EGY in this article) characterizes not only Jews from the main part of the Kingdom of Prussia, but also those from Great Poland.27 Additional arguments come from the consideration of the lexicon. Only a few words of Romance origin are shared by all Yiddish varieties: StY bentshn ‘to bless’, leyenen ‘to read’, tsholnt ‘a kind of Sabbath dish’ and a few others. Yet, a number of Romance lexical elements integrated to SWY are unknown or marginal not only in EY, but also in EGY and, to a lesser extent, CzY. Among them: or(e)n ‘to pray’, pülts(e)l / pilts(e)l ‘maiden’, brayen / prayen ‘to invite’ (all discussed in Section 3), memern ‘to commemorate the dead’, sargenes ‘shrouds’, porshen ‘to remove the forbidden fat and veins from meat, ‘to porge’, dormen ‘to sleep’, baven ‘to drink’, tetshen ‘to blow the Shofar in a synagogue’ and frimzelikh ‘noodle’ (compare Beranek 1965: 24–27, 80, 86, 92, 93).28 Their semantics and phonology imply that here we are dealing not with borrowings from French and/or Italian that could take place rather recently when Yiddish varieties from Central and Eastern Europe were formed already. Many of them belong to the substratum of SWY. In contrast to this Romance substratum of SWY, in EY, EGY and oten in CzY too, we ind a small number of old words of (mainly western) Slavic origin that are likely to be a part of the substratum of these dialects: treybern ‘to porge (meat)’ in contrast to western porshen, CzY deyde / EY and EGY zeyde/zayde ‘grandfather’ and bobe ‘grandmother’ (contrasting to SWY harle and frale, Beranek [1965: 100–101]), and, most likely, also StY beylik ‘white meat (of fowl)’, preydik ‘fore-quarter (of animal)’, srovetke ‘whey’ 27 He also distinguishes (1) “Swabian” Yiddish, also spoken in the Halberstadt area of Prussia, (2) Yiddish of the “Roman Emperor” provinces (that is, the Habsburg Empire), (3) Yiddish from Little Poland and Lithuania. In this article, they roughly correspond to SWY, CzY and EY, respectively. 28 Certain modern maps show the presence of these western forms in Prague contrasting to typical eastern forms found in Moravia (Beranek 1965: 92). Such geography implies that Prague, the largest community of this area, underwent during the last centuries an inluence from the West, while the language in Moravia kept more archaic variants. 126 Alexander Beider (Weinreich 1973: Vol. 2, 202–203), hoyl ‘bare, pure’, and par(e)v(e) ‘neither dairy nor meat (food)’ (Beranek 1965: 79; Eggers 1998: 140). A large number of other lexical isoglosses related to various components of Yiddish also place SWY apart from EY, EGY, and CzY.29 Phonological peculiarities in certain words from the Hebrew component also unify EY and CzY and separate them from SWY (see, for example, Beranek 1965: 36, 63–65). DuY appears to be a transitional dialect that underwent inluences from, on the one side, the East (principally, Bohemia and Poland) and, on the other side, the Jewish speech from the south-western German-speaking provinces.30 This intermediate position is closely related to the history of the Ashkenazic community of Amsterdam formed ater migrations of Jews from various parts of Europe: western Germany, Prague and Eastern Europe. In general linguistics, genetic links between various dialects or languages are oten established according to the criterion of shared innovations. If we take into account phonological features of Yiddish varieties that appear to be unrelated to those of surrounding German dialects, only one innovation seems to be panYiddish: the merging of short relexes of MHG ä, ë and e, but the absence of merging of long relexes for, on the one hand, MHG ä and ë and, on the other hand, MHG e {V7}. However, it is isolated from other phonological developments and its exact reasons and sources are obscure enough. As a result, it can not be taken for a doubtless example of innovation that took place among Jews of one particular region before spreading out to various parts of Europe. At any case, even if such common proto-development indeed took place, it concerned only one part of the system of the stressed vowels. As a whole, such systems – as discussed in Beider (2010) – were diferent for Proto-WY (the ancestor of SWY) and Proto-EY. These systems represent main phonetic innovations of Yiddish in comparison to German dialects. The Proto-EY system implies such important feature as merging of the relexes of MHG ei and ê in one phoneme and those for MHG ou and ô in another phoneme. Other innovations do not concern all Yiddish varieties as well. All of them are relatively recent dating from the period when Yiddish varieties seem to be formed already. The change of the negative preix un- to um- is shared by CzY and EY and is partly known in Alsace too. However, it is not found in DuY, EGY and the main part of SWY and, moreover, is unknown before the 17th cen- 29 See, for example, Beranek (1965: 23, 46, 74, 77, 78, 86, 89–91). Manaster Ramer (1997: 209– 210) makes a similar observation. In order to avoid misunderstanding provoked by inappropriate standard classiication, he uses the terms “Westerly Yiddish” and “Easterly Yiddish.” The former covers (among others) SWY and DuY. The latter encompasses (among others) EY, CzY and EGY. 30 Note that DuY uses, for example, the same words as SWY for ‘daddy’ and ‘mummy’, while a word from EY for ‘grandfather’. Unity of the German component of Yiddish 127 tury. For a number of other innovations known in modern EY we ind references in Jewish sources from Prague (17th to 18th centuries) and no references in EGY, DuY and SWY.31 The fronting /u/ > /y/ (> /i/) concerned only CzY, a part of EY (namely, CY and SEY), and is known from Jewish sources written in the 16th century in northern Italy and central Germany. To summarize the information presented above in this paragraph, one can say that the consideration of innovations realized by Jews in the domain of phonology demonstrates that there is no set of phonetic innovations that could be assigned to putative Proto-Yiddish from which all modern Yiddish varieties would be descendants. It also allows showing additional links between EY and CzY. No shared innovations are common to SWY, DuY and EGY. In morphology, the situation is slightly diferent. Yiddish varieties include a large number of nouns with no equivalent in German dialects, in which several German suixes are added to the root. As it was demonstrated in Timm (2005: 84–94), numerous words of this group are due to the tradition of biblical translations internal to Jewish communities.32 The plural suix –s (Timm 2005: 100–108) is also pan-Yiddish. It is in lexicon and semantics that peculiarities of the German component of Yiddish in comparison to German appear to be most important. In this domain, a large layer is due to the tradition of biblical translations internal to Jewish communities. Many of these characteristic elements appeared in the Middle Ages within the Rhenish communities and during the following centuries spread out from West to East. Other peculiar elements – some of which are also of medieval western origin – spread out inside of Jewish communities due to numerous migrations of rabbis and religious teachers. However, the importance of the lexical layer common to Yiddish varieties should not be exaggerated. Firstly, numerous dialectal lexical diferences can be observed too. In many cases, they are correlated with neighboring German dialects and, therefore, in principle, could correspond to relatively recent borrowings made by Jews from these German dialects. Still, for elements from the common layer too we do not ind doubtless evidence about their old presence in the vernacular language of Jews from various parts of Europe. References to some of them in biblical glossaries printed in Prague or Kraków during the 16th century are good testimonies to the propagation of the Ivri-taytsh tradition from West to 31 Among them: a regular change of sibilants /s/, /š/ and /z/ into africates (/ts/, /tš/ and /dz/ respectively) ater /n/ or /l/; the introduction of /d/ before /l/ when a diminutive suix -l is added to a stem ending in /n/. 32 Many of them appear already in R9 and during the following centuries spread out from West to East. 128 Alexander Beider East. Nevertheless, they are insuicient to draw a conclusion about these words being a part of the vernacular language already. Moreover, as noted in Section 3, in some aspects, these early works are non-representative of the language spoken at that time by Jews who lived on territories of central and eastern Poland, as well as in Grand Duchy of Lithuania. Secondly, a number of common elements in modern Yiddish varieties – unattested in early western sources – clearly result from mass migrations of Jews from Eastern Europe to the West that started ater the Cossack wars of the mid17th century. To them are likely to be due the presence in Yiddish varieties of Western Europe of such words as lokshn ‘noodle’, khotshe ‘at least’ and khape(n) ‘to catch’. Also note that we ind in DuY not only individual words but also morphological patterns using such diminutive suixes of Slavic origin as -nik, –tshik, –ke, and –enyu (Beem 1959: 23). The linguistic inluence of eastern communities was also indirect. For example, from a testimony written at the end of the 17th century, we know that Dutch Jews were regularly sending young men to Poland to learn, while numerous scholars and teachers from Poland were present in Western Europe (Landau 1901: 24). Thirdly, numerous features found in various modern Yiddish varieties and not present in standard literary NHG immediately cease to look as purely Jewish idiosyncrasies if we open German dialectal dictionaries. As discussed above, SWY is closely related to EF, while EY with Bohemian and – apparently during a more recent stage of its development – Silesian. These dialects are in many aspects intermediary between CG and UG. Bohemian was formed ater migrations of German colonists from the areas of Bavarian, EF and ECG (Schwarz 1962: map 59). Both EF and Silesian, in turn, were also inluenced by Bavarian. In this situation, oten a feature common to SWY and EY can actually be due not to the putative Proto-Yiddish but to a simple fact that they are (or were) present in both EF and Bohemian. Among the examples are {C17, U3, M3}.33 Fourthly, a large number of elements shared by modern Yiddish varieties that look peculiar if we compare them to modern NHG appear not to be innovations made by Jews: they rather reveal features that became archaic in the written language used by German Christians. Disappearing of some of them from NHG is due to the natural development of the standardized literary German language from which many forms judged to be too dialectal were gradually withdrawn and cer- 33 To this, one can also add features that are present in numerous High German dialects but absent from standard literary NHG: inal /x/ in the words cognate with NHG Höhe ‘height’ and Schuh ‘shoe’ (StY heykh and shukh); the variant nit ‘not’ (NHG nicht); un ‘and’ (NHG und); absence of inal –t in the 3rd singular of the verb ‘to be’ (StY iz, NHG ist). Unity of the German component of Yiddish 129 tain features of Low German were introduced. Other elements disappeared due to the extremely important role that the language of Martin Luther’s translation of the Bible (1534) played in the development of modern German. Numerous characteristics of his language were due to his native UpS dialect, while the spelling of certain words was due to his conscious eforts of retrieving forms that he considered to be etymologically correct (Timm 2005: 49–52). Until the late 18th century in Western Europe and the start of the 20th century in Eastern Europe Jews were not afected by these innovations in written German, and, as a result, many German archaisms are kept in modern Yiddish varieties. They should not be erroneously taken as evidence of the existence of Proto-Yiddish. Fithly, considering features common to Yiddish varieties, one should not be mislead by the conventional classiication of dialects used today in Yiddish linguistics. Here an interesting example comes from two lists presented by Timm of twenty expressions combining [noun (mainly from the layer of religious terms) + verb from the German component], one list in EY and another written in the vernacular language of a Jewish informant from Hamburg (Timm 2005: 8). Verbs used are the same in all but one of these expressions while, in principle, other verbs with a close meaning exist in Yiddish. As a result, the selection of identical verbs can not be fortuitous: it is clearly related to a common tradition. According to Timm, this comparison shows the unity of the German component of WY and EY and, therefore, of Yiddish as a whole. From a synchronic point of view this argument is convincing. However, it should not be extrapolated to draw such a general conclusion about the genesis of WY and EY. Indeed, here we are dealing with expressions directly related to the religious sphere that is a domain in which a kind of uniformity obtained can be easily explained not via common roots but via gradual changes brought about due to mobility of Ashkenazic rabbis and religious scholars. Moreover, the consideration of Hamburg as being representative for studying characteristics of WY appears to be inappropriate. As DuY, the local dialect is transitional between EY and WY. On the one hand, the variety of Yiddish spoken in that city exhibits such typical WY features as {V3, V12} (Beranek 1965: 2–3, 75). On the other hand, according to numerous other isoglosses, it appears closer to CzY and/or EY than to SWY (compare Beranek 1965: 12, 30, 58, 66, 67, 69, 80, 86,34 90, 95, 101), while for ‘daddy’ and ‘mummy’ in Hamburg (exactly as in EGY) one inds words from CzY and those from SWY too used as variants (Beranek 1965: 81–82). 34 This isogloss is particularly important for our discussion because it deals with one of expressions listed by Timm: ‘to blow Shofar’. The verb appearing in the testimony from Hamburg is cognate with StY blozn (also found in CzY) while the verb used in SWY is tetshen. 130 Alexander Beider 6 Conclusions In general linguistics, modern German and Yiddish are usually considered to be descendants of one common ancestor. This approach is quite natural: the German component of Yiddish is by far the most important, by the number of its elements and especially by its role in grammar, morphology, syntax, lexicon, and phonology of this language. During the initial development of Yiddish linguistics, this consideration was also without doubt: it characterizes studies by Landau, Sainéan, Gerzon, as well as the work of the most brilliant representative of this early “Germanistic” school, Bin-Nun (Fischer). However, today’s Yiddish linguistics also uses a parallel classiication that assigns Yiddish to a speciic group of Jewish languages. The elaboration of its linguistic aspects is mainly due to Weinreich (1973) who, without refusing the importance of the link between Yiddish and High German, insisted on deriving Yiddish from putative Romance-based speciically Jewish language(s) Laaz/Loez spoken by ancestors of Ashkenazic Jews in northern France and/or Italy. According to his model, German Jews never spoke the same language as their Christian neighbors because their language, from the very beginning, was a fusion language that merged Hebrew-Aramaic, Romance and – gradually growing in numbers – High German elements. Its immediate consequences were dramatic for Yiddish studies: (1) placement of the “birth” of Yiddish to the period – more than one thousand years before today – when the irst Jewish communities appeared in German-speaking territories; (2) determination of the Rhineland as the “cradle” of Yiddish; (3) automatic assignment of the label “Yiddish” to any German-based language/dialect spoken by Jews. During decades that followed ater Weinreich work was published, his approach contributed to the oblivion of the results achieved by the “Germanistic” school and to the concentration on “Judeo-centric” models. These new models mainly questioned some of the conclusions made by Weinreich,35 but not the essence of his approach. The importance of ideas by Weinreich can also be measured by the fact that they were of great inluence to the studies by two scholars who in many other aspects are really opposite to each other, Wexler (2002) and Timm. The former showed how one can develop Weinreich’s idea about an intimate link between various Jewish languages in the worst way creating a theory of “relexiication” whose arguments and methods of obtaining information contradict all major methodological principles elaborated by general linguistics during 35 For example, there was an attempt to displace the “cradle” for Yiddish from the Rhine to the Danube considering Yiddish to be unrelated to the language of Rhenish Jews (Faber and King 1984: King 1992, 1993; Katz 1993; Eggers 1998). Unity of the German component of Yiddish 131 the last two centuries. The latter took proit of innovative Weinreich’s ideas in the best way writing Timm (2005), a study, fundamental for Yiddish linguistics that shows the channel through which numerous speciically Jewish idiosyncrasies gradually became widespread in various Ashkenazic communities. This channel corresponds to Ivri-taytsh, that is, the tradition of biblical translations whose irst traces are found in medieval Rhenish communities. Certain of these idiosyncrasies have been linked by Timm to the tradition of biblical translations developed previously by French Jews (Timm 2005: 33–40). In other words, thanks to her opus magnum, something that before could be considered just Weinreich’s intuitive guess received a strong corroboration. In addition to elements that entered Yiddish through Ivri-taytsh, among other speciically Jewish pan-Yiddish elements one also counts: a layer of words related to the religion (at least, partly propagated thanks to the mobility of rabbis and religious scholars), a signiicant part of the Hebrew-Aramaic component, and numerous given names of Semitic, Romance or High German origin (Beider 2004: 233–237). It is important to stress that these elements of Yiddish oten have their roots on the Rhine, exactly as suggested by Weinreich. Are they really suicient to follow his idea of the Rhenish origins of Yiddish? In part, the answer is conventional because all depends on the deinition used. However, the question has substantive aspects too. According to major principles of general linguistics, the elements from the above layers, lexical and to a lesser extent morphological, can never be considered to form the basis for a language. This basis is typically High German in all fundamental aspects of the language. In this context, it would be inappropriate to consider Yiddish to be also a descendant of a putative Romancebased speciically Jewish language Laaz/Loez. In other words, the above elements could be of large importance for the gradual separation of Yiddish from German but in no case do they form Yiddish as a language. If during the period of this separation, High German would still be a totally uniform language, then Weinreich’s approach to origins of Yiddish could be valid. However, the situation is not that simple. Since the end of the First Millennium C.E., important dialectal diferences existed already inside High German separating it into Central and Upper German. During the irst centuries of the Second Millennium C.E. numerous subdivisions appeared inside Central and Upper German. We had no single evidence that Jewish vernacular speech was diferent from that of surrounding Christians by that time (but for the presence in it of certain speciically Jewish elements). Quite on the contrary, Jewish sources compiled in Western Europe before the 16th century show a close relationship to peculiarities of local Christian dialects, and, most important, one inds no evidence about any kind of homogeneity of these sources. In other terms, it would not be an exaggeration to say that to a great extent they represent German texts spelled using the Hebrew alphabet. Jewish texts 132 Alexander Beider compiled during the 16th century are diferent. Firstly, Jewish documents from that time fall into one of two rather homogeneous groups: western and eastern. The former covers documents from western German-speaking provinces and northern Italy and is closely related to East Franconian and, to a lesser extent, Swabian. The latter comprises texts from Prague and Kraków that exhibit features showing their link to the colonial Bohemian dialect of German. Secondly, for the 16th century we can be sure to deal already with idioms that are distinctly diferent from those used by local Christians: the Hebrew component is fully incorporated; one can observe numerous other speciically Jewish peculiarities as well the irst hints about the mixture of elements taken from diferent German dialects. Properly speaking, it would be inappropriate to see western and eastern Jewish idioms as dialects of one Judeo-German language, the ancestor of modern Yiddish. Such consideration would necessarily imply that both of them descend from the same common speciically Jewish ancestor (Proto-Yiddish) that in turn descends from one speciic German dialect. This is precisely false: German donor dialects were diferent for the two Jewish idioms in question. However, due to the presence in both western and eastern Jewish languages of numerous common elements, either inherited from the time when these languages were not separated from corresponding German dialects yet, or due to permanent exchanges between Jewish communities, as well as to closeness of High German dialects that served as basis for them, there certainly could be a feeling among Jews of the existence of a common Ashkenazic vernacular language. Modern Yiddish varieties descend from these two early Jewish idioms, western and eastern. As a result, it would be appropriate to speak about the existence of an independent Proto-WY (formed ater its separation from its East Franconian/Swabian German ancestor) and Proto-EY (formed ater its separation from its Bohemian ancestor). These independent Jewish proto-languages were suggested in Beider (2010) taking into account only the systems of stressed vowels of modern Yiddish varieties. The information presented in this paper ater a much more detailed analysis – allowing for a large inventory of additional phonological and numerous other features and their realization in early Jewish texts – corroborates not only the general idea of the existence of these protolanguages, but also their suggested approximate temporary frames. If in the sources of the 16th century these two subdivisions are already easily visible, it would be appropriate to place the creation of Proto-WY and Proto-EY to the 15th century, and at any case, ater the Black Death. The importance of Jews from Bohemia-Moravia for the development of Ashkenazic communities of Eastern Europe is relatively well corroborated by historical facts and therefore the Bohemian basis for Proto-EY is not a surprise. More enigmatic appears to be the process of gradual uniication of western Jewish communities under a Proto-WY with Unity of the German component of Yiddish 133 East Franconian/Swabian basis. Perhaps, a detailed historical study of the demography of medieval western communities could shed some light to this question. It would also be appropriate to address the question of geographic sources of Ashkenazic Jews who populated northern Italy during the 14th to 16th centuries. In these new territories on which local Christians spoke a diferent language, some kind of linguistic uniication of Jews coming from diferent German dialectal areas was unavoidable. Jews from northern Italy and their printing houses surely played an important role in the general cultural development of western Ashkenazic Jewry until the beginning of the 17th century. They could be of importance for the development of modern WY too. The classiication of Yiddish varieties currently used in Yiddish linguistics is inappropriate. Based on one conventionally chosen isogloss, it makes obscure the actual genetic links that exist between these Jewish idioms. Yiddish varieties spoken in Alsace, Switzerland, and the Rhineland (SWY) descend from Proto-WY. For all of them, the expression “WY dialects” is appropriate. EY dialects spoken in Eastern Europe and Yiddish in Czech lands (currently oten seen as a subdialect of WY) all descend from Proto-EY. During its development, EY (and especially the system of its consonants) underwent an important inluence of Silesian dialect spoken by German colonists in Polish towns. Similar changes are also found in Yiddish in Czech lands, most likely, through the intermediary of EY. Yiddish varieties spoken in East Germany and the Netherlands – that are usually also seen as subdialects of WY – actually are transitional idioms resulting from the fusion of elements coming from dialects descending from Proto-WY and Proto-EY, the latter being for East Germany clearly more important than the former. References Althaus, Hans Peter. 1971. Die Cambridger Löwenfabel von 1382. Berlin & New York: Walter de Gruyter. Banitt, Menahem. 2005. Le Glossaire de Leipzig. Vols. 1–4. Jerusalem: Académie nationale des sciences et des lettres d’Israël. Beem, Hartog. 1959. Jerŏsche. Jiddische Spreekwoorden en Zegswijzen uit het Nederlandse Taalgebied [Heritage: Yiddish proverbs and expressions from the Dutch linguistic area]. Assen: Van Gorcum–Prakke. Beider, Alexander. 2004. The birth of Yiddish and the paradigm of the Rhenish origin of Ashkenazic Jews. Revue des études juives 163. 193–244. Beider, Alexander. 2010. Yiddish proto-vowels and German dialects. Journal of Germanic Linguistics 22(1). 23–92. Beranek, Franz. 1965. Westjiddischer Sprachatlas. Marburg: N.G. Elwert (In this article, references are given to map numbers.). 134 Alexander Beider Bin-Nun, Jechiel. 1973. Jiddisch und die deutschen Mundarten: unter besonderer Berücksichtigung des ostgalizischen Jiddisch. Tübingen: Max Niemeyer. Birnbaum, Salomo A. 1979. Yiddish. A survey and a grammar. Toronto: University of Toronto Press. Blosen, Hans. 1986. Teilweise unorthodoxe Überlegungen zu einigen Problemen des Jiddischen. In Friedhelm Debus & Ernst Dittmer (eds.), Sandbjerg, 85. Dem Andenken von Heinrich Bach gewidmet, 161–187. Neumünster: Karl Wachholtz. Eggers, Eckhard. 1998. Sprachwandel und Sprachmischung im Jiddischen. Frankfurt: Peter Lang. Faber, Alice & Robert D. King. 1984. Yiddish and the settlement history of Ashkenazic Jews. The Mankind Quarterly 24. 393–425. Fischer, Hermann. 1904–1936. Schwäbisches Wörterbuch. 6 vols. Tübingen: H. Laupp. Frakes, Gerold C. 2004. Early Yiddish texts, 1100–1750. Oxford: Oxford University Press. Friedrich, Carl Wilhelm. 1784. Unterricht in der Judensprache und Schrit. Prenzlau: Ragoczy. Fuks, Leo (ed.). 1965. Das altjiddische Epos Melokîm-Bûk. 2 vols. Assen: Van Gorcum-Prakke. Gerzon, Jacob. 1902. Die jüdischdeutsche Sprache, eine grammatischlexikalische Untersuchung ihres deutschen Grundbestandes. Cologne: S. Salm. Gold, David L. 1986. Has the textbook explanation of the origins of Yiddish and Ashkenazic Jewry been challenged successfully? The Mankind Quarterly 26. 339–363. Grimm, Jacob & Wilhelm Grimm. 1854–1960. Deutsches Wörterbuch. 16 vols. Leipzig: S. Hirzel. Guggenheim-Grünberg, Florence. 1976. Wörterbuch zu Surbtaler Jiddisch (Beiträge zur Geschichte und Volkskunde der Juden in der Schweiz 11). Zürich: Juris. Hakkarainen, Heikki J. 1973. Studien zum Cambridger Codex T-S.10.K.22. Vol. 3: Lexikon. Helsinki: Suomalainen Tiedakatemia. Heide, Manfred Gernot. 1974. Graphematisch-phonematische Untersuchungen zum Altjiddischen. Der Vokalismus. Bern-Frankfurt-am-Main: Lang. Heide, Manfred Gernot. 1977. Die h-Graphen in älteren Jiddisch [H-graphs in older Yiddish]. In Hermann-Josef Müller & Walter Röll (eds.), Frägen des älteren Jiddisch, Kolloquium in Trier 1976, 4–15. Trier: University of Trier. Heilig, Otto. 1898. Grammatik der ostfränkischen Mundart des Taubergrundes und der Nachbarmundarten. Leipzig: Breitkopf. Jelinek, Franz. 1911. Mittelhochdeutsches Wörterbuch zu den deutschen Sprachdenkmälern Böhmens und der mährischen Städte Brünn, Iglau und Olmütz. XIII. bis XVI. Jahrhundert. Heidelberg: Carl Winter. Katz, Dovid. 1993. East and West, khes and shin and the origins of Yiddish. In Israel Bartal, Ezra Mendelsohn & Chava Turniansky (eds.), Studies in Jewish culture in honour of Chone Shmeruk, 9–37. Jerusalem: The Zalman Shazar Center for Jewish History. King, Robert D. 1992. Migration and linguistics as illustrated by Yiddish. In Edgar C. Polomé & Werner Winter (eds.), Reconstructing languages and cultures, 419–439. Berlin & New York: Walter de Gruyter. King, Robert D. 1993. Early Yiddish Vowel Systems: A Contribution by William G. Moulton to the Debate on the Origins of Yiddish. The Field of Yiddish. Fith Collection. David Goldberg (ed.). Evanston, IL: Northwestern University Press, pp. 87–98. König, Werner. 1978. Dtv-Atlas zur deutschen Sprache. München: Deutscher Taschenbuch. Krämer, Julius. 1965–1997. Pfälzisches Wörterbuch. 6 vols. Wiesbaden-Stuttgart: F. Steiner. Landau, Alfred. 1895. Das Deminutivum der galizisch-jüdischen Mundart. In Johann Nagl (ed.), Deutsche Mundarten, 46–58. Vienna: Carl Fromme. Unity of the German component of Yiddish 135 Landau, Alfred. 1901. Der Sprache der Memoiren Glückels von Hameln. Mitteilungen der Geselschat für jüdische Volkskunde 7. 20–67. Manaster Ramer, Alexis. 1997. The polygenesis of Western Yiddish – and the monogenesis of Yiddish. In Irén Hegedus, Peter A. Michalove & Alexis Manaster Ramer (eds.), IndoEuropean, Nostratic, and Beyond: Festschrit for Vitalij V. Shevoroshkin. Journal of Indo-European Studies (Monograph Number 22), 206–232. Washington, DC: Institute for the Study of Man. Manaster Ramer, Alexis & Meyer Wolf. 1996. Yiddish origins: The Austro-Bavarian problem. Folia Linguistica Historica 17. 193–209. Martin, Ernst & Hans Lienhart. 1899–1907. Wörterbuch der elsässischen Mundarten. 2 vols. Strasbourg: K.J. Trübner. Mieses, Mateusz. 1924. Die Jiddische Sprache: eine historische Grammatik des Idioms der integralen Juden Ost- und Mitteleuropas. Berlin & Vienna: B. Harz. Mitzka, Walther. 1963–1965. Schlesisches Wörterbuch. 3 vols. Berlin: Walter de Gruyter. Moser, Virgil. 1929. Frühneuhochdeutschen Grammatik. Band 1: Lautlehre. Teil 1: Orthographie, Betonung, Stammsilbevokale. Heidelberg: Carl Winter. Moser, Virgil. 1951. Frühneuhochdeutschen Grammatik. Band 1: Lautlehre. Teil 3: Konsonanten. Heidelberg: Carl Winter. Müller, Joseph. 1928–1971. Rheinisches Wörterbuch. 9 vols. Berlin & Bonn: Klopp. Müller-Fraureuth, Karl. 1911–1914. Wörterbuch der obesächsischen und erzgebirgischen Mundarten. 2 vols. Dresden: W. Baensch. Paul, Hermann. 1998. Mittelhochdeutsche Grammatik, 24th edn. Tübingen: Max Niemeyer. Röll, Walter (ed.). 2002. Die jiddischen Glossen des 14.–16. Jahrhunderts zum Buch >Hiob< in Handschritenabdruck und Transkription. Teil 1: Einleitung und Register, Teil 2: Edition. Tübingen: Max Niemeyer. Sainéan, Lazare. 1902. Essai sur le judéo-allemand et spécialement sur le dialecte parlé en Valachie. Mémoires de la Société de linguistique de Paris 12. 90–138, 177–196. Sauerbeck, Karl Otto. 1970. Grammatik des Frühneuhochdeutschen. Beiträge zur Laut- und Formenlehre, 1. Vokalismus der Nebensilben. Part 1. Heidelberg: Carl Winter. Schmeller, Johann Andreas. 1827–1837. Bayerisches Wörterbuch. 4 vols. Stuttgart & Tübingen: I.G. Cotta. Schnitzler, Leopold. 1966. Prager Judendeutsch. Ein Beitrag zur Erforschung des älteren Prager Judendeutsch in lautlicher und insbesondere in lexikalischer Beziehung. Gräfeling bei München: E. Gans. Schwarz, Ernst. 1962. Sudetendeutsche Sprachräume. München: Robert Lerche. Staub, Friedrich & Ludwig Tobler (ed.). 1881–2000. Schweizerisches Idiotikon. Wörterbuch der schweizerdeutschen Sprache. 16 vols. Frauenfeld: Huber. Stopp, Hugo. 1973. Grammatik des Frühneuhochdeutschen. Beiträge zur Laut- und Formenlehre, 1. Vokalismus der Nebensilben. Part 2. Heidelberg: Carl Winter. Stopp, Hugo. 1978. Grammatik des Frühneuhochdeutschen. Beiträge zur Laut- und Formenlehre, 1. Vokalismus der Nebensilben. Part 3: Die Entsprechungen nichthaupttoniger mhd. Vokale außer e. Heidelberg: Carl Winter. Timm, Erika. 1977. Jüddische Sprachmaterialien aus dem Jahre 1290: Die Glossen des Berner kleinen Aruch – Edition und Kommentar. In Hermann-Josef Müller & Walter Röll (eds.), Frägen des älteren Jiddisch, Kolloquium in Trier 1976, 16–34. Trier. University of Trier. Timm, Erika. 1987. Graphische und phonische Struktur des Westjiddischen. Tübingen: Max Niemeyer. 136 Alexander Beider Timm, Erika. 2005. Historische jiddische Semantik. Tübingen: Max Niemeyer. Timm, Erika (ed.). 1996. Paris und Wiene [Paris and Viena]. Tübingen: Max Niemeyer. Timm, Erika & Liliane Gehlen. 1996. The language of Paris un’ Viena. In Chone Shmeruk (ed.), Paris un’ Viena, 303–320. Jerusalem: Publications of the Israel Academy of Sciences and Humanities. Weill, Emmanuel. 1921. Le Yidisch Alsacien-Lorrain. Paris: Durlacher. Weinreich, Max. 1958. Bney hes un bney khes in ashkenaz: di problem – un vos zi lozt undz hern [Ashkenazic Bney hes and bney khes: the problem and what it teaches us]. In Shlomo Bickel & Leibush Lehrer (eds.), Shmuel Niger bukh, 101–123. New York: YIVO Institute for Jewish Research. Weinreich, Max. 1973. Geshikhte fun der yidisher shprakh [History of Yiddish language]. New York: YIVO Institute for Jewish Research. Weinreich, Max. 1993. Geschichte der jiddischen Sprachforschung. Atlanta, GA: Scholars press. Wexler, Paul. 2002. Two-tiered relexiication in Yiddish: Jews, Sorbs, Khazars and the Kiev-Polessian Dialect. Berlin & New York: Mouton de Gruyter. Žirmunskij, Viktor Mixajlovič. 1956. Nemeckaja dialektologija [German dialectology]. MoscowLeningrad: Akademija Nauk SSSR. Zivy, Arthur. 1966. Jüdisch-deutsche Sprichwörter und Redensarten. Basel: Victor Goldschmidt. Zuckerman, Richard. 1969. Alsace: an outpost of Western Yiddish. In Marvin I. Herzog, Wita Ravid & Uriel Weinreich (eds.), The ield of Yiddish. Third collection, 36–57. The Hague: Mouton.