— 71 — How Yiddish Polonisms may uncover Iranian, Turkic and other Asian speech in the White Croat State in Galicia (A chapter in the new field of »Silk Road Linguistics«) Paul Wexler Tel-Aviv University, IL 69978 Tel-Aviv, wexlerpaul7@gmail.com Govorci jidiša so kot trgovci na Svilni poti med 9. in 13. stoletjem prevze - li kakšnih pet tisoč afrazijskih, zlasti perzijskih jezikovnih posebnosti. Na Poljskem so se judovski govorci, ki so bili predvsem iranskega izvora, najprej naselili blizu ali znotraj galicijske Bele Hrvaške, ki je razpadla do 10. stoletja. Veliko nezapisanih iranskih in turkotatarskih vplivov v poljskem jidišu so nato prekrili polonizmi, ki v obliki in pomenu spominjajo na prvotno iransko in turkotatarsko izrazje. Članek se trudi raziskati, koliko polonizmov v polj - skem in ukrajinskem jidišu (pravzaprav ‘prikritih irano-turcizmov’) lahko obnovi podobo izumrle iranščine in turkotatarščine, govorjene na slovanskih in germanskih tleh. Druge možnosti za zdaj ni. As traders on the Silk Roads from the 9–13 th centuries, Yiddish speakers acquired about 5,000 Afro-Asian, mainly Persian, linguistic components. In Poland, these Jews, largely of Iranian origin themselves, first settled in or near the Galician White Croat State, which disappeared by the 10 th century. As unwritten Iranian and Turkic became obsolete, many such influences, also found in Polish Yiddish, came to be replaced by Polonisms, that resembled the original Iranian and Turkic terms in form and meaning. The present paper will explore how many Polonisms in Polish and Ukrainian Yiddish (essentially “covert Irano-Turkisms”) can reconstruct aspects of the extinct Iranian and Turkic once spoken in Slavic and German territories. So far, there is no other means available to us to accomplish this task. Ključne besede: poljski jidiš, iranščina, turkotatarščina, kitajščina, stari ju - dovski jeziki, ‘lingvistika Svilne poti’ Key words: Polish Yiddish; Iranian; Turkic; Chinese; old Jewish languages; “Silk Road linguistics” 1 Competing views of the origin of Yiddish It has long been suggested that Yiddish is a form of Middle High German that arose when Jews speaking Judaised variants of French and Italian emigrated 1.01 Izvirni znanstveni članek – 1.01 Original Scientific Article — 72 — Slavia Centralis 1/2022 Paul Wexler to the Rhineland and Bavaria in the 9–10 th centuries, adopting and adapting regional German dialects that came to be known as Yiddish. Yiddish became marginally Slavicised in eastern Germany (probably more so than the coterri - torial Christian dialects of German, as, apparently, some Sorbs, Polabians and other German Slavs converted to Judaism in order to participate in the Silk Road trade, largely in Jewish hands). After the 13 th century , Yiddish became increasingly Slavicised in the monolingual Western and Eastern Slavic lands where most Yiddish speakers came to reside. Hence, Yiddish should prove to be an exceptional laboratory for the study of Slavo-German linguistic and cultural contacts that began over a millennium ago—first with Polabian and Sorbian, and lastly with Russian in the late 18 th century. Yiddish has also been expected to provide unique insights into the histories of Middle High and early Modern German. I do not share these assumptions, mainly because I do not think that Yid - dish is a Germanic language (for the “old” and “new” views, see Weinreich 2008 and Wexler 2002, 2021, respectively). Its German component, limited almost entirely to the lexicon, is highly idiosyncratic, while Yiddish syntax and phonology are overwhelmingly Slavic. Hence, Yiddish can only be defined as a Slavic language, relexified to German and other superstratal lexicon (on relexification, see Wexler 1997 and other articles in the same volume). Impres - sionistically speaking, about a little over one third of the Yiddish vocabulary is German, but a similar proportion is artificial “Germanoidisms”, created by Yiddish speakers, which either do not exist at all in German, or exist in German with different meanings. Approximately half of the so-called Hebra - isms are “Hebroidisms” created by Yiddish speakers—on the model of the inventory of Arabic cognates used in Persian (which renders Yiddish more Arabicised than the Judeo-Ibero-Romance languages spoken in Spain and Portugal, which are famous for being far poorer in Arabisms than any dialect of Christian Ibero-Romance!). Yiddish is, pointedly, almost entirely free of idiosyncratic Slavoidisms. These facts mean that Yiddish is a uniquely merged Eastern and Western Slavic language (Upper Sorbian and pre-Ukrainian, i.e. Kiev-Polissian-Galician), where most of the lexicon has been “relexified” to German(oid) elements, according to Slavic and Iranian, and, to a lesser extent, also Turkic patterns of discourse. Recently, I have uncovered a massive overt and covert Asian (mainly Iranian, Turkic, Chinese, and even a tiny Tokharian and Mongolian element), and a minor overt African impact on Yiddish (see Wexler 2021), reflecting the routes of peripatetic Jewish merchants along the Silk Roads. Yiddish was invented (along with other Jewish languages) probably by the 9–10 th centuries to be a secretive language for Jewish merchants operat - ing on the Afro-Eurasian Silk Roads from the 9–13 th centuries. It therefore goes without saying that the idosyncratic nature of Yiddish Germanoidisms makes it very dangerous to extrapolate from Yiddish to German. Still, Yiddish will continue to be of major interest to Germanists, as well as to Slavists, but in entirely new ways: As a heavily Iranianised and mildly — 73 — How Yiddish Polonisms may uncover Iranian, Turkic and other Asian speech … Turkicised and Sinicised Slavic language. As such, Yiddish may assist in iden - tifying hitherto unrecognised (i) Iranianisms (mainly Persianisms?) and Turki - cisms (probably from the closely related Khazar and Chuvash), and even Sini - cisms and a few Mongolianisms, in Slavic languages, and the same components, along with Slavisms, in German, and (ii) The Slavic and German locales where Iranian and Turkic (and Chinese and Romani?) were probably spoken, by mixed Slavo-Turkic and Slavo-Iranian confederations (some involving Mongolians), at the time of the Jewish settlement in these lands by the 9 th century. Iranian and Turkic probably became obsolete in Europe about two centuries later. Romani, an Iranianised Indic language, apparently spoken by staff working on Jewish caravans along the Silk Roads, can also assist in identifying overt and covert Iranianisms in Slavic languages and German. The Silk Road may account for the geographical spread of Roma throughout Eurasia. These topics have yet to be raised and explored by Romanologists. 2 The goals of the present paper The present paper will try to reconstruct the Iranian, Turkic and other Asian elements very likely spoken in the Irano- and Turco-Slavic confederations in the White Croat State in Polono-Ukrainian Galicia. There is also the possibility that Chinese, Romani and Mongolian merchants were also active in Polish and Ukrainian Galicia, and, possibly, also resided in Irano- and Turco-Slavic confederations. Mongolian skeletons have been found in an Avar cemetery in northern Serbia, thought to date from the 8–9 th centuries (at Čelarevo, in the Vojvodina, to the north of Novi Sad; Avars were a Turkic people probably speaking Iranian); stones with rudimentary Hebrew inscriptions and religious decorations and inscriptions have also been found at the same site (they can be viewed in the historical museum in Novi Sad). The Mongolians may have been long-distance merchants who had converted to Ju - daism. Chinese merchants were encountered less frequently on the Silk Roads a millennium ago than Iranians (Zoroastrians, Muslims, Jews), but my recent discovery of a wealth of likely Chinese influences in Yiddish (less so in other cryptic Old Jewish mercantile languages) means that the possible presence of (Judaised?) Chinese merchants in Galicia requires further study—by historians and linguists alike. In many of the files presented below, I have cited Chinese terms with strong formal and semantic resemblance to Iranian, Turkic and/or Arabic (synonyms, or semantically close terms). I have always assumed that China was the most likely venue for the diffusion of such terms to Old Jewish languages. However, I should not rule out the possibility that the Sinicisms were acquired by Yiddish on Slavic territory. A Chinese, and possibly also Romani presence in Slavic or German territories, could most likely come from converts to Judaism who were partners of Iranian Jews who also worked on the caravans that plied the Silk Roads. In file #13 below in the discussion of a — 74 — Slavia Centralis 1/2022 Paul Wexler Yiddish bread term ( xale) with Chinese, Iranian, and/or Arabic connections, I emphasised that the Yiddish term (believed erroneously by speakers of Yiddish to be of Hebrew origin), as well as possible surface cognates in Arabic and Persian, has, historically, never appeared, to the best of my knowledge, with a diminutive suffix—but such a suffix is possible in some dialects of Mandarin (e.g. in Běijíng), and is required in the Polish loan, chałka. This crucial fact suggests that Polish may have acquired the term (attested from Chinese all the way to German!) directly either from Jewish or non-Jewish merchants active in Przemyśl/Peremyšl (the earliest known Jewish mercantile settlement in Galicia in the 10 th century). If Yiddish never had a diminutive option, the reason could be that Yiddish speakers received the term in its non-Chinese form. For Romani loans of Hebrew origin, see Wexler 2021. An examination of Polonisms in Romani might reveal whether Romani merchants may have settled in Galicia in this earlier period (I discussed detailed parallels in the Polish components shared by Yiddish and Romani in Wexler 1986). If so, then some Roma may have left India and Iran earlier than is usually presumed. In recent years, I have discovered that Yiddish has over 4,500 Afro-Asian components, both overt and covert, on all levels of the language, though pri - marily in the lexicon. The main Asian contributor to Yiddish is Persian. The two main reasons for this fact are: (i) Until c.1200, the overwhelming major - ity of Jews in the world (consisting mainly of converts of diverse ethnicities) resided in the Iranian Empire (where the Iranian Talmud, composed largely in Judeo-Eastern Aramaic—also the administrative language of the Iranian Empire for centuries—was compiled by the 5–6 th centuries). (ii) Jews acquired special trading privileges in the late 9 th century from the Holy Roman Empire and the Táng Dynasty, that allowed them to circulate freely on the Silk Roads connecting Mainz and Xī’ān, Andalusia, Africa, the Near East and India (and, marginally, Indonesia). The Jews rose to prominence as peripatetic merchants because they were neutral in the religious conflicts that developed by the 9–10 th centuries between Islam and Christianity, Buddhism and Hinduism, and they lacked a proselytising or political agenda. The prominent role of the Jews on the Silk Roads required the invention of cryptic colloquial languages, which were invented by the process of relexification (a technique originally used hitherto by them, but only in the production of literal Hebrew Bible “translations”; in practising widespread colloquial relexification they were probably follow - ing Iranians and Arabs; a primary reason for the massive disparities among contemporary Arabic dialects is that the original local languages, e.g. Coptic, Berber and Aramaic, were first relexified to Arabic lexicon; technically, the results would not have been Arabic, but “relexified Coptic”, etc.). Non-Jews could understand relexified Old Jewish languages with great difficulty. With - out the Silk Roads, the Jewish languages would not have been invented, and widespread conversion to Judaism by mainly pagans interested in participating in the lucrative Silk Road trade would not have taken place. The very survival of the Jews would have been unlikely after 1200 without the Silk Roads. — 75 — How Yiddish Polonisms may uncover Iranian, Turkic and other Asian speech … The evidence that the Jews arose as a commercial guild with a unique religion in Iran is diverse. Aside from the plethora of Iranian legal concepts and terminology in the Talmud, and Iranian customs practised by the Jews worldwide, I should point out that the two Hebrew ethnonyms, Ashkenaz(ic) and Sefard(ic), which are used by almost 90% of the world’s Jews, are themselves of Iranian(ised) origin and mean, respectively, ‘Scythian’, a geographically widespread Iranian tribe, and Sardes, the name of a Greco-Irano-Lydian city in western Anatolia, near Izmir. The remaining Jews in the world do not have unique ethnonyms. Moreover, Ashkenaz(ic), originally cited in the Bible as an Iranian Scythian territory near Armenia, in the late first millennium A.D. first came to denote Iranian-, and then, Slavic-speaking Jews, in the Near East and Western Europe. Since the late 11 th century in Europe, the term Ashkenazic has denoted Yiddish speakers in the German lands. Ashkenaz(ic) was still being used natively to denote Iranian-speaking Jews in Uzbekistan in the 14 th century, and in the Caucasus and Crimea as recently as the early 20 th century. The Ashkenazic Jews historically spoke Iranian, Slavic and Slavic relexified to German, Hebrew and other lexicon—in that order—which came to be known as Yiddish; the Sefardic Jews, who abandoned Sardes after its destruction in the early 7 th century in a war between the Byzantine and Iranian Empires, settled in Andalusia and North Africa, where they invented Judaised variants of Arabic, Berber and, eventually, Ibero-Romance. The first international Jewish trading guild known to have operated on the Silk Roads was from Rādhān, in southeastern Iraq, in the Iranian Empire, by the 9 th century, and was known as the Radhanites. The Iranians played a major role on the Silk Roads, due to their central geographical location. Southeastern Poland (Małopolska) and the neighbouring Western Ukrainian lands were on the Silk Roads, and it was in this part of Poland that Jewish merchants first settled in the 9–10 th -centuries. The first international cryptic Jewish merchant language was Persian, which was known in the 10 th century as Lotera’i (interpreted by Jews as Hebrew ‘not the Torah [language]’, due to the plethora of Hebraisms and artificial Hebroidisms). The glottonym, however, also used by non-Jewish Iranian merchants (who shared the secret language with Jews), is probably ultimately of Indic origin. More likely etyma are Hi/ Ur lutrā 1 ‘exiled, dumped, driven away, rejected; sycophant; babbler, tell-tale, 1 The following abbreviations are used with examples: ag.—agentive, Ar—Aramaic, Arab—Arabic, Arm—Armenian, Ashk—Ashkenazic, Aves—Avestan (the language of Zoroastrian scripture), Balk—Balkan, Bojk—Bojkian (a Ukrainian ethnographic group in the Carpathian Mountains of Ukraine, Slovakia, Hungary and Poland), Br—Bela - rusian, Bulg—Bulgarian, c.—century; circa, C—Common, Cant—Cantonese, Carp— Carpathian, Ce—Central, ch.—chapter, Chin—Chinese, Chr—Christian, Chuv—Chu - vash (a Turkic language, closely related to Khazar and proto-Bolgar), Cl—Classical, Cri—Crimean, Cz—Czech, d.—died, D—Digor dialect of Ossete, dial.—dialectal, dim.—diminutive, Džud—Džudezmo (Judeo-Spanish), Džuh—Džuhuri (a Judeo- Iranian language spoken in Daghestan and Azerbajdzan), E—Eastern, ed.—edited, — 76 — Slavia Centralis 1/2022 Paul Wexler blab; silly person; backbiter, slanderer; mischief-maker’, possibly related to Skt lótra ‘booty , loot’ ( see P latts 1 960: 95 1 ; Mayrho fer 1 96 7 , fasc. 1 9 : 1 1 4 ) . Turner cites * luṭṭati ‘plunders’; loptra- ‘stolen property’ (1966: 644, #11078; 649, #11145). Bailey (1979: 24) also suggests Hi lūṭ ‘robbery, plundering’ (< Skt lótra ‘booty’; this is the source of Anglo-Ind loot ‘booty’ [1788]); see also Hi/Ur luṭnā ‘to become entangled/lean/weak/poor; degenerate; to deteriorate’, lūṭijā ‘robber’. There are formally and semantically similar words in Chinese that could either be sources of (J)Pe lotera’i or derivatives of the Indo-Iranian glottonym, thus creating an isogloss that links Persian with Chinese and German, see e.g. OCantChin lau ‘to chatter’, lo ‘to prattle; dribble’, lö’ ‘to confuse’, löt ‘to ramble’; see also the first component of ShangChin lóe gāw ‘to talk nonsense, prattle on’ (with ‘to talk’). See also MandChin láo ‘to chatter’, lào ‘to gossip’, (dial.) ‘to chat’ (the latter two terms share a common character), lán ‘confused chatter, incomprehensible babble’, fèi liào ‘good-for-nothing’ (lit. ‘waste prod - ucts’), wú lài ‘rascal, rogue, scoundrel’ (lit. ‘not a rascal’), liú làng zhě ‘vagrant, vagabond’ (lit. ‘degenerate’ + ‘unrestrained’ + ‘person involved in’), lüè ‘to rob, plunder’, lĭ yŭ ‘slang’ (lit. ‘vulgar, unrefined’ + ‘language, dialect, speech’), lŭ lüè ‘to loot, pillage’ (with lŭ ‘to capture, seize’). In my view, the “Indo-Sino-Iranian” glottonym is also the basis of G Lotter(bube) ‘sluggard’, Uk lotr ‘scoundrel’, Pol lotr ‘ sco un dre l ; thi e f’ ( see further details in Wexler 2021, ch. 2, sect. #264). In the German lands, Jews used the term He lwtjr /loter/ to denote the areas where Judeo-Persian was Egyp—Egyptian, Eng—English, f—feminine, fasc.—fascicle, fl.—flourished, fn.— footnote, Fr—French, G—German, Geor—Georgian, Gk—Greek, G’oid—German - oid (i.e. artifical German), H—High, He—Hebrew, Hi—Hindi, Hung—Hungarian, I—Iron dialect of Ossete, Ind—Indic, Iran—Iranian, J—Judeo-, Jagh—Jaghnobi (an Eastern Iranian language spoken in Tadjikistan, the sole survivor of Soghdian), Jap— Japanese, Jazgh—Jazghulami (a Southeastern Iranian language, spoken in Tadjikistan), Kar—Karaite, Khot—Khotan (a major oasis town in southwestern Xīnjiāng, China), Kor—Korean, Kurd—Kurdish, L—Low(er), Lat—Latin, Latv—Latvian, Lemk— Lemkian (an ethnic group in the Transcarpathian Ukraine, southeastern Poland and northeastern Slovakia, on and near the location of the White Croat State), lit.—literally, Lith—Lithuanian, m—masculine, M—Middle, Mand—Mandarin, Mod—Modern, Mong—Mongolian, N—North(ern), O—Old, Oss—Ossete (an Eastern Iranian lan - guage, spoken in the Caucasus), Ott—Ottoman, Pal—Palestinian, Part—Parthian (an extinct Northwestern Iranian language, spoken in Iran and Turkmenistan), Pe—Persian, pl.—plural, plt.—plurale tantum, Pol—Polish, R—Russian, Rom—Romani, Rtw—Rot - welsch (German slang), S—Southern, Se—Serbian, sect.—section, Shang—Shanghai, Skt—Sanskrit, Sl—Slavic, Sogh—Soghdian (an Eastern Iranian language, spoken in Uzbekistan, Tadjikistan and China until the 8c A.D. and the major lingua franca of the Eastern Silk Roads until the arrival of Persian), st—standard, Syrc—Syriac (a dialect of Aramaic), Talm—Talmudic, Tat—Tatar, Tc—Turkic, trsl. —translated, Tu—Turkish, Uk—Ukrainian, Ur—Urdu, USo—Upper Sorbian, V—Vulgar, W—Western, Y—Yid - dish, Yem—Yemeni, ZoroPahl—Zoroastrian Pahlavi (Middle Persian spoken up until the 7c. A.D.). My transcription of Modern Persian follows Rubinčik 1970. — 77 — How Yiddish Polonisms may uncover Iranian, Turkic and other Asian speech … spoken, though all scholars mistakenly believe that it denoted G Lothringen, Fr Lorraine. This is unlikely , since the Hebraism denoted various areas that did not overlap with the non-Jewish named areas. The Radhanite guild must have been multiethnic and multilingual almost from its very inception. As such, the Radhanites resembled the many Irano-Turkic professional confederations that migrated from Asia to Europe in the course of the first millennium A.D. and intermarried with Slavs. In the non-Judaic confederations, the Iranians or Turks were the minority elite, while Slavs were the subject majority; eventually, the minority assimilated to the Slavic majority. Very likely the Irano-Turkic elites imparted linguistic, religious and cultural influences to the pagan Slavs before assimilating. This is a new topic for Slavic linguistics, which Yiddish helps to explore. The only members of the mixed confederations to survive to our day with a unique language are the Jews and the Roma. The ethnically and linguistically varied Jewish confederations that settled in the Slavic and German lands probably retained Irano-Turkic speech longer than the coterritorial non-Jewish confederations—for purposes of secrecy and professional group closure, and because they alone, as international merchants, travelled back and forth to their ancestral lands in Asia. In addition to creating the first unique Jewish languages towards the end of the first millennium A.D., the Radhanites also created a unique written and partly spoken Hebroid norm for common use by all the linguistically diverse merchants, by relexifying their native languages to Hebrew lexicon and by inventing Hebroidisms, on the model of various languages. In the case of Yiddish, the models for Germanoidisms were mainly Slavic, but also included Iranian and Perso-Arabic, Turkic and even Chinese patterns of discourse; the models for the invented Hebroidisms in Yid - dish and the selection of genuine Semitic Hebraisms were determined mainly by the cognate Arabisms embedded in Persian. Much of this Hebrew/Hebroid corpus of spoken and written Yiddish has also been accepted by Modern Israeli Hebrew, though modern-day language planners were, officially, reluctant to do so. The rather large Hebrew/Hebroid component of Yiddish had to be engineered in common with the Hebrew/Hebroid component of other Old Jewish cryptic trade laguages. Hence, it is not surprising that patterns of discourse for the use of secretive components could be licensed simultaneously by a number of the native languages of the merchants. An open question is whether caravans tended to be manned primarily by speakers of the same language, or by merchants fluent in different languages. The latter alternative is more inviting, since, if merchants made the full trip between Germany and China (as I believe they did), there would have been a need for considerable linguistic competence. Ac - cording to the eye-witness account of the Jewish merchants in Baghdad made in the late 9 th century by a local Persian official (ibn Khordadhbeh, d.c.912; see details in my 2021, ch. 1.7), the merchants could converse in about six major Silk Road languages. Since some of the glottonyms are ambiguous, I presume ibn Khordadhbeh subsumed several languages under some glottonyms—thus giving us about 12 languages in all. — 78 — Slavia Centralis 1/2022 Paul Wexler 3 Probable Irano-Turkic settlements in the German and Slavic lands There are altogether nine means for uncovering the locations of Irano-Turkic settlements in the German and Slavic lands: (i) Archaeological data. (ii) The use of a surviving non-Slavic ethno-glottonym (e.g. Sorb, Serb, Croat, Czech, Obodrite, etc.). (iii) Historical descriptions of confederations (e.g. the Irano- Slavic “White Croat State”; see details below). (iv) Slavic ethnographic prac - tices of Irano-Turkic origin (e.g. wedding customs). (v) Toponyms denoting the presence of Jewish traders’ way stations. (vi) Evidence of linguistic geography (e.g. the Eastern Slavic “pseudo-dual” has parallels in Bulgarian and Iranian). (vii) Slavic beliefs in Iranian forebears (e.g. Sorbian slaves in 11 th -century Spain claimed they were descended from Persians—suggesting that they came from still extant Irano-Slavic confederations in their German homeland). (viii) The relative volume of Iranianisms in the coterritorial Slavic languages and Ger - man (e.g. Polish has a particularly large Iranian inventory, possibly because of the presence of the White Croat State). (ix) Known mass conversions (prob - ably conducted superficially by merchants) of pagans to Judaism, in order to avoid slavery being imposed by German and Turkic slavers, and in order to participate in the lucrative Silk Road trade, which was a quasi-monopoly of peripatetic Jewish merchants. To the best of my knowledge, the White Croat State and similar states have never been examined by historians of the Jews or Slavic languages; yet, the White Croat State is especially important for Jewish, Slavic and Iranian lin - guistics and ethnologies. It is mentioned by (i) The Byzantine ruler, Constantine Porphyrogenitus (b.905–d.959, who writes that White Croatia consisted of parts of Bohemia, Poland and Lusatia). (ii) King Alfred the Great of England (fl. 897–901) writes of the (Lat) Horithi ‘Croats’ in the Bohemian mountains. (iii) The so-called Bavarian Geographer (9 th c.). (iv) Arabic sources (e.g. ibn Rusta, 10 th c., using data from 842–847; Gardizi and al-Mas‘udi, 10 th c.: see Labuda 1949: 194–262; Czekanowski 1957: 194, 205, 239, 252, 319, 338, 360–363). Pivtorak (1993: 87) believes the Carpathian and Bojkian dialects of Ukrainian developed on a substratal Iranian White Croat dialect. In future, a detailed com - parison of these Ukrainian dialects and Yiddish would be very much welcome. Ibrāhīm ibn Ja’qūb, a 10 th -century Catalan Jewish merchant and diplomat from Tortosa, visited the White Croats in Galicia, and knew the Iranian name of the White Croats. He may have been of Iranian origin who knew their name from his homeland (Tortosa, the Near East?), or learned it from the White Croats, some of whom apparently still spoke Iranian at the time of his visit to White Croatia (see Łowmiański 1963, 2: 61–62, 117). An Andalusian Hebroid source of the 10 th century shows that Jewish emissaries from the White Croat State had visited Andalusia. I propose to follow the etymology suggested by Vas - mer (1923: 118ff; 1958, 3: 261) and Shevelov (1965: 483): OIran (fšu)haurvatā ‘herdsman’. A form closest to the Vasmer-Shevelov reconstruction is found only in the Arabic writings of Ibrāhīm ibn Ja’qūb: ḥ’ jr w’s /ḥirwās, ḥajrawās/. — 79 — How Yiddish Polonisms may uncover Iranian, Turkic and other Asian speech … The name ‘Croat’ appears in a variety of forms in Arabic-language reports, see ḥ(u)ruās(i)ja in the writings o f the North African al-Idrīsī 1 2c. (Lewicki 1954, 2: 56; see also Dvorník 1949: 271, fn. 15, who supports an Iranian origin of the name Croat, see ibid.: 279–286, 309–311; Kmietowicz 1978: 24–5 gives variant spellings). See also item #15 below. The importance of the White Croat State for Slavic (including Yiddish) studies stems from my belief that it may have been one of the remaining en - claves of Iranian speech in Europe towards the end of the first millennium, and that it was located along the Irano-Arab Trade Roads. The Radhanite merchants, who settled in this part of Poland first, would have resembled the White Croats by being bilingual in Slavic and Iranian (though not necessar - ily in the same Iranian and Slavic languages), in addition to speaking other, unshared, languages. It is a striking fact that the dialects of Yiddish spoken in Ukraine reveal the largest Slavic component of all the Yiddish dialects, primarily from lo - cal Eastern Slavic dialects, while Polish Yiddish has more Eastern Slavic than Polish components. Though Yiddish may well have been first invented in the Khazar Empire (where pre-Eastern Slavic was also very widespread), and, subsequently, became concentrated in the Slavic and German lands to the west, and obsolete in the Caucasus, it would be imprecise to ascribe the large Eastern Slavic component of Polish Yiddish to original westward mi - grations of Iranian Jewish merchants involved in the Silk Road trade. This is because, for several centuries, Jewish Silk Road merchants were travel - ling back and forth across Eurasia, and not just migrating to the west, and there is also a small Polish component in Ukrainian Yiddish, especially in dialects neighbouring Poland. There are several plausible explanations for the presence of a relatively small quantity of shared Polonisms in Polish an d Ukraini an Y i ddis h di al ects : ( i ) S o m e P o l o nisms co ul d ha v e spread to Ukrainian Jews in the wake of the political and cultural Polonisation of the Belarusian, Ukrainian and Lithuanian lands; however, these influences would date from several centuries after the demise of a quasi-Jewish monopoly on the Silk Roads. (ii) In order to reduce to a minimum regional variants of Yiddish that might arise from differences in form and meaning characteris - ing cognates within disintegrating Western and Eastern Common Slavic (e.g. between pre-Sorbian and pre-Ukrainian), the two original Slavic languages of Jewish merchants, there is some indication that merchants took Polonisms as a neutral choice—possibly given the central position of Polish speech. (iii) I find that a number of the (new?) Polonisms embedded in Polish and neighbouring Galician Yiddish, may have replaced Yiddish Iranianisms and Turkicisms with which they shared form and meaning. I suspect that such Polonisms were intended to be (cryptic?) replacements for Iranianisms that were either poorly or no longer understood by many peripatetic Jewish Silk Road merchants, speaking languages other than Turkic and Iranian. If my assumption is correct, then the largely unexpected Polonisms in Polish and — 80 — Slavia Centralis 1/2022 Paul Wexler Ukrainian Yiddish could give us clues to the nature of the Iranian and Turkic speech used in the Irano-Slavic confederations established on the Slavic and German lands. Obviously, we cannot know with any precision which Iranian languages were spoken in the White Croat and other confederations; possible candidates could have been Western Iranian dialects such as Persian and Median, or Eastern Iranian Ossete. Once we have an exhaustive collection of Yiddish Iranianisms and Turkisms (and maybe Sinisms) reconstructed from Polish, we will need to make similar investigations of overt and covert Iranian and Turkic influences in the coterritorial non-Jewish Slavic speech and in the other Old Jewish Silk Road languages. I believe that Polonisms adopted by the local Yiddish speakers in Polish Galicia as replacements for the local obsolescent (Judeo-)Iranian and Turkic, could have been circulated among non-Galician Polish (and Ukrainian) Yid - dish dialects by the Radhanite merchants. Most of the pan-Yiddish Polonisms were most likely motivated by underlying Iranian (and other Asian) semantic, lexical and/or phonological consideratons, and definitely did not need to have been cognates of Polonisms. The new Yiddish Polonisms would have retained some cryptic quality in the Yiddish of Ukraine and Belarus’. There is a serious methodological problem confronting us. In addition to White Croat Iranianisms, Yiddish dialects have a much larger Iranian and other Asian component (e.g. from Turkic languages, Mongolian, Tokharian, Indian languages, Chinese, and even some Africanisms—both overt and covert), col - lected by the peripatetic Jewish merchants. Furthermore, not all Iranianisms in Galicia would have been replaced by similar-sounding Polonisms in the local Yiddish speech. Some of the obsolescent Iranianisms could have been retained outright by Galician Yiddish. In the near future, we will need to look for ways of distinguishing between Asianisms acquired on the territory of Polono- Ukrainian Galicia and those brought to the latter territories by the peripatetic Jewish merchants from points much further east. 4 The term ‘Bird’s milk’ in Afro-Eurasian languages Before we look at Iranianisms that may well have triggered the adoption of similar, mainly non-cognate, Polonisms in Galician Yiddish (and which may have spread marginally to Ukrainian dialects to the east), it would be useful to explore an example of a Eurasian isogloss shared by a very large variety of languages, both Jewish and non-Jewish, e.g. Slavic, Turkic, Iranian, Arabic, Aramaic, Armenian, Latin, Greek, Rumanian, Hungarian, Hindi/Urdu, Finno- Ugric, Chinese, and some Caucasian languages: ‘Bird’s milk’. In this example, all target languages use native elements. The reader will quickly see that the Yiddish form could, thus, have been licensed by various individual languages, and maybe even by a set of languages. The etymological and methodological problems are compounded by the absence of historical documentation in most — 81 — How Yiddish Polonisms may uncover Iranian, Turkic and other Asian speech … languages. However, the expanse of the isogloss suggests that the term devel - oped in a cryptic mercantile, possibly Jewish, language. ‘Bird’s milk’ is an expression found in most Slavic languages (but not, at present, in Upper Sorbian). Some scholars have suggested that the componen - tially identical (i) Y fojglmilx is an exact parallel of the Gk tōu pouliōu to gála that, in both languages, has the metaphorical meanings ‘unimaginable wealth, something fantastic, very rare’, replacing an older synonymous ClGk ὀrniθōn gála ‘ an y marv e ll o us dain ty ; g ood f o rtun e ’ , w hi c h is f o un d in Aristop han es (5–4 th c. B.C.: for the relevant passage in his play Wasps, see MacDowell 1971: 202) and (ii) That Greek-speaking Judaists may have been the carriers of the expression to non-Mediterranean European languages (Dieterich 1931: 337, fn. 3; M. Weinreich 1956: 629; 1973, 3: 72, 4: 261; Gerhardt 1974–1977; Wexler 1992: 93–94). (iii) Also in support of a Greek source for the nearly pan-Slavic and Yiddish terms is the fact that languages exposed to Greek cultural and linguistic influence, or to Hellenised Old Church Slavonic, also have the expres - sion, see e.g. Bulg ptiče mleko, Se ptičje mleko, Tu kuş sütü in the Balkans, and Br ptušynae malako, Uk ptaše/ptašyne/ptyče moloko (and ptašynoho moloka jomu lyš brakuje ‘there is nothing more he can desire’, lit. ‘bird’s milk to him only is lacking’). It is unclear why there are three semantically equivalent adjectives in Ukrainian; perhaps they were once not synonymous. For sources of the Ukrainian expression ‘bird’s milk’, see Ivčenko (1998: 114); for further discussion of the expression in Polabian and other Slavic languages, see Ben - veniste (1970: 23); Kostov (1971); Hinze (1974: 510–511); Buyaner (2005). The presence of the term in Finno-Ugric languages may be due either to Russian and/or to direct Radhanite contacts. Aside from Yiddish (also ‘something very delicious’), a sole unique food meaning for ‘bird’s milk’ is found only in Hung madártej, Rum lapte de pasăre ‘dessert’ . The designation of a dessert perhaps developed in Hungarian and Rumanian due to their physical contiguity in Eastern Europe, but Hungarian might have acquired the expression and this meaning prior to the migration from southern Russia to Hungary. Neighbouring Pol ptasie mleczko (with ‘milk’ dim.), in addition to the meanings ‘something rare/delicious/unobtainable’ can also mean ‘chocolate-covered marshmallow’ (since the 1930s at least), which goes together with Hungarian and Rumanian where the meaning is only ‘dessert’. The Persians and Greeks could have become familiar with one another’s culture and language from the exploits of Alexander the Great in Asia and the familiarity of Persians with the tragedies of Sophocles and Euripides (see Götze 1923; Ettinghausen 1972: 7; Miller 1997: 24). However, there are four reasons to oppose a Greek source for the ubiqui - tous expression ‘bird’s milk’: (i) Aristophanes could well have acquired the expression from the Persians. (ii) The term may have reached the Slavs not via Greek or Old Church Slavonic, but via Radhanite Judaic merchants or Irano-Slavo(-Turkic?) White Croats in southeastern Poland or elsewhere . (iii) The concept is particularly popular in Yiddish, which also has synonymous Y — 82 — Slavia Centralis 1/2022 Paul Wexler kačkemilx ‘balm, balsam; incredible foods’, and antonymous ‘tasteless dishes; idle story, nonsense; ineptitude’ (lit. ‘duck’s milk’ with Y kačke < Uk kačka ‘duck’ and milx ‘milk’ < G Milch). (iv) A Greek source becomes even more dubious since we find the term in Afro-Asian languages, like Iranian, Hindi/ Urdu, Neo-Eastern Aramaic, Palestinian/Iraqi/Egyptian Arabic; see e.g. Neo- ChrEAr xulwid sippurta (Urmi dialect, Iranian Kurdestan), Arm terčuni kat and Pe širè morġ—presumably the source of the Hindi/Urdu, Aramaic, Arabic and Caucasian terms ‘bird’s milk’ (the Persian term also meant ‘bat’). It is unclear if the Chinese use of ‘unicorn’ or ‘phoenix feather’ to express the meaning of Y fojglmilx is part of the same Eurasian isogloss, but it should be noted that the unicorn and feather motifs in Chinese parallel the bird motif of Iranian and other languages. In ancient Chinese, the phoenix was known as the king of the birds, with the head of a bird, neck of a snake, throat of a swallow, back of a tortoise and tail of a fish; it had five colours, and was very tall. The Chinese phrases are fèng máo lín jiăo ‘ so m ethin g rare ; rare as p h oenix f ea th ers an d unicon horns’ (lit. ‘phoenix’ + ‘feather’ + ‘female unicorn’ + ‘horn’) and lín jiăo fèng zuì (lit. ‘female unicorn’ [spelt with no less than three characters; qí ‘male unicorn’ is not used in the phrase] + ‘horn’ + ‘phoenix’ [of either sex] + ‘mouth’). G Vogelmilch is known at present only in the meaning ‘Star of Bethle - hem; ornithogalum (herbaceous perennial flowering plant)’. See also VLat lacte gallīnāceum ‘something uncommonly pleasant or sweet’ (lit. ‘chicken’s milk’), which surfaces in the writings of Petronius (1 st c. A.D.). The many influences shared by Chinese and Persian and the length of mercantile contacts of several centuries due to the presence of Iranians (including multilingual Iranian Jews) in several Chinese cities, makes Chinese an attractive licensing source for Yiddish. Note that the Chinese and Yiddish terms for ‘bird’s milk’ both share initial /f/ and /m/, in fojgl + milx and fèng máo. A coincidence? A geography that suggests an Iranian imput which spread both eastwards and westwards clearly spotlights Radhanite merchants as the carriers. No target languages use Iranian morphemes in their common phrase, only native morphemes. Since the expression in Persian was formerly ambiguous, the sec - ond meaning being (now archaic) ‘the mammal bat’, we would have expected Judaists to have relexified both terms; alternatively, the Judaists may have been exposed only to the former meaning, if we can extrapolate from Persian, which also disambiguated širè morġ ‘bird’s milk; bat’, by acquiring xuffaš ‘bat’ (< Arab xuffāš). The link between the ‘bat’ and ‘milk’ deserves further study, since the motif of the bird simurġ, a creature with a mammalian head and the body of a large benevolent female bird, originated in Sassanid Iran, with the bird commonly found in Byzantine silk textiles in the centuries fol - lowing the 6 th century. Hence, it seems reasonable to claim that either the term began in Iranian territory and was subsequently diffused, at least westward, by Radhanite merchants, or else began in Greek territory and then spread to Iran, where it could have been taken up and diffused in various directions by Radhanite merchants. A similar-sounding term, but with a dissimilar meaning — 83 — How Yiddish Polonisms may uncover Iranian, Turkic and other Asian speech … is Arm sira-marg ‘peacock’, with the second term < Oss marġ ‘bird’ (Bailey 1979: 403; Martirosyan 2010: 318); the Armenianism might have strengthened the popularity for Radhanite merchants. An important clue to the origin of Y fojglmilx could come from the existence of morġ in the meaning ‘hen’ in the Median dialect of Ramändi, Iran (Yarshater 2002: 442), but more geographical details about this meaning are needed, since this meaning is not unknown in Persian. This could be significant, since both the territory of the ancient Medes and the White Croat State were called ‘(land of the) mountains’ in a variety of sources. If the phrase ‘hen’s milk’ was typical of Median speech a millennium ago, then it might have been taken by (Judaic?) Median migrants to the White Croat State in southeastern (Little) Poland. Of course, we cannot determine if Median Judaic merchants brought the term to the White Croats, or vice versa, or whether the term was known to both groups independently. Significantly, the only non-Iranian example of the expression ‘hens’ milk’ known to me is G Hühnermilch (li t. ‘ h en ’ s milk ’ , o f unkn o wn age; of Persian origin?); see also related(?) Eng hen’s teeth. Might the German term be the source of the English phrase? This would be a rare case of putative Iranian (or Chinese) influence extending as far west as Germany and England. If the German expression is old, then it too may be of White Croat origin, though the Sorbs themselves may have been another Irano-Slavic confederation. Moreover, the suspected Iranian Germanism uses ‘hens’ in the plural. Gerzon (1902: 87) also cites the variant Y fejglmilx with ‘birds’ pl.; an Arabic dialect also uses the plural. The expression ‘bird’s milk’ is also found in Mediterra - nean Arabic dialects, but not always with the Yiddish-Greek-Slavic meaning, e.g. urban EgypArab laban ‘ōṣfūr, laban ‘aṣfūr (the latter lit. ‘milk of a bird’) has the Greek, Yiddish, Slavic, etc. meaning, in addition to ‘that’s impossible’ (cited by Hary 2009: 97), but urban PalArab laban ‘aṣāfīr means ‘exaggera - tion’ (lit. ‘milk of birds’). It is significant that the Hellenised Balkan Slavic languages in the 7–9th centuries became separated from the Eastern and Western Slavic languages by Hungarian and Rumanian (present-day Rumania was not originally the main location of Balkan Romance speakers: See Wexler 1997)—in which the expression (at least now) only seems to designate ‘vanilla pudding’. The Hungarian-Rumanian-Polish innovation is an intrusion into the larger Slavic zone of ‘bird’s milk’ between the c. 6–9 th centuries, and raises the possibility that Slavic languages north and south of the Hungarian-Rumanian-Polish speech territories could have received the expression from different sources—including from local Iranian confederations. Interestingly, more expressions and meanings are found in Y fojglmilx, kačkemilx than in any of the coterritorial European non-Judaic and Eurasian Judaic languages. Unfortunately, we cannot determine the age of the culinary term in Hungarian, Polish, Rumanian—or whether these languages ever had the original Greek meaning. Moreover, there is some evidence that the use of the expression in the Eastern Slavic languages is (at least now) restricted — 84 — Slavia Centralis 1/2022 Paul Wexler geographically; e.g. some Belarusian speakers (e.g. in the Njasviž area) claim not to have heard the expression at all (Nosovič 1870 cites it for Mahilëw, and informants cite it for Polissja and Homel’). If the absence of the expression in at least some Belarusian dialects proves to be old, it would support the hypothesis of different sources for Southern and Eastern Slavic. 5 Reconstructing Iranian and Turkic speech in Polish and Ukrainian Galicia The following 23 files contain examples of Irano-Turkisms that may have once circulated in the Yiddish, Iranian and Turkic speech of the Western Slavic lands, to judge from Polish (and sometimes also German and Hebrew/Hebroid) replacements in Galician (or, more broadly, Polish and Ukrainian) Yiddish; oc - casionally, the Yiddish forms of Slavic (Polish or Ukrainian) morphemes may themselves manifest Persian or Turkic influence. On rare occasions, a Yiddish Polonism appears to have replaced a Chinese term of similar form and meaning to an Iranianism. It is impossible, given the present state of our knowledge, to determine whether Chinese merchants were present in Irano-Turko-Slavic settlements in the Slavic or German lands, but this is a distinct possibility—in view of the well-known settlements of Arab and Iranian merchants in various Chinese trading centres. In fact, descendants of the Jewish merchants who settled in China at the end of the first millennium still exist there in small numbers, and are familiar with the origins of their Near Eastern ancestors.. In the files below I have also given some examples of Asianisms in Polono- Ukrainian Yiddish that have not been replaced by Polonisms, in order to estab - lish my claim that Yiddish has acquired a wealth of Asianisms and Africanisms (amply documented in my 2021) along the Silk Roads and/or from other Old Jewish Silk Road languages (e.g. possibly from the largely extinct Khazar, or from Karaite; Iranian languages such as Judeo-Soghdian[?] and Džuhuri, from Judeo-Georgian, Judeo-Berber, Judeo-Kwara and other Ethiopic languages). A major question of interest is why Galician Yiddish (in Poland and neigh - bouring Ukraine, in the territory of the White Croat State) has maintained (along with all other dialects of Yiddish in Eastern Europe) such a large corpus of Asianisms, without Polish replacements. I can think of five possible answers: (a) There were no matching Polonisms available (i.e., with similar form and meaning to the retained Asianisms). (b) Galician Yiddish was influenced by the non-Galician Yiddish dialects. (c) Many Yiddish speakers may have ceased to understand Judaised versions of Iranian and Turkic by the end of the first millennium, and, thus, had dif - ficulty in identifying Asian candidates in Yiddish for possible replacement by Polonisms. (d) Presumably, Africanisms and Asianisms in the semantic field of Jewish religion or culture would not have been candidates for replacement by — 85 — How Yiddish Polonisms may uncover Iranian, Turkic and other Asian speech … Polonisms, since they would have already been Hebraised (a confirming example would be Y xale ‘festive Sabbath bread’, most likely of Chinese and/ or Iranian and/or Arabic origin, but assumed to be from Biblical Hebrew). (e) The unreplaced Asianisms in Galician Yiddish did not circulate in the Asian languages of the Slavo-Iranian and Slavo-Turkic confederations located on the Slavic lands, but only in the Judaised dialects of Iranian and Turkic. (Above, I noted that Sorbian slaves in 11 th -century Spain claimed they were descended from Persians—suggesting that they came from still extant Irano- Slavic confederations in their German homeland. A topic awaiting study is the relative chronology of Asian language obsolescence in different parts of Europe and North Africa, among Jews and non-Jews.) The examples follow below: 1. Y pente ‘chain, fetter’, penten ‘to chain’ < Pol pęto, pętać; cognate Uk putaty ‘to clog, hobble, fetter’ is also retained as Y puten ‘to confuse, (en)tangle’. The Yiddish Polonisms may have replaced the formally and semantically similar (but unrelated?) Pe bänd ‘band, fastening, tie, chain’ or pabänd ‘fetters; knot; belt’ (with ‘foot, leg’). Pan-Yiddish has also acquired a second Polonism: Y culončen ‘to connect, join to, bind, tie’ < Pol łączyć, połączyć się ‘to join, unite’ , possibly borrowed because of partial formal and semantical similarity to l-h-ġ of Pe molhäġ kärdän ‘to join’ (< Arab mulḥaq ‘joined; additional’ and Pe ‘to make’). 2. Y farblondžen ‘to go astray’ and farplonte(r)n ‘to confuse (things)’ < Pol zabłądzić ‘to go astray’ (see also item #11 below), (za)plątać ‘to entangle, mud - dle (up)’. Yiddish also utilises cognate Uk bludyty ‘to err, mistake; go astray, lose the way, ramble about’, as Y bludne ‘stray, lost’ (< Uk bludnyj ‘wandering, roaming; burdened with sin’). The first Yiddish Polonism could have replaced an unrelated but similar Pe bärhudän, birah šodän ‘to wander, stray’ (with šodän ‘to become’). Y plonte(r)n ‘to confuse’ probably gained /r/ from Y plonter ‘confusion, muddle’, but lost it optionally if the Polonism was intended to replace Arab falaṭa ‘to overlook, be inattentive to’, falaṭ ‘unexpected event’, filāṭ ‘surprise’ which lack /r/ (> Pe filat ‘anything sudden and unexpected’). 3. Yiddish posssibly preferred Pol łąka over cognate Uk luka ‘meadow, field deluged by spring waters’ because of its similarity with Y lan ‘glade, meadow, field’ (< Uk lan ‘large piece of tilled land, grainfields’, itself < Pol łan ‘unit of cultivated land’, and assumed to be from MHG lê[he]n, lêhn ‘loaned prop - erty; measure of field; land given to a vassal by lords for military service or as payment’). Alternatively, Y lan and lonke ‘lawn, meadow’ could have been acquired in the White Croat State from Oss (I) lænk, (D) lœncœ ‘hollow, cav - ity, depression; low field; valley; ravine’ and/or Tu alan ‘field, area; sphere, domain, public square’, respectively. — 86 — Slavia Centralis 1/2022 Paul Wexler Y senezenc, sjenoženc, sanozenc, šanozenc ‘meadow’ have the form of Pol sianożęcie ‘haymaking’ (lit. ‘hay’ + ‘cutting’) but the meaning of cognate Uk sinožat’, Br senažac’ ‘meadow’; were the Yiddishisms licensed by the formal similarity with unrelated Pe čämän ‘meadow’? 4. Y štam has the meanings ‘trunk, stem’ of (MH)G Stamm, also ‘race, clan, pedigree; tree’, but the additional meanings of ‘race, clan, pedigree’ are mainly available to Yiddish only as later post-relexificational Germanisms. The mean - ing ‘tree’ is not available to Y štam . For G ‘pedigree’, see also Y jixes ‘descent, pedigree, lineage, parentage, Judaic aristocracy’ (< He jḥws / j ixus/) an d Y šalšeles-hajoxsn ‘chain of geneology, aristocratic pedigree, dynasty’ (< He šlšlt hjḥsjn /šalšelet hajuxasin/, lit. ‘the chain of lineages’). The German practice of expressing ‘tree’ and ‘family tree, descent’ by a common term finds a parallel in Pe šädžär ‘tree’/ šädžärè ‘pedigree’, šädžärät (onnäsäb) ‘family tree’ (lit. ‘trees of’ + the family’ < Arab šadžara ‘tree’, šadžar ‘trees, shrubs’, an- ‘the’, nasab ‘lineage’). If Persian licensed the two sets of meanings of the Germanism in Yiddish, then the semantics of the German loan in USo štom, štóm ‘tree, trunk; flock’, may also reflect Iranian influence—which could support the hypothesis that Iranian was spoken in the Irano-Sorbian community. Interestingly, the optional use of ‘letter’ in Y jixes- briv ‘pedigree’ (< G Brief ‘letter’) parallels Pe näsäbnamè ‘geneology, pedigree’ with namè ‘book; letter’. An open question is why the meaning ‘tree’ is not available to Y štam (see instead bojm < G Baum). In Steingass’s compendious dictionary of contemporary and archaic Persian (1892), there is a very large number of terms for both generic and specific trees that begin with /b/, which could have prompted Yiddish speakers to accept G Baum. Similarly, Steingass lists a very large list of Persian terms for ‘tree’ that begin with /d/, as in Slavic, but Yiddish has no such forms. Yiddish also has cryptic ejc ‘tree’ (< He ‘c /ec/) in the meanings ‘tree of life, knowledge’; we may assume Yiddish once had the Hebraism with the simple meaning ‘tree’, since the latter was used in German Rotwelsch (professional slang) variants, from the early 19 th century . Yiddish shows no trace of G Stammbaum ‘family tree’. 5. Y jar ‘dale; gulch, ravine’ may be a direct loan < Uk jar ‘ravine, ditch; cliff; steep bank of a river’, Pol jar ‘ cany on, ravine ’ or < the ultimate source lan - guage, Tc * jār. The geography of the term is significant: Carpathian Ukrainian and Slovak have the unique form jarok ‘ditch, channel, groove; creek’ < Hung árok ‘irrigation ditch’ < Tc jār(ok) . Yiddish, Hungarian, Slovak and Carpathian Ukrainian may all have acquired Tc jār(ok) directly from Central/Eastern Eu - ropean White Croat or some other Slavo-Asian confederation. 6. Y jatke ‘butchery’ < Uk jatka ‘butcher’s stall, booth at a market’ or from the Polish etymon, jatka, directly (though a synonymous dim. - kè is also native to Persian dialects; see the next entry). Alternatively, Yiddish may have replaced — 87 — How Yiddish Polonisms may uncover Iranian, Turkic and other Asian speech … an earlier Sogh jātē ‘meat; flesh’ or Jagh jōta ‘meat’ (not found in Persian) by the similar (cognate?) Ukrainianism or Polonism. The possibility that the Slavism is itself of Iranian origin needs study. 7. Yiddish uses two Ukrainian variants of the native root for ‘flower’: kvejt and cvit ‘flower’. In Eastern Slavic, kv can alternate with cv before CSl * ě in a small number of roots: see Uk kvit ‘flower’ and cvit ‘flower; blossom’. Kv appears mostly in northern and eastern Ukrainian; cv prevails in southwestern Ukrainian. Shevelov (1979, 56–57) suggests that UkWLemk kvitka probably < Slovak, but it appears to be common in the area formerly occupied by the White Croats (dim. - kè is also dialectal Persian, including Median). See also Pol kwiat ‘flower’. Note also the partial similarity of both Yiddishisms with Pe oškufè, šekufè, šokufè ‘flower(ing), blossoming’, išguft ‘ expanding (blossoming) of a flower’, šekuftän, šekoftän, šekofidän, šegoftän ‘to open (flower)’. Pe -kä is also used in Ch u v as h an d P ers i an R o mani, see e . g . Ch u v dim. kača(ka) ‘goat’ (Benzing 1959: 716). On the Persian Romani use of Pe dim. -kä and -čä, see Ivanow (1922: 378) and Matras (2002: 24). On the Tc dim. -V k, see Clauson (1972: 183); see also Chuv - ke, and its possible appearance in Yiddish. The diminutive infix with -k- is also available in Armenian and Sanskrit. On Arm -ik dim. in Romani, see Bläsing (2002: 107). Persian dialects also have diminutive suffixes which are identical to Y -ele, -l(e): e.g. the dialect of Hamadan (also in Media!) uses - lä, as in ähmädlä ‘little Ahmad’ (Fück 1955: 171 and fns. 51–52). Thus, while Y - l, -(e)le dims. appear to < G - l, see Y mejdl, (S)G Mädel ‘girl’ (vs. N, stG Mädchen ‘girl’), they might also be < Iranian and/or Turkic, which (later) predisposed Yiddish to follow South German norms. The dialect of Rajj (the political and cultural base of the Medes!) uses -kä, see e.g. häkä dim. ‘little Hasan’ vs. stPe häsänäk (< Arab ḥasan); see also (Rajj) hämäkä dim. ‘little Ahmad’ (< Arab aḥmad). See also KhotSaka dim. - ka, as in pūrä ‘son’ > pūrakä dim. If Yiddish acquired the diminutive suffix from Median Persian, we might assume that some of the Jewish Radhanite merchants joined fellow Median communities in establishing Irano-Slavic confederations in Galicia. 8. Uk hubka ‘sponge; tree fungus’ > Y hubke, gubke dim. ‘fungus on a tree’, while cognate Pol gąbka is the immediate source of Y gom(b)ke ‘sponge’. Dena - salisation appears to have taken place in Eastern Slavic in approximately the 9th century, but before that, the Eastern Slavism > Tat gömba ‘mushroom’, perhaps via Volga proto-Bulgar, which may be the source of Chuv kămpa ‘mushroom’. Thus, Y gom(b)ke ‘sponge’, despite its Polish form, may actually be a loan from Chuvash, which, subsequently, attached a Perso-Turko-Slavic diminutive suffix; in the latter case, the etymon may have also been used in the Iranian and/or Turkic speech of the White Croat State. Chuvash is the only surviving language in the small Turkic subgroup of which the extinct Khazar is also a member, and was also spoken in the Khazar Empire. Some Chuvash speakers — 88 — Slavia Centralis 1/2022 Paul Wexler may have joined the Khazars in converting to Judaism in the 9 th century for the purposes of working with Jewish traders on the Silk Roads. Tat gömba and Chuv kămpa ‘ mushroom ’ coul d both account f or the mid front vowel of Y gembe ‘mouth; chin’ (alongside gombe ‘chin’), if not Pol gęba , gąbka ‘(animal’s) mouth; sponge’. Note also the use of two German variants in Yiddish: Y švom, švojm ‘mushroom; sponge; fungus’ (the latter suggest two distinct German lexifier dialects and/or two chronologies of acquisition; see MHG swam, swamp, swamme). The Polish, Yiddish and German doublets for ‘mushroom’ may all be imitating the variety of forms in Turkic languages. 9. Uk xlop ‘peasant, rustic; man, male’, xlopec’ ‘ bo y , lad; j ac k in cards ; ap - prentice; bachelor’ < Pol chłop ‘peasant; boor’. Of this set, Yiddish has xlop ‘peasant; rustic’, but shows no trace of Uk xolop ‘serf’. The etymology of these Slavisms is uncertain, but possible etyma are Pe xäläf ‘deputy, vice-regent, suc - cessor, heir; prodigy, generation; son; deserving’, surface congeners, Tc xalfa ‘apprentice, servant’, Tu halef ‘surrogate, replacement’ ( all ultimately < Arab xalaf ‘substitute; good son’, a cognate of xalafa ‘to supplant, follow; be stupid [young man]’—the latter [Iranian?] meaning is found in Polish but not in Yid - dish). The Persian meanings all have non-German expression in Yiddish, see e.g. G’oid noxkumer ‘successor’ (vs. G Nachkömmling), alongside Y jaršenen ‘to inherit’, memale-mokem ‘successor’, ile ‘wonder child, prodigy’, dor ‘gen - eration’, bonim ‘sons’, ben ‘the son of’, zojxe zajn ‘to be worthy of’ (< He jrš /jaraš/ ‘he inherited’, mml’-mqwm /memale-makom/, lit. ‘he fills’ + ‘place’, ‘lwj /ilui/, dwr /dor/, bnjm /banim/, bn /ben/, zwxh /zoxe/ ‘he is worthy of’). Y xiluke-dejes (lit. ‘divisions, parts of views’) and maxlojkes ‘quar rel’, are two variants of the common Hebrew root x-l-q, close formally to Uk-Pol x-l-p. He x-l-q is semantically identical to Arab xalaf, cited above, which also expresses the meaning ‘quarrel’: Arab xilāf ‘discord’, ixtilāf ‘difference of opinion’ > Pe xilaf ‘contradicting, rebelling against’, èxtilaf ‘disagreement, discord’. The two Yiddish variants of He x-l-q ma y addi ti o nall y ha v e been licensed by the doublets Arab nizā’, munāza’a and xiṣām, ixtiṣām, xuṣūma ‘dispute’ (> Pe niza, monazèät, èxtisam, xusumät). He /xelek/ ‘part’, a cognate of Y xiluke-dejes and maxlojkes ‘quarrel’, is also used in Yiddish as xejlek ‘section, part, share, division; portion’. The pair of Y xejlek and xiluke-dejes may have been licensed by the near-minimal pairs of Arab qism ‘part, share, allotment; division, section’ and qisṭ ‘part, share, allotment; measure, extent’ (> Pe ġism ‘part, portion; division’, ġist ‘division, portion, share’). Iranian could have licensed the use of the Hebraisms in Yiddish in the White Croat State or other Irano-Slavic confederation in Europe—but we cannot exclude acquisition in the Persian homeland either. 10. Y xitre ‘crafty’, xitrak ‘sly person’, xitrekejt, xitrost ‘cunning; craft’, xitre - ven ‘to use cunning’ < Pol chytrek, Uk xytrun ‘sly person’, xytryj, xytrist’, xytruvaty. The semantic gaps between Uk xytryj ‘complicated; clever’ and USo — 89 — How Yiddish Polonisms may uncover Iranian, Turkic and other Asian speech … chětry ‘agile, quick, nimble, skillful; quite, considerable’ could have licensed the retention of a Polish Slavism in Yiddish, along with Y pike(je)x ‘clever man’, bistre ‘agile’, luftik ‘nimble’ (the latter two have original meanings), bekies ‘skill’, boke ‘skilled’, berje m, f ‘ s kill ed perso n ’ ( th ere is n o g en der in Persian except in some Arabisms), niškošedik ‘considerable’ < He pjqḥ /pikeax/, Uk bystryj, G luftig ‘breezy’, He bqjwt /bkiut/, bqj /baki/, brjh /berja/, Pol niezgorszy ‘fair, tolerable’ and/or Uk nezhiršyj ‘not bad, good enough, tolerable, passable’. Y xitre, etc. may originally have been replacements for the formerly similar Chuv xitre ‘pretty; good, luxurious, magnificent, splendid; pompous; costly, rich’. I doubt that Y niškošedik ‘considerable’ ~ Pol niezgorszy ‘fair, tolerable’ and Uk nezhiršyj ‘not bad’. The second syllable - koše- perhaps < He qšh /kaše/ ‘hard; difficult’—a cognate of synonymous Arab qāsin (> Pe ġäsi ‘hard [in feeling]’). Y boke ‘skilled (person)’ may have been licensed by a formally and semanti - cally similar Pe-Tc bakši ‘learned doctor; Buddhist priest’, which is the source of Chin bó-shì ‘learned doctor; expert, erudite’ (written with the characters for native bó ‘to be well-informed’ + shì ‘scholar’). See also Chin bó-qià ‘well- informed’ (with ‘extensive’) and bó-yă ‘erudite’ (with ‘refined’), borrowed as Jap bokushi ‘pastor, cleric, chaplain’. I note that the original meaning of the source Mong bakši, baxši was ‘Mongol shaman ’ , which later came to denote ‘religious Buddhists, Buddhist doctors’, possibly during the Mongol period in Iran (13 th c.). The Hebroidism, Y boke , might actually have been licensed by an original Mongolianism. The latter might also have been an epithet of Rašīd ad- Dīn (b.c.1247–d.1318), the Jewish vizier of the Ilkhan ruler of Iran, and author of the most important source for the history of the Mongol Ilkhanate period and the Mongol empire; if so, this could well have made the term popular in so many languages, including Yiddish. But are Y xitre, etc., really Slavic? After all, they also share form and mean - ing with Arab xaṭara ‘to remember something’, xaṭura ‘to be eminent, high in rank, important’, axṭara ‘to be in danger (due to illness); be the equal or rival of; warn; lay a bet, risk’, xaṭar ‘danger’, xaṭir ‘dangerous’, xāṭir ‘occurrence of a thought; pleasure, fancy, sake’, which surface as Pe xätär ‘danger, threat; risk; alarm’, xatèr ‘thought, idea; remembrance’, etc. The negative connota - tions of Pe xätär may have prompted the Yiddish acquisition of Polonisms like Y mondrak, etc. ‘wiseguy’ < Pol mądry ‘wise’ in the White Croat State. The latter now surfaces in Yiddish beyond the borders of Poland. The cognate Pol mędrek ‘wiseguy’ may be the basis of Y šmendrik ‘ smart -al ec k ’ , wi th a pejorative š- (< Turkic), possibly reinforced by Pol męda ‘creep’. Pol mądry has Iranian cognates, see Aves mązdra- ‘wise; sensible’, mąz-dā-, maz-da ‘to keep in one’s memory’, and a formally and semantically close (but not cognate) Pe modärräb ‘experienced’ (< Arab mudarrab). Indeed, the Slavism has Iranian cognates, see e.g. Aves mązdra- ‘wise; sensible’, mąz-dā-, maz-da ‘to keep in one’s memory’, which support the hypothesis that the Yiddish Polonism replaced an earlier Iranian cognate, and maybe even two, in the White Croat State. — 90 — Slavia Centralis 1/2022 Paul Wexler 11. The etymology of Y grajz ‘error, mistake’, etc. has puzzled scholars for over a century. Suggested etyma are He grj’h /gri’a/ ‘subtraction, diminution’ and Uk hrix ‘sin; error’, which are plausible, given Uk pohrišyty ‘ to s in ( a little); err’, ohrix ‘ error’ . Baltic also o ffers a plausible semantic shift in Latv grèizs ‘crooked, curved’ > ‘bad, unjust, false’ (the final - zs might explain Y -z). Alternatively, the meaning ‘sin’ frequently develops the meaning ‘false’ via an intermediary stage of ‘error’, see Chuv sue, suja ‘lie; untruthful, pretended’, suj ‘to lie’, maybe < Chin tsùi ‘sin; vice, defect, flaw; guilt, blame’. Besides Uk hrix, see also Pe xäta ‘error; transgression’, xäta kärdän ‘to err; sin’ (< Arab xaṭā’ [and Pe ‘to do’] with a Hebrew cognate in Yiddish). However, the above analyses are phonologically not entirely convincing, and Baltic association with Yiddish is relatively late and minor (though an Estonian lingua franca was in use in the Judaised Khazar empire). In rejecting the Hebrew and Slavic sources cited above, I would derive Y grajz ‘error’ < grajzl ‘curl’ and krajzl ‘ripple, ruffle’, etc., created from the solitary G kraus ‘curly, frizzy’; (newer) ‘muddled, confused’, Kraus, Krause ‘ruff, frill’, krausen ‘to pleat; knit (brow)’, kräuseln ‘ to cur l (hair ) , ruffl e ; p l ea t ’ ( 1 6 th c . ) . Whil e the root is already attested in Middle High German, the meanings ‘muddled, confused’ are too recent to have influenced Yiddish, and MLG krūs ‘curled; dishevelled; convoluted’ comes from a dialect to which Yiddish speakers were never exposed directly. There does not appear to be a German match for the Yiddish variant with gr-, and certainly no precedent in German for the specific meaning ‘to err’. However, there are doublets in Slavic and Iranian that could motivate the Yiddish minimal pair. A Ukrainian model for the coexistence of ‘err(or)’ and ‘(to) curl’ exists in Uk (i) mylaty ‘to wind, reel; shake; twist, twirl, roll up; embroil, entangle’ and cognate (ii) mylytysja ‘to err, be mistaken’, or in (iii) krutyty ‘to twist; turn; twirl, wind’/ perekrutyty ‘to twist; turn, twirl, whirl; overwind; corrupt, distort, misinterpret, misconstrue; garble, con - fuse’. Still, the Y kr- and gr- variants may reflect different chronologies of diverse German source dialects. Yiddish may also have created doublets to distinguish between ‘curl of smoke’ and ‘curl of hair’, see Y krajzlen (zix) ‘to curl, ruffle’, krajzl ‘ripple, ruffle’, grajzl ‘curl’, grajzlen (zix) ‘ to cur l (hair)’, respectively. Persian also provides not only a model that parallels the two Ukrainian roots cited above, but even a precise mirror image of the two Yiddish “Germanisms”, grajz(l) and krajzl. Consider the minimal pair of Pe ġält ‘rolling’ < ġältidän ‘to roll (from place to place); trill (in music), warbling’, spelled either with the Arabic letter /t/ ت or pharyngealised /ṭ/ ط vs. ġälät ( spelt only with pharyn - gealised /ṭ/ = Pe [t]) ‘error, mistake; going astray; guilt; erroneous, incorrect’. The first root appears to be native Persian, but the second term < Arab ġalaṭ ‘error’; /ṭ/ appears only in the spelling of Persian Arabisms, and always has the phonetic value of Pe /t/. Yiddish speakers fluent in Persian might have copied the spelling and pronunciation of the Persian pair onto a single German root. — 91 — How Yiddish Polonisms may uncover Iranian, Turkic and other Asian speech … See also the minimal pair of Jazgh ġalat ‘error’/ ġǝlǝt ‘to sink (in sand, mud, snow)’—a kind of “error”. Other links with Iranian come to mind, such as consonant metathesis, typical of Eastern Iranian languages. For example, if Pe läġzèš ‘ s li p ; erro r; going astray’, läġzidän ‘to err, take a false step; slip, slide, glide’ had become via metathesis * ġälzèš or *ġärzèš, then we could anticipate a parallel shift in Y krajz- > grajz: See also Džuh gərəz ‘evil thought, secret intention, hidden spite/malice/anger/hostility’. Pe läġzèš might also have been the motivation for using Y b(e)laz ‘in a foreign (usually Romance) language (used as vernacular equivalents of written Hebrew or Aramaic words); otherwise known as’ (maybe < BiblHe b- /be-/ ‘in’ + l’z /la’az/ ‘Egyptian’). The possibility that Y grajz, etc. was coined in Galicia is supported by the retention of the semantically similar Polonism in non-Polish Yiddish: blondžen ‘to stray, ramble (having lost one’s way)’, cited in example #2 above. Finally, the link between ‘error’ + ‘to curl, twist’ can be motivated by the possibly cognate Chin miù ‘lie, error’ and jiŭ , jiū, jiăo ‘to twist’. The latter pair in Japanese is pronounced byū ‘mistake’ and kyū ‘to twist thread (rope); put all in one place’. Thus, Yiddish ‘error’/’to twist’ would be the western-most member of an isogloss extending from Yiddish in the Slavic and German lands (but not in Slavic and German!) to Japanese. This is a good example of the huge benefit for Slavic and German linguistics that follows from the innova - tive comparative Yiddish research within a Silk Road framework that I am proposing in the present study. 12. For ‘dull’, Yiddish retains a Polish root, see e.g. Y temp(ik) ‘dull, blunt; obtuse, dense’ < Pol tępy ‘dull’ (unrelated to G stumpf ‘ dull, blunt’). There is no trace of the root vowel of the cognate USo tupy, Uk tupyj ‘dull’. I as - sume that the pan-Yiddish use of Pol temp(ik) may have begun in the White Croat State during the obsolescence of (Judeo-)Iranian there. An Iranian homophonous pair, such as Džuh kut¹ ‘lump, clog, ingot, bar (of metal)’/ kut² ‘dull, blunt; dullard, blockhead, dolt, dunce, slow-witted person’ could have licensed the use of the relexified Pol tępy as Y temp ‘dull, blunt; obtuse, dense’ (relexification is mandatory when the substratal language has mini - mal or near-minimal pairs, because they can almost never be imitated in the superstratal lexifier language). 13. Yiddish expresses ‘cabbage/lettuce stalk’ by glomp/b (also ‘dummy’) (see Pol głąb ‘depth; stump/stalk’) in Southeastern Poland and at locales in Western and Carpathian Ukraine. In addition, there is also pan-Y kačn, kačen¹, kačan ‘cob, cabbage stump’ (< Pol kaczan < Tu koçan). The Yiddish of the former White Croat State appears to have also licensed the Polonism, because it resembled Pe käläm ‘cabbage’. I would also speculate that Pe ġäläm ‘pen; reed’ (< Arab qalam) also licensed the acquisition of Pol głąb because of its similarity in form and meaning to G Halm ‘blade; stalk; straw’. — 92 — Slavia Centralis 1/2022 Paul Wexler The meaning of Y glomp/b ‘ dumm y ’ was poss ib l y also supported b y the formally similar Uk hlupyj ‘stupid’, hlupak ‘stupid fellow’, nehlup ‘very stupid fellow’ (perhaps before the lenition of the Uk * g > h), Pol głupawy ‘foolish’, głupi ‘stupid’, with an attempt to “Polonise” the Yiddishism. Other possibilities are that Y glomp/b (i) Became the sarcastic opposite of the similar Pol głęboki ‘deep’ in the sense of ‘intellectual, profound’ and/or (ii) May have crossed with Y gojlem ‘dummy; golem, an artificial man’ (thought to be < He gwlm /golem/ ‘larva, pupa, embryo; shapeless matter; awkward person, robot, ignoramus; leg - endary automaton made of clay’). However, the source of Y gojlem is possibly the Pol golemy ‘large’ or Cz holomek ‘unmarried youth, servant’—or, better, the Pe ġulam ‘boy; adolescent; servant, slave’ < Arabic. Curiously, Russian Romani reveals a Polonised term golumbo ‘pigeon’, which is either < Pol gołąb [gówomp] or is a blend of the Polonism with the cognate R golub’. Hence, there is the possibility that Romani was also Polonised in the White Croat State just like Yiddish (though many Roma resided in Iran for several centuries upon leaving India), and also replaced an earlier Iranian term, see e.g. similar Pe kaludž ‘pigeon’ (the Roma probably worked with Jew - ish merchants on the caravans; the Rom gadžo ‘non-Rom’ may be < He/Y goj ‘non-Jew’, originally ‘nation’; j > dž in Iranian, Romance, etc.). Consider also the RRom dlèngo ‘long’ with a pseudo-Polish shape (vs. Pol długi ~ Uk dovhyj ), pointing to the White Croat State; see Pe ludžija ‘long’. Finally, Y kačen² ‘to roll’ (< Uk kačaty ‘to roll, spread by rolling’) may have been retained in Yiddish, unrelexified to the G sich wälzen ‘to roll ’ , because o f the near -homonymous kačen¹ ‘cob, cabbage stump’ (which lacked a parallel pair in German), and/or by Indo-Iranianisms, such as Rom katel ‘to spin’, Hi kāt ‘to spin’ . On the matter of ‘to roll’, there is another interesting Yiddish link with Chinese, Polish and the White Croat State. The Y xale ‘braided festive bread’ is regarded by Jews and non-Jews alike as a quintessential Jewish holiday and Sabbath bread (see also item #23 below). All linguists regard the etymon as He ḥllh /xala/ , but this is unconvincing, since the latter denotes something ined - ible! It means solely the ‘tithe of unbaked dough presented to the priests in the Temple in Jerusalem’ (see BalkDžud xala ‘unbaked bread given as a tithe’). There are a host of semantically and formally similar terms in Arabic ( ġallah ‘grain; bread’ > Pe gälle ‘grains’), but the most attractive one is the Pe āl ‘a dangerous goddess who had the potential to cause various types of damage and destruction to the harvest’. In European contexts, the goddess could offer protec - tion to girls reaching the age of puberty; she then had to be given the locks of hair of the girls as an offering. The offering was symbolised by the baking of a plaited bread which was consumed only on holidays or special occasions, such as weddings, funerals, etc. The knowledge of the bread term and of the goddess and her functions with girls who have reached puberty has reached German, where the goddess is known as Frau Holle ‘Madame Holle’. It could have been introduced into the German lands either by Radhanite Jewish or non-Jewish merchants. There is a plausible Chinese link in the formally and semantically — 93 — How Yiddish Polonisms may uncover Iranian, Turkic and other Asian speech … similar Chin huā juăn ‘bread roll’, i.e., ‘a steamed twisted bread roll!’ (with huā [xwa] ‘flower, blossom; florid, fancy pattern’ + juăn [džuen] ‘to roll [up]; roll; scroll; volume’; classifier for small rolled objects); in Běijíng Mandarin, the term would be huā juăr, with Chin /x-/ and dim. /-r/ licensing Y /x-/ and /-l-/ (/r/ ~ /l/ is a widespread correspondence). Persians of all religious persuasions resided and operated in China for several centuries (evidently to a far greater degree than was the case in the Holy Roman Empire in Germany), and a braided bread is also known in Iran. The Chinese expression could have inspired Jews to seek a Hebraism of similar form and meaning for Jewish languages (it is exceptionally rare in other Jewish languages), or else the Hebroidism was the source of the Chinese expression—first coined by Chinese Jews. Also, it is entirely possible that the Chinese bread term entered Yiddish in the White Croat State for the first time via the Pol chałka. What makes the latter an attractive source of Y xale is the absence of * chała without the diminutive suffix/infix. The Beijing -r also has a diminutive function! There is no diminutive suffix possible with the Yiddish Hebroidism to the best of my knowledge. 14. Y drong, prent ‘bar’ < Pol drąg, pręt ‘rod, bar, stick’. Yiddish lacks syn - onymous Uk prut. The Y drong could very likely be connected to the White Croat state, since Persian has the formally and semantically similar diräk ‘pole, mast’. Y drong ‘pole; tall man’ allows two plurals: drongen and drenger, but I cannot determine their geography. Persian has pl. - an ~ Y pl. -en, but lacks a parallel of Y - er p l . A poss i b l e suppo rt f o r th e P o l o nism co ul d also co m e from the Chin zhàng¹ ‘stick, staff (for a specific purpose)’ < OChin * draŋ[?], ONWChin * daŋ (c. 400 A.D.). Y prent resembles Pe bärn ‘a bundle of rods’ . 15. Tarnów PolY krātiš ‘strange, incomprehensible’ might come from G Krau- twelsch (lit. Kraut ‘weed’ + Welsch ‘incomprehensible Romance speech’), an expression still used in the 1920s in the Tyrol, and synonymous with G Kauderwelsch, Kinderwelsch ‘incomprehensible language’ (< kaudern ‘to speak unclearly’ [of uncertain origin] and Kinder- ‘children’). But I could also imagine the reverse derivation, i.e. G Krautwelsch < PolY krātiš, theoretically ‘Croat language’ (now the name of a not entirely comprehensible Slavic lan - guage, or formerly the name of an incomprehensible Iranian language)—since Tarnów is located just to the west of the former White Croat State. Slavic- and Yiddish-speaking Jews in the Irano-Slavic area could have turned “Croat” into ‘strange, incomprehensible’ once Iranian speech had become obsolescent, just as G Welsch ‘Welsh’ > ‘incomprehensible Romance speech ’ . ( See the discus - sion of the original Iranian term for ‘Croat’ found only in the Arabic writings of Ibrāhīm ibn Ja’qūb cited before the data). The Pol Y krātiš itself may have replaced the original Iranian form of the ethno-glottonym. 16. Y pečatek, pečatik, počatek, počatik ‘beginning (of business), first sale of the day’ < Uk počatok ‘beginning’, but Y počontik < Pol początek ‘beginning, — 94 — Slavia Centralis 1/2022 Paul Wexler opening; origin’. The Polonism may be a local replacement of the Pe däšt¹ ‘initiative (in trade); a monetary gift at a holiday’ in the former White Croat State, which is homophonous with Pe däšt² ‘desert; plain, field’ (expressed by many non-Germanisms in Yiddish). In relexification, it is usually impossible to find (near-)minimal pairs of Germanisms or Slavisms for the Persian minimal pair, hence lexical items are blocked from the superstratal German or Slavic “lexifier language”. Another source for the Yiddishisms could be Pe èstèftah ‘opening; beginning; capture; asking assistance’ < Arab istiftāḥ ‘opening of a market; first sale of the day’ < the root f-t-ḥ ‘to open’, partially similar to Sl p-č-t-k (OHe pharyngeal ḥ merged with velar x in Yiddish and Ashkenazic Hebrew/Hebroid, perhaps on the model of Perso-Arab ḥ, which merged with h). If Pe èstèftah never meant ‘market opening’ (it means ‘opening; beginning’ in general), then the Jews could have been exposed directly to Arabic. 17. Y cere ‘complexion’ either < Uk cera (< Pol cera with sibilant confusion < Ital cera [č]), or Pe čehrè ‘face; cheek; appearance’; girl’s name. A synonymous Y cure ‘face’ (affective and pejorative in Northeastern Yiddish), appears to be derived < He cwrh /cura/ ‘form; picture; countenance; majestic appearance; expression; physiognomy’—but never *‘face’, to judge from the etymological Hebrew spelling. The WY cure (spoken in the German-speaking lands, Holland, Alsace, Switzerland, Hungary, western Czech and Slovak lands), and Rtw cure can only mean ‘form’. The meaning He cwrh /cura/ ‘form’ is found in Yiddish only in a few prepositional phrases, such as becures-odem ‘ in the form o f a man’ (< He bcwrt ‘dm /becurat adam/). Perhaps Western Yiddish dialects in Germany had less or no contact with Iranian settlements. I believe the Y cure ‘face’ reflects a unique meaning of the cognate Ara - bism in Iranian languages. In contrast, the double meanings of ‘face; form’ are divided between two allomorphs in Uk tvar ‘creature, being; face, visage’/ tvoryty ‘to compose, form; create, produce’ and Pol twarz ‘face’/ tworzyć ‘to form’. Still Pol twarz and Uk tvar in th e m eanin g ‘ f ace ’ ma y also re fl ect a Persian influence. Y cure ‘face’ < Pe surät ‘face; portrait; form; list’, sär-o-surät ‘face’ (lit. ‘head; beginning; end’ + ‘and’ + ‘face’) < Arab ṣūra ‘form, picture’ (*‘face’). Jews speaking the Borudžerd dialect of Judeo-Persian use surät for ‘face’, but in their secretive trade lexicon (Lotera’i), they prefer punim (< He pnjm /panim/— the source of the Yiddish pronunciation, [ponem]), which would be unknown to non-Jewish Persians. Like Persian, Turkish has both the meanings of the original and the Irani - anised Arabism, but they are formally distinct in Tu suret ‘form, shape, picture; copy’ (< Persian < Arabic) vs. surat ‘face; mien’ (< Persian); thus, the origin of Y ‘face’ could also be ascribed to Turkic. There is also a Chinese link in the Y parcef-ponem ‘mug’ (< He prcwf-pnjm /parcuf-panim/, lit. ‘mug’ + ‘face’). Most of the Chinese terms for ‘face’ are bimorphemic, and some involve morphemes that begin with one or two bilabial consonants, see e.g. Chin bíao-miàn (lit. — 95 — How Yiddish Polonisms may uncover Iranian, Turkic and other Asian speech … ‘surface’ + ‘appearance’), miàn mào (with ‘face’), miàn duì (with ‘opposite’; see also duì miàn ‘opposite; to be face to face’), miàn bù ‘face’ (with ‘part’). At present, Džuh sirot, surot, surèt and sürèt have only the meanings ‘pic - ture, portrait, drawing, representation, portrayal, image; form, appearance’, but not ‘face’. This is valuable evidence that Džuhuri may not ever have been spoken in the White Croat State. However, the progression of ‘form’ > ‘face’ took place in another root, to judge from Džuh sir-sifèt ‘face, physiognomy’ (< sir/sǝr ‘face’, nor mally ‘head’, and sifet ‘description, quality, attribute; form; like, resembling’ < Arab ṣifa). The meanings ‘face’ and ‘form’ can coexist in many languages, not neces - sarily in contact with Yiddish, but the original underlying term is not ‘form’, as in Turkish, Iranian and Yiddish, see e.g. Kurd rang ‘face; appearance; form’ < ‘colour, paint; shine’. The semantic shift of Eng mug from ‘cup’ > ‘face’ supposedly took place because of the custom of decorating cups with a human face. Might the coexistence of the two meanings in English be, alternatively, a reflection of Iranian influence? 18. There is a possible Iranian motivation (in the White Croat state?) for a Slavic near-minimal pair, e.g. Pe därd ‘pain; sickness, illness, suffering’/ ard ‘flour’ ~ Pol mąka ‘f lour’/ męka ‘torment, pain’, Uk, Br múka ‘pain, torment, torture’/ muká ‘flour’ (under Iranian influence?). CePolY menčen ‘to torment’ (< Pol męczyć) may have replaced the synonymous Pe därd, but pan-Yiddish (at least now) lacks Pol mąka ‘flour’ and męka ‘torment, pain’. Nor does East - ern Yiddish preserve the Uk múka ‘pain, torment, torture’ and muká ‘f lour’. Curiously, Polish Yiddish—except for the centre of the country—does not use the Polish root for ‘to torment’ with a historically nasalised vowel, but rather uses the Eastern Slavic/Upper Sorbian cognate with /u/: mučen. Hence, it is possible that Polish or Upper Sorbian Yiddish could have originally copied the near-minimal pair of Pe därd ‘pain; sickness; suffering’/ ard ‘flour’. If the Sorbs were, like the White Croats, partly of Iranian origin, then the local Yiddish dialects might also have retained Iranianisms there for a longer period of time than in Eastern Slavic Yiddish not spoken on an Iranian substratum. But the end result in all dialects of Yiddish would be the same—initial blockage of the German equivalents. A Sorbian source for the Y mučen ‘to torment’, found in all dialects of Yiddish except Central Polish Yiddish, is more attractive than the Uk mučyty, because the Y mučen is also found in Poland to the west of CePolY menčen, in an area which once had a Sorbian population . In Sorbian Yiddish, we would expect the same reflex of the Common Slavic nasal vowels as in Eastern Slavic Yiddish. Perhaps future research can help to determine whether Iranian speech prevailed longer in the White Croat State in Polish Galica or in the Sorb lands. In any event, the existence of pan-Yiddish Polonisms, and possibly also of pan- Yiddish Sorbianisms, could corroborate the presence of Irano-(Turco-?)Sla vic — 96 — Slavia Centralis 1/2022 Paul Wexler confederations, such as the White Croats and “White Serbs”. On poss i b l e Sorbianisms in Yiddish, see my 1991. A Polish term is also found in Northwestern Belarusian, among Catholic speakers, see, e.g., zmenčycca ‘to get tired’ (< Pol zmęczyć się ). Thus, theoreti - cally, the Polonism may have reached Yiddish not directly from Polish but from Polonised Belarusian—in either case, to replace a prior Iranianism. A formally similar pair of morphemes for ‘flour’ and ‘exhaustion’ (i.e. ‘pain’) is also found in Chin fén ‘f lour’/ fá ‘to exhaust’, which shows the po - tential extent of this isogloss. 19. It is instructive to examine one example which shows how Polish and Bela - rusian have retained an original meaning of an Iranian or Judeo-Aramaic term, even though the coterritorial Yiddish dialects have become formally Slavicised. The Y paskudnik and paskudnjak m ‘scoundrel’ (with a non-Slavic use of the agentive suffix; see below), paskude m, f ‘nasty action/thing; scoundrel’ may < Uk paskudnyk ‘scoundrel’, Pol paskudnik ‘scoundrel; fright’. Perhaps a common form for both animate genders reflects the fact that Persian lacks gender in native and many Arabic words, even though Arabic has gender? The ultimate source of the Yiddish and other Slavic terms is either Aramaic or Iranian, see e.g. OIran pškwč ‘griffin’, as in Oss p’ak’undzä via Geor p’ask’undzi, Arm paskouč and ZoroPahl bškwč < * pati-škuvači- ‘swooping down upon’—an epithet of a raptor bird; see also MPe baškuč ‘winged monster’, Syrc psjnq’ ‘bird of heaven’ and TalmEJAr pjšqnc’, pwšqnc’ /pišqancā/, /pušqancā/, pšqc’ /puškaca, puškuca/(?) ‘female raven’. The Yiddish terms may well have under - gone formal Ukrainianization, to match Uk paskudnyk m ‘foul/nasty thing/ person, scoundrel, loathsome person’, paskuda m, f ‘foul thing; scoundrel’. The prevailing view among Slavists is that the Slavic terms are native < Uk paskúdnyj < pá- + skúdnyj ‘scant, parsimonious; indigent, insufficient’ and a cognate of ščadyty ‘to spare, save, economise; have mercy on’, with pa- mark - ing spatial or temporal nearness, similarity, identity; insufficiency of a quality/ property (Reczek 1991, 55–56). Uk paskúdnyj is unlikely to < pá- + skúdnyj, since the stress does not fall on the first syllable, characteristic of CSl pá-compounds, like the Uk pásynok ‘stepson’ < syn ‘son’. Significantly, Polish lacks a cognate of the Uk skudnyj in this meaning; see, however, OPol (p)oskundzić ‘to blame, censure; dishonour, disgrace’ (the Pol - n- resembles the forms of Ossete, Aramaic and Georgian). I know of only one Slavic example with initial stress, but with /o/ in the first syllable: Br póskudz’ ‘good-for-nothing’, dial. póškudz’ (Stawbcy, west of Minsk) ‘blood-sucking insect; evil person’ (with Persian-like meanings), alongside the Br paskudnik. There is one further curious point: The Uk skudnyj may be a cognate of ZoroPart škand, škast and Pe šekän ‘break’, šekästän ‘to break’ . If Yiddish speakers thought that the notion of ‘(to) break’ was sufficiently close to the meaning of ‘raptor bird, griffin’, i.e. ‘a bird that commits evil’, they might have been inclined to Ukrainianise the form of the Persianism after the — 97 — How Yiddish Polonisms may uncover Iranian, Turkic and other Asian speech … relexification process of Judeo-Slavic to German had ended (thus forming “Yid - dish”). It is highly significant that the Pol paskuda ‘revulsion; monster; creep’ and paskudnik ‘fear; scoundrel’ (like the Belarusian cited above) retain mean - ings closer to the original Irano-Aramaic meanings than the current Yiddish or Eastern Slavic forms. Might this fact reflect the presence of Iranian speakers in White Croatia a millennium ago, who could have preserved the original Iranian meanings? Unfortunately, we have no information on whether the Judeo- Aramaism ever survived in Yiddish, only to be Slavicised subsequently. The Irano-Aramaic term could have become Ukrainianised for two reasons: (i) The idiosyncratic use of a Slavo-Turkic agentive suffix, Y - njak (see details below). (ii) The presence of the term in Persian and contiguous Persianised languages would have reduced the element of secrecy—at least in Iran. Sl -nik m ag. appears o ften in Yiddish, mainly with S lavic roots, but oc - casionally with German and Hebrew components as well. However, the distri - bution of the suffix with Slavic roots in Yiddish differs not infrequently from that of the suffix in other Slavic languages, see e.g. Y nudnik ‘pest, bore’ vs. Pol nudziarz, Uk nudjar, which employ a cognate of Eng, G - er. This fact can be explained in two ways: (i) Yiddish makes independent use of shared Slavic suffixes in order to maintain distance from the coterritorial Slavic languages. (ii) The Yiddish suffix did not enter from a Slavic language. The second point is suggested by - nik and by a second rare Yiddish agentive suffix, also of Slavic origin, - njak m ag., which differs from Slavic distributional norms. This suffix clearly entered Yiddish from a Turkic intermediary. Consider the example of Y paskudnjak ‘scoundrel’, which coexists with the genuine Slavic synonym with Uk -nyk: paskudnik. Y -njak, along with -nik, could have been acquired < Chuv - n’ăk ag., as in Chuv robotn’ăk ‘worker’ < R rabotnik, Uk robitnyk (R *robotnjak, Uk *robitnjak). This would account for the idiosyn - cratic distribution in Yiddish of both the Tc-ESl -njak and -nik, in violation of non-Judeo-Slavic. Hence, while the suffixes are of Slavic origin, they may not have been received by Yiddish directly from a Slavic source. (Research is required to explore the possibility that some instances of Sl - njak in other Slavic languages are motivated by Chuvash or another Old Turkic language.) If Eastern Slavic agentive nouns entered Chuvash at a very early date, they could have been acquired by Old Jewish languages either from Chuvash, or even from the now largely extinct Khazar. I cannot say whether Yiddish originally used the Chuvash variant of the Slavic suffix productively, since after contact with Chuvash was lost, Yiddish could have undergone broad Ukrainianisation, which could have eliminated all but a handful of Y - njak forms. The Y nudnik m, nudnice f ‘bore’, along with nudne ‘boring’, very likely have an Iranian link. Besides having been motivated by the similar He nm’s /nim’as/ ‘boring; tiring’, there is a still closer resemblance with other formally and semantically similar Persian roots: (i) Pe nädžes ‘unclean; defiled, defamed; ugly; bad (name); poor (play); low (spirits), vile (people)’, nèdžasät ‘lack of cleanliness; impurity; dirt, excrement’ (< Arab nadžasa ‘impurity’) may also — 98 — Slavia Centralis 1/2022 Paul Wexler offer a surface congener to Yiddish. (ii) Semantically related Y negišes ‘vexa- tion, persecution’ looks like it derives, on formal grounds, from He ngjšwt /negišut/ ‘accessability’, but, on semantic grounds, it probably derives rather from He ngś /nagas/ ‘he pressed, oppressed; urged’ (with sibilant confusion in Yiddish, which could be a feature acquired < some Iranian and Turkic lan - guages, as well as Polish and Belarusian dialects). Note also YemJArab (ši) nigeš, lit. ‘something’ + ‘that disturbs (me)’ and the formal and semantic paral - lels between Perso-Arabic and Chin náo ‘to trouble, disturb’—another reason to use a Slavism with an idiosyncratic use of a Slavic agentive suffix in Yiddish. 20. Ashk and ModHe paxot o joter ‘more or less’ displays the rare word order of ‘less or more’. There is presently no trace of this expression in Yiddish itself (but see BalkRom majčǝř majbut ‘more or less’ , lit. ‘less-more’). The Hebrew (and Romani?) expression could have been coined on the model of the Pol mniej więcej or USo mjenje (albo) bóle, lit. ‘less (or) more’, the only two Northern Slavic languages with this rare word order. The absence of the term ‘or’ points to a rule productive in Iranian and Chinese, inter alia. Spoken Modern Yiddish has mer-vejniker, mer vincik, fil-venik with German components (~ G mehr oder weniger, with obligatory ‘or’, also possible in Yiddish as mer oder vejniker). Curiously, the initial consonants of both components in mer-vejniker, mer vincik match those of the componentially opposite Pol m(niej) w(ięcej). The original pattern of discourse behind ‘less-more’ might have been Iranian, see e.g. Džuh kəm-ambar (lit. ‘little’ + ‘a lot’), Pe kämabiš (lit. ‘little’ + ‘more’) ‘more or less’. Also, the Polish and Upper Sorbian word order of ‘less-more’ may derive from the Iranian languages once spoken on those two Slavic territories (in addition, the ethno-glottonym Sorb may be of Iranian origin). The Hebroid order of ‘less or more’ was either brought directly from an Iranian language in the Near East, or acquired in Polish or Sorbian lands. It may have survived in Mediaeval Hebrew, because Hebrew was then still an unspoken language, less affected by Slavic considerations than colloquial Yiddish. Finally, see Chin chà bù duō, lit. ‘lacking not more’ and Uk menš(e)-bil’š(e), lit. ‘less more’—following Persian norms? 21. Y a majontek mit gelt ‘large fortune, possessions’ < G ein ‘a’, mit ‘with’, Geld ‘money’, Pol majątek ‘property, fortune’. The Pol majątek itself might have re - placed Pe mal ‘riches, money; property’ (< Arab mal ‘wealth, personal property’). 22. The Y nextn ‘yesterday’ < naxt ‘night’ is componentially similar to the synonymous ‘yesterday’ < ‘evening’, found in USo wčera < wječor, Uk učora < večir, Br zawčora < večar, an d i den ti cal to Uk nič ‘night’/Bojk(Carp)Uk snočy ‘yesterday evening’, because of the latter’s use of ‘night’, rather than the standard Ukrainian use of ‘evening’. The Carpathian facts have a parallel in Serbian and Croat. It is also not out of the question that Bojkian Ukrainian was influenced by the Iranian speech of the White Croats in nearby Galicia (see the map of Western Ukrainian ethnic groups in Kuzela 1963, 1: 281). — 99 — How Yiddish Polonisms may uncover Iranian, Turkic and other Asian speech … Turkic languages also provide an appropriate model for Yiddish: See OTc tün ‘night’, and in some modern Turkic languages, also ‘yesterday’, as well as OttTu dün ‘night’ (14 th c.)/ düne gün ‘yesterday’ (with ‘day’) (16 th c.); CriKar tün ‘night’/ tune gin, tune gün, tünèn ‘yesterday’. See also the near minimal pair of Chin yè ‘night’/(O) yoh ‘yesterday’. 23. The Y (Lith-Br) kitke ‘plaited bread’ could be an original borrowing from Slavic, or a later Slavic replacement for a similar pre-existing Pe gis(u) ‘tress, braid, tail of the hair, woman’s long hair, curls’. I make this suggestion tenta - tively, since I cannot determine whether the Persianism ever denoted a ‘plaited bread product’, but at least the latter ‘hair’-term could have inspired the Y ‘plaited bread’. The bread meaning is absent in both Pol kitka ‘ponytail ’ and Uk kytký plt. ‘braided strands of hair over the ear’ (see the minimal pair with kýtky pl. ‘tufts, crests, plumes’). Hence, ‘plaited bread’ could be a Yiddish in - novation, stimulated by the Y xale, also not originally a bread term (see item 13 above). A link between the /-t-/ of the Yiddishism and /-s-/ of the Persianism might have been created by the facultative change of postvocalic He /t/ > /s/ in the Yiddish pronunciation of Hebrew/Hebroid. However, a White Croat source is put in doubt by two facts: (i) The Pol kitka m eans ‘ po n ytail ’ an d ( ii ) N o Irano-Slavic confederation is known to have existed in Lithuania or Belarus’, which is where we find the sole example of Y (Lith-Br) kitke as a bread term. 6 Conclusion In conclusion, I hope to have shown that, despite its exclusively European geog - raphy at present, Yiddish has maintained the form and function of its Asianisms (and Africanisms) with amazing fidelity. For example, all the Asian languages spoken by a Muslim majority have created a periphrastic verbal conjugation for the integration of Arabic loans (under Persian and possibly also Aramaic influence), using native auxiliary verbs (no parallel rule exists in German for foreign components). While Yiddish has not followed the Islamic languages in incorporating a very large overt Arabic component, it has, nonetheless, joined the Islamic linguistic Sprachbund in developing a periphrastic verbal conjugation in Yiddish for the integration of large numbers of pre-eminently Hebrew and Hebroid elements, which almost always have the Arabic surface congeners used in Persian; these periphrastic verbs usually require auxiliary verbs of German origin (though German lacks a rule parallel to the Persian and Yiddish periphrastic conjugations). It is unclear if Yiddish originally accepted the original Persian Arabisms which it later replaced by Hebrew and Hebroid cognates, or whether the speakers of Yiddish only used Perso-Arabisms as a guide for determining their own Hebrew inventory, but never actually used the Perso-Arabisms overtly in spoken Yiddish. In any event, Yiddish underwent largescale (covert) Arabicisation, without requiring much direct contact with — 100 — Slavia Centralis 1/2022 Paul Wexler Arabic speakers or literature. The same may be true of Romani. Most students of Romani are fond of asserting that the relative paucity of Arabisms in Near Eastern Romani is a sign that the Roma exited Iranian-Iraqi speech territory prior to the Arabic/Islamic invasion—i.e. by the 9 th century . The experience of Yiddish suggests that this assertion is dubious. It may come as a surprise to learn that Yiddish is far more Arabicised (though almost entirely covertly) than Judezmo (Judeo-Spanish, with mainly overt Arabic loans)—spoken in the Iberian Peninsula until the end of the 15 th century, and in Morocco and the successor states of the Ottoman Empire until the present. Thus, Yiddish (with its Slavic grammar and phonology) may be a “peripheral language” vis-à-vis P ers i an, P erso- Arab i c , T urki c an d Chin ese in terms of its geography, but it is a fellow “Irano-Arabo-Turko-Chinese core language” in terms of its abiding respect for the form and meaning of its Ira - nianisms, Turkicisms, Arabisms and Sinicisms. In contrast, Iranian languages spoken in proximity to Persian often alter the form and/or meaning of their Persian loans. Thus, Slavic Yiddish is a “core language” vis-à-vis Persian,, while numerous Iranian languages reveal serious formal and semantic alterations in their Persian component—in tandem with German and the Slavic languages which inherited but largely distorted Iranianisms received from the Jewish merchants and their neighbouring Irano-Slavic confederations. The new merger of “Yiddish (or, more broadly, comparative Jewish) lin - guistics” with “Silk Road Linguistics” that I have propounded here will require not only a close look at all the Afro-Asian languages in contact with Yiddish across the Afro-Eurasian Silk Roads, but also a comparison of Yiddish with its Iranian and Iranianised cousins: Džudezmo, Judeo-Berber, Džuhuri, Judeo- Arabic, Judeo-Persian, Judeo-Georgian, Judeo-Chinese and Karaite (a kind of “Judeo”-Crimean Turkic; Karaism began as a Jewish sect in the 8 th century in Iraq-Iran that rejects the authority of the Talmud; Iranian Karaites were also active on the Silk Roads)—almost all of which developed in the shadow of Iranian. The uniqueness of Yiddish is not only that (i) It (alone or together with Romani) can identify with varying degrees of certainty the locations of Irano-Turko-Slavic confederations in Europe, or that (ii) It has a mass of largely unique Asian influences and a modest number of Africanisms, but that (iii) It appears to be the most faithful to Iranian norms, even after being physically removed from Irano-Turkophone communities for centuries. While Iranian (and other Asian) norms were often lost after the host languages became extinct in the Slavic and German lands, the unusual acceptance of Polonisms in Galician Yiddish is a clue to the missing Asian corpus. The loyalty of Yiddish to its Irano-Turko-Slavic purveyors of linguistic (and even cultural) enrichment can be shown dramatically in the term Y cimes ‘fruit/ vegetable stew’, which I would derive < Chuv şimĕş ‘fruit, vegetables; various foods’; the term was never lost in Yiddish and, in fact, has spread to Polish and Ukrainian. Its survival in Yiddish may depend on the fact that it is not attested in Persian. (This fact raises the question of whether Turkic languages — 101 — How Yiddish Polonisms may uncover Iranian, Turkic and other Asian speech … were becoming extinct along with Iranian, or survived longer in Galicia. Al - ternatively, Asian food and dress terms have enjoyed a striking longevity in Yiddish up to the present.) Surface congeners in other target languages have undergone extreme mutilation, as in Tu yemiš, Damascus, Cairo Arab jamiš, Hung gyümölcs ‘fresh fruit’, Jap ume ‘apricot-like plum (tree)’, Chin méi(zi), Kor mejshe ‘plum’ only, Uk cymes, cimes, cimys ‘tasty food; something very good’, Br cymus ‘the most interesting, essence; sweet food made from carrots; something very tasty’. Brzezina (1979: 87–88) defines JPol cymes (< Yiddish), as ‘a Judaic tasty tidbit; something great/wonderful’ (but derives it < MHG zuomüese or zimbiss > G zum Imbiss). Karłowicz, et al. (1900, 1) gloss the Polish surface cognate as ‘something that is the best; choice; tasty; sour cream; flower; rarity; suitable, worthy’; the Rtw cimes has developed the meanings ‘turnips, carrots, beans’ out of the meaning ‘stewed fruits and vegetables’. Another example of Yiddish “core” behaviour, is the pair of Y šabaš ‘a tip given to musicians at a wedding by guests who dance at the wedding’ + šibeš ‘trifle; small coin’ < Pe šabaš ‘a tip given to musicians at a wedding by guests who participate in the dancing; bravo!’ + (unrelated) sèpas ‘gratitude, thanks’. In contrast, see the semantically and/or formally distorted Neo-KurdJAr šābāše ‘wedding gifts; money for musicians’; Geor šabaš ‘w edding tip to the musi - cians––both minus the condition that only guests who dance at the wedding party need give the tip; Se džȁba ‘hospitality; for free ’ ( with a T urkish form and meaning); R šabáš ‘rest time; enough!’, R, Br, Uk šabáški ‘wood cuttings taken home by carpenters from work (for kindling)’ (with a Slavo-Iranian di - minutive suffix), R dial. baš(li) ‘money; a two-kopeck coin’ (with a possibly Iranian diminutive suffix), bašljat’ ‘to earn money’, bašlëvyj ‘having a lot of money; financial’; Pol szabas(z) ‘finished; enough!’; truncated Hung sáp [šap] ‘pimp’s or brothel owner’s profit; hush money; metal or wood cuttings’. All the languages except Yiddish lack a reflex of the Pe sèpas ‘gratitude, thanks’. A second term is possible in European target languages, but only in truncated form (as in the Russian and Hungarian examples). It remains to be seen whether the Asian Islamic languages, like Yiddish, largely imitate the Arabic inventory of Persian, despite independent access to Islamic literature in Arabic. For example, unlike Persian, Dari and Tadjik are much poorer in Arabic components. It is important to remember that citing parallels between Yiddish and Asian languages, such as Iranian and Chinese, has a strong and weak interpretation: In the stronger case, Asian Silk Road languages are quite likely the sources of Yiddish features—many of which were further passed on to Slavic languages and German; in the weaker case, Yiddish and the other Asian languages are at the very least members of common Eurasian isogloss bundles. Chinese Jews still recall their Iranian roots; Ashkenazic Jews do not. It is time for the latter to recognise that the popular narrative that Yiddish developed in Germany as a variant of Middle High German in the 9–10 th centuries is un - supported by facts. Yiddish began as a cryptic trade language, in tandem with a — 102 — Slavia Centralis 1/2022 Paul Wexler number of other Old Jewish languages, exploiting the process of relexification. It probably began when Judeo-Persian became obsolete in the Khazar Empire, and was replaced by a newly emerging merged Judeo-Slavic language (specifically, Sorbian and pre-Ukrainian and pre-Belarusian). In the Khazar Empire, Judeo- Slavic acquired a mass of Iranian and other Asian components—including a smattering of Mongolian and Tokharian elements (see examples in my 2021), probably when Judeo-Turkic and Judeo-Iranian became obsolete there. Judeo- Slavic, in turn, became relexified to German mainly in the Southern German lands (and possibly through an intermediary stage of Gothic enrichment in the Khazar Empire, though some isolated Judeo-Slavic dialects survived roughly up to the 13 th century in Czech territories, and up to the early 1 7 th century in Belarusian territories) to become the Iranianised Slavic language called Yiddish ‘Jewish’ since the 16 th century. The chain of three Jewish mercantile languages is matched by an identical chain of semantic changes that affected the ethno - nym Ashkenaz(ic) that I mentioned at the beginning of this article: From the meaning of ‘Iranian-speaking Jews’ to, and overlapping with ‘Slavic-speaking Jews’ to, and overlapping with ‘Yiddish-speaking Jews’. The White Croat State appears to have disappeared by the end of the 10 th century, when some Iranian Croats migrated to the northwestern Balkans to found a new Croatian State, and others remained in Galicia as speakers of Polish and Ukrainian. It was at that point—shortly after the demise of the White Croat State—that some Polish Jews, speaking either unrelexified “Judeo- Slavic” or “Judeo-Slavic, newly relexified to the German lexicon (i.e. Yiddish)” would have begun to incorporate a small but not trivial number of Polonisms (some of which spread to neighbouring Ukrainian Galician Yiddish) to replace the dwindling corpus of obsolescent Iranian speech—probably consisting of Judeo-Iranian and other Asianisms brought to Galicia from the Near East a few centuries earlier, by both Jewish and non-Jewish Iranians. I hope that, in the near future, other scholars will join me in seeking to expand our knowledge of (i) The Afro-Asian corpus of Yiddish and other Old Jewish languages developed for use on the Silk Roads, (ii) The mutual influ - ences between Jewish languages and the Iranianised Indic speech of the Roma (among the presumed partners of the Radhanite merchants on the Silk Road caravans) and (iii) The locations of all non-Jewish trading confederations that were active in the White Croat State. REFERENCES Harold W. BAILEY, 1979: Dictionary of Khotan Saka. Cambridge, London, New York, Melbourne: Cambridge University. Émile BENVENISTE, 1970: Old Iranian. Current Trends in Linguistics. Ed. Thomas Sebeok. The Hague: Mouton. 6, 9–25. Johannes BENZING, 1959: Das Tschuwaschische. Philologiae turcicae fundamenta. Eds. Jean Deny, et al. Wiesbaden: Aquis Mattiacis, Steiner. 1, 695–751. — 103 — How Yiddish Polonisms may uncover Iranian, Turkic and other Asian speech … Uwe BLÄSING, 2002: Ein Beitrag zu Zigeuner-Jargons, Geheimsprachen und zum argotischen Wortschatz in der Türkei. Einige Bemerkungen zu Caferoğlus Liste aus dem Wortschatz der Elekçi bei Alaçam (Bolu). Iran and the Caucasus 6/1–2, 103–168. Maria BRZEZINA, 1979: Języki mniejszości narodowych w tekstach literackich i folk- lorystycznych 1. Południowokresowa polszczyzna Žydów. Warsaw-Krakow: Państwowe wydawnyctwo. David BUYANER, 2005: On the etymology of Middle Persian baškuč ( ‘winged mon - ster’). Studia iranica 4, 19–30. Gerald CLAUSON, 1972: An etymological dictionary of pre-thirteenth-century Turkic. Oxford: Clarendon. Jan CZEKANOWSKI, 1957: Wstęp do historii słowian. Poznań: Instytut Zachodni. Karl DIETERICH, 1931: Zur Kulturgeographie und Kulturgeschichte der byzantinischen Balkanländer. Byzantinische Zeitschrift 31, 37–57, 334–50. Francis DVORNIK, 1949: The making of Central and Eastern Europe. London: Polish Research Centre. Richard ETTINGHAUSEN, 1972: From Byzantium to Sasanian Iran and the Islamic world. Three modes of artistic influence. Leiden: Brill. Johann FÜCK, 1955: ‘ Arabīya. Recherches sur l’histoire de la langue et du style arabe. Paris: M. Didier. Dietrich GERHARDT, 1974–1977: “Vogelmilch”–Metapher oder Motiv. Semantische Hefte 2, 1–77; Semantische Hefte 3, 81–86. Jacob GERZON, 1902: Die jüdisch-deutsche Sprache. Eine grammatisch-lexikalische Untersuchung ihres deutschen Grundbestandes. Frankfurt am Main: J. Kauffmann. Albrecht GÖTZE, 1923: Persische Weisheit in griechischem Gewande. Zeitschrift für Indologie und Iranistik 2, 60–98, 167–177. Benjamin H. HARY, 2009: Translating religion. Linguistic analysis of Judeo-Arabic sacred texts from Egypt. Leiden-Boston: Brill. Friedhelm HINZE, 1974: Zum Stand der Erforschung des Dravänopolabischen. Zeitschrift für Slavistik 19, 507–512. Wladimir IVANOW, 1922: An old Gypsy-Darwish jargon. Journal of the Asiatic Society of Bengal n. s. 18, 375–383. Anatolij IVČENKO, 1998: Zasady etymolohičnoho analizu ukrajins’koji narodnoji frazeolohiji. Zbirnyk Xarkivs’koho istoryko-filolohičnoho tovarystva n.s. 7, 105–122. Jan KARŁOWICZ, Adam KRYŃSKI and Władysław NIEDŹWIEDZKI (eds.), 1900- 1927: Słownik języka polskiego 1-8. Warsaw: Nakładem prenumeratorów. Franciszek KMIETOWICZ, 1978: Die Titel der Slawenherrscher in der sog. “Anony - men Mitteilung”, einer orientalischen Quelle (Ende des neunten Jahrhunderts). Folia orientalia 19, 13–33. Konstantin KOSTOV, 1971: Nasleden li e ot praslavjansko vreme dnešnijat slavjanski poslovičen izraz ot ptiče mljako? Bălgarski ezik 21/2–3, 255–259. Zenon KUZELA, 1963: Tribal division and ethnographic groups. Ukraine. A concise encyclopaedia. Ed. Volodymyr Kubijovyč. Toronto: Ukrainian National Association, University of Toronto. 1, 280–286. — 104 — Slavia Centralis 1/2022 Paul Wexler Gerard LABUDA, 1949: Pierwsze państwo słowiańskie. Państwo Samona. Poznań: Księgarnia Akademicka. Tadeusz LEWICKI, 1954: Polska i kraje sąsiedne w świetle “Księgi Rogera” geografa arabskiego z XII w. al-Idrîsî’ego 2. Krakow-Warsaw: Polska Akademia Umiejętności. Henryk ŁOWMIAŃSKI, 1963: Początki Polski: Z dziejów Słowian w 1 tysiącleciu n.e. 2. Warsaw: Państwowe Wydawnictwo Naukowe. Douglas M. MACDOWELL (ed.), 1971: Aristophanes, Wasps. Oxford: Oxford University. Hrach K. MARTIROSYAN, 2010: Etymological dictionary of the Armenian inherited lexicon. Leiden, Boston: Brill. Yaron MATRAS, 2002: Romani: A linguistic introduction. Cambridge: Cambridge University. Manfred MAYRHOFER, 1956–1980: Kurzgefasstes etymologisches Wörterbuch des altindischen. Heidelberg: C.Winter. Margaret C. MILLER, 1997: Athens and Persia in the fifth century BC: A study in cultural receptivity. Cambridge: Cambridge University. Ivan Ivanavič NOSOVIČ, 1870: Slovar’ belorusskogo narečija. St. Petersburg: Izdanie Otdelenija Russkogo Jazyka i Slovesnosti Imperatorskoj Akademii Nauk. Hryhoryj P. PIVTORAK, 1993: Ukrajinci: Zvidky my i naša mova? Kiev: Naukova dumka. John T. PLATTS, 1960: A dictionary of Urdū, Classical Hindi and English. Oxford: Oxford University. Józef RECZEK, 1991: Polszczyzna i inne języki w perspektywie porównawczej. Wrocław, Warsaw, Krakow: Zakład Narodowy im. Ossolińskich. Jurij Aronovič RUBINČIK, 1970: Persidsko-russkij slovar’ 1 –2. Moscow: Sovetskaja ènciklopedija. George Y. SHEVELOV, 1965: A prehistory of Slavic. The historical phonology of Com- mon Slavic. New York: Columbia University Press. – –, 1979: A historical phonology of the Ukrainian language. Heidelberg: C. Winter. Francis Joseph STEINGASS, 1892: A comprehensive Persian-English dictionary, in- cluding the Arabic words and phrases to be met with in Persian literature. London: Routledge and Kegan Paul. An on-line reverse English-Persian version of this dictionary is available at dsal.uchicago.edu. Ralph L. TURNER, 1966: A comparative dictionary of the Indo-Aryan languages. London-New York-Toronto: Oxford University. Max VASMER, 1923: Die Iranier in Südrussland. Leipzig: Markert and Petters. – –, 1953–1958: Russisches etymologisches Wörterbuch 1-3. Heidelberg: Carl Winter. Max WEINREICH, 1956: Yiddish, Knaanic, Slavic: the basic relationships. For Roman Jakobson. Essays on the occasion of his sixtieth birthday. Eds. Morris Halle, Horace G. Lunt, Hugh McLean and Cornelis H. van Schooneveld. The Hague: Mouton. 622–32. – – [VAJNRAJX], 1973: Gešixte fun der jidišer šprax 1–4. New Y ork: YIVO. English translation, trsls. Shlomo Noble and Joshua A. Fishman. Ed. Paul Glasser. New Haven 2008: Yale University. — 105 — How Yiddish Polonisms may uncover Iranian, Turkic and other Asian speech … Paul WEXLER, 1986: Exploring the distinctive features of Wandersprachen. (The case of European Romani and Jewish languages.) Mediterranean language review 2, 5–47. – –, 1991: Yiddish—the fifteenth Slavic language. International journal of the sociology of language 91, 1–150, 215–225. – –, 1992: The Balkan substratum of Yiddish. A reassessment of the unique Romance and Greek components. Wiesbaden: Harrassowitz. – –, 1997: The relexification hypothesis in Rumanian. Relexification in creole and non- creole languages. (With special reference to Modern Hebrew, Haitian Creole, Romani and Rumanian). Eds. Julia Horvath and Paul Wexler. Wiesbaden: Harrassowitz. 162–188. – –, 2002: Two-tiered relexification in Yiddish. (Jews, Sorbs, Khazars and the Kiev- Polessian Dialect). Berlin: Mouton de Gruyter. – –, 2021: Silk Road Linguistics: The birth of Yiddish and the multiethnic Jewish peoples, 9–13 th centuries. (The indispensable role of the Arabs, Chinese, Iranians, Slavs and Turks.) Wiesbaden: Harrassowitz. Ehsan YARSHATER, 2002: The Jewish dialect of Kāshān. Jerusalem studies in Arabic and Islam 27, 439–462. KAKO LAHKO POLONIZMI JIDIŠA RAZKRIJEJO IRANSKO, TURKOTATARSKO IN DRUGO AZIJSKO GOVORICO V GALICIJSKI BELI HRVAŠKI (POGLAVJE NA NOVEM PODROČJU ‘LINGVISTIKE SVILENE POTI’) Narečja poljskega in vzhodnoslovanskega jidiša imajo skromen delež polonizmov in znaten delež vzhodnoslovanskih posebnosti. Izjema je galicijska Bela Hrvaška (mešana slovansko-iransko-turkotatarska zveza, ki je razpadla do 10. stoletja). Tam ima jidiš precej več poljskih posebnosti, saj so veliko nezapisanih iranskih in turkotatarskih iz - razov tamkajšnjega jidiša zamenjali polonizmi podobne oblike in pomena. Ti ‘prikriti irano-turcizmi’ pomagajo obnoviti izumrlo iransko in turkotatarsko izrazje judovskih in nejudovskih govorcev. Primer za to je jid. pente ‘veriga’, penten ‘ukleniti’, culončen ‘prikleniti’ < polj. pęto, pętać, łączyć, połączyć za perz. pabänd ‘okovi’, molhäg ‘zvezan’. Tu pa tam se poljske zamenjave pokažejo za perzijske, turkotatarske in druge azijske. Nekaj sorabizmov in germanizmov v jidišu vrh tega kaže oblikovne in pomenske zveze z azijskimi posebnostmi (prim. mnenje, da so Lužiški Srbi delno iranskega izvora). Naštel sem tudi nekaj nezamenjanih azianizmov v poljsko-ukrajinskem jidišu. Ti prika - žejo obogatitev jidiša z azianizmi in afrikanizmi na Svilni poti. Možni vzroki, zakaj je galicijski jidiš (ob vseh drugih narečjih jidiša v Vzhodni Evropi) ohranil toliko azianiz - mov, ki jih poljščina ni zamenjala, so naslednji: (a) vedno ni bilo primernih polonizmov; (b) na poljski jidiš so vplivala negalicijska narečja; (c) veliko govorcev jidiša je opustilo iranščino in turkotatarščino preden so galicijski Turkotatari in Iranci izginili in zato niso videli potrebe za zamenjavo; (d) nekateri azianizmi v okviru judovske vere in kulture so se že hebraizirali, prim. jid. xale ‘praznični sobotni kruh’, zelo verjetno kitajskega oziroma iranskega oziroma arabskega izvora, pripisoval pa se je biblični hebrejščini; (e) nezamenjani azianizmi so krožili med nejudovskimi govorci. Če zadnja razlaga drži, bi bilo razvidno, katere azijske jezike so govorili pripadniki slovansko-iranske in slovansko-turkotatarske zveze na slovanskih tleh. Nadaljnje raziskave naj določijo, ali ima poljščina azianizme, ki jih jidiš ne pozna.