Skip to main content

Full text of "Notes on Judeo-German Phonology"

See other formats


STOP 



Early Journal Content on JSTOR, Free to Anyone in the World 

This article is one of nearly 500,000 scholarly works digitized and made freely available to everyone in 
the world by JSTOR. 

Known as the Early Journal Content, this set of works include research articles, news, letters, and other 
writings published in more than 200 of the oldest leading academic journals. The works date from the 
mid-seventeenth to the early twentieth centuries. 

We encourage people to read and share the Early Journal Content openly and to tell others that this 
resource exists. People may post this content online or redistribute in any way for non-commercial 
purposes. 

Read more about Early Journal Content at http://about.jstor.org/participate-jstor/individuals/early- 
journal-content . 



JSTOR is a digital library of academic journals, books, and primary source objects. JSTOR helps people 
discover, use, and build upon a wide range of content through a powerful research and teaching 
platform, and preserves this content for future generations. JSTOR is part of ITHAKA, a not-for-profit 
organization that also includes Ithaka S+R and Portico. For more information about JSTOR, please 
contact support@jstor.org. 



NOTES ON JUDEO-GERMAN PHONOLOGY 
By E. Sapir, Ottawa. 

A GRATIFYING phase of Germanic study in recent years 
is the constantly increased attention paid to the modern 
spoken dialects. That the dialects still spoken by the 
rural population of Germany, for instance, have often 
preserved archaic features in vocabulary, phonology, mor- 
phology, and syntax, where the literary ' Gemeinsprache ' 
is less conservative, is well known. Thus, attention may 
be called in passing to the fact that many of the dialects 
in Middle and Upper Germany still observe the distinction 
in pronunciation between short open e (<O.H.G. and 
M.H.G. e, as in ge'ban, geben) and short close e due to 
z'-umlaut of a (as in O.H.G. bemro, cf. Gothic battza), 
while, as is well known, the ' Gemeinsprache ' has levelled 
the distinction completely. By such archaic features the 
modern dialects are often able to throw a great deal of 
light on the history of the language ; moreover, they are 
generally more easily handled, from the purely linguistic 
standpoint, than the literary monuments of Old and Middle 
High German, in that they are immediately accessible to 
study and are not distorted, particularly in regard to 
phonetics, by orthographic imperfections. 

While the German dialects now spoken within the 
confines of Germany, Austria, and Switzerland are being 
diligently and profitably studied, little has as yet been 

231 



232 THE JEWISH QUARTERLY REVIEW 

done in the way of scientifically examining the various 
dialects spoken by the Jews of Lithuania, Russian Poland, 
Galicia, southern Russia, and Roumania. 1 When one 
recollects that these Judeo-German or 'Yiddish' dialects 
have, since the beginning of the modern period (in the early 
part of the sixteenth century), developed in comparative 
isolation from the main body of German dialects and that 
they have been subjected to the influence, chiefly lexical, 
of the Slavic vernaculars (Polish, Russian, and Little 
Russian) on the one hand, and of the sacred Hebrew 
tongue on the other, it becomes clear that we are here 
dealing with a complex of linguistic conditions that must 
prove highly instructive to the student of language. 2 

1 Besides Leo Wiener's two articles on Judeo-German in The American 
Journal of Philology, XIV, pp. 41-67 and 456-82 (phonologically unreliable 
because modern literary German, instead of Middle High German, is taken 
as the point of departure) and L. Saingan's study (' Essai sur le Judeo- 
Allemand et specialement sur le dialecte parl6 en Valachie ') in Memoires de 
la Socie'te de Linguistique de Paris, XII, pp. 90-138, 176-96 (treats of 
Roumanian Judeo-German), we have Jacob Gerzon's Die jiidisch-deutsche 
Sprache, eine grammatisch-lexikalische Untersuchung ihres deutschen Grund- 
bestandes (Frankfurt am Main, 1902), treating mainly of the Lithuanian 
Judeo-German of Homel (Government of Mohilev). Valuable as Gerzon's 
work is, it is much less satisfactory in its treatment of the phonology (pp. 
20-35) tnan of 'he morphology and syntax ; in particular Gerzon has failed 
to point out the absence of quantitative differences in the vowels of stressed 
syllables and the development of voiced stops in final position, both of which 
are characteristic features of Judeo-German when contrasted with other 
High German dialects. The present study, though late to appear, was 
completed before access was had to Gerzon's work, so that the material 
here presented is the result of independent investigation. The dialect here 
treated is the form of Lithuanian Judeo-German spoken in the Government 
of Kovno. Further references to works on special points in Judeo-German 
may be found in L. Wiener's History of Yiddish Literature in the Nineteenth 
Century (New York, 1899), pp. 12-24 (chapter on ' The Judeo-German 
Language'). 

a The following taken from Gratz's Geschichte der Juden (vol. 9, p. 64) 



JUDEO-GERMAN PHONOLOGY — SAPIR 233 

The conditions are, in fact, not dissimilar to those that 
obtained in the development of the English language — 
isolation from the main body of the vernacular and con- 
siderable foreign influence. On the whole, the student 
of Judeo-German will be inclined to see a less extensive 
foreign influence in the case of Judeo-German than in 
that of English ; the basis has remained thoroughly German, 
the foreign accretions and influences are, at best, of only 
secondary importance. 

Before proceeding to the sketchy phonological obser- 
vations I have to offer, it may not be inappropriate to 
call attention, by way of illustration, to some of the more 
interesting archaic features that Judeo-German presents. 
In vocabulary many Middle High German words now 
obsolete or, at any rate, not in common use in literary 
German, have been preserved in full vigourby Judeo-German. 
Such are {dm 'son-in-law' (<M.H.G. eidem) ; sver 'father- 
in-law' (<M.H.G. sweher) ; snur 'daughter-in-law' 

will serve as historical basis of the above remarks. GrStz's statements apply 
to the period 1496-1525. ' Aber nicht bloss deutsche Talmudkunde haben die 
jfldisch-deutschen Flachtlinge nach Polen verpflanzt, sondern auch die deutsche 
Sprache— in ihrer damaligen Beschaffenheit ; sie impften sie den eingeborenen 
Juden ein und verdrangten nach und nach aus deren Munde die polnische 
Oder ruthenische Sprache. Wie die spanischen Juden einen Teil der euro- 
paischen oder asiatischen Tiirkei in ein neues Spanien verwandelt haben, so 
machten die deutschen Juden Polen, Littauen und die dazu gehorigen 
Landesteile gewissermassen zu einem neuen Deutschland. . . . Mehrere 
Jahrhunderte hindurch zerfielen daher die Juden in spanisch Redende und 
deutsch Sprechende, gegen welche die Italiens als eine wenig zahlende Klasse 
verschwand, da auch hier die Juden Spanisch oder Deutsch verstehen 
mussten. . . . [Die polnischen Juden] verehrten [die deutsche Sprache] wie 
ein Palladium, wie eine heilige Erinnerung, und wenn sie sich auch im 
Verkehr mit Polen der Landessprache bedienten, im trauten Familienkreise, 
im Lehrhause und im Gebete behielten sie das Deutsche bei. Sie gait ihnen 
nachst dem Hebraischen als eine heilige Sprache.' 



234 THE JEWISH QUARTERLY REVIEW 

(<M.H.G. snur); tor 'dare' (<M.H.G. tar,gitar); zdgr 
'clock' (<M.H.G. seiger) ; haint 'to-day' (<M.H.G. 
hint ' this night ') ; and many others. In phonetics, Judeo- 
German has, for instance, not levelled M.H.G. i and ei into 
ai, but has kept them apart as at and e respectively ; e.g. 
vats 'white' and ix vis 'I know' (< M.H.G. wiz, and ich 
weiz, respectively; contrast modern literary German weiss 
for both). In the case of zamd ' sand ' an Indo-Germanic 
m has been preserved that has in practically all other 
Germanic dialects been assimilated to n — cf. Greek &/j.a6os 
' sand ' < *samadhos. A large number of archaic features 
are found also in the morphology, The old dative singular 
in -en of weak feminines (M.H.G. der zungen, der mitten, 
but modern German der Zunge, der Mitte) is preserved 
in stereotyped phrases like in dr mitn drin ' right in the 
midst of it '. The M.H.G. feminine noun heit ' manner ' 
preserved in modern German only as derivative suffix in 
abstract nouns (e.g. Kiihnheit, Menschheit) survives in 
Judeo-German in adverbial genitives in -r (h)e"t (e.g. blindr dt 
' blindly ' < blinder heit). The preterito-present verb M.H.G. 
touc has in modern German been levelled to the great class 
of other verbs, while Judeo-German still has er tig ' he is of 
account' (contrast modern German er taugt). The old 
imperative Id ' let ' survives in phrases like 16 ntir (or 16 mix) 
' let me ' (contrast modern German lass mich). In syntax, 
the double negative may be mentioned as an archaic 
feature, though something should here be perhaps ascribed 
to Slavic influence. 

It would, however, be erroneous to suppose that the 
Judeo-German dialects are on the whole more archaic than 
modern literary German. They are not. In morphology 
particularly great simplification has taken place. The 



JUDEO-GERMAN PHONOLOGY — SAPIR 235 

preterite has disappeared in favour of the periphrastic perfect 
(e. g. er hot gtzin = er sah). The dative and accusative (at 
least in Lithuanian Judeo-German, which dialect alone is 
here considered) have disappeared as such and have been 
merged into an objective case, partly dative and partly- 
accusative in form (e.g. er git mir = er gibt mir ; er zit 
mir = er sieht mich). The ending -er preceded by umlaut 
and umlaut alone have greatly spread as plural signs (e.g. 
pletsr 'places'; teg ' days' <*tege for tage). The umlaut 
of the second and third persons singular of strong verbs 
has in most cases been levelled out (er zit = er sieht; er 
fait = er fallt; er le"ft = er lauft, cf. ix le"f= ich laufe). 
A number of weak verbs have followed the analogy of 
strong verbs in their participle (e. g. gfkrSgn ' obtained ' 
as participle of krign by analogy of such verbs as fardrisn 
— -fardrdsn; gtsdtn = geschiittet; ongltsimdn = angeziin- 
det). The third person reflexive has been generalized for 
all persons and numbers (e.g. ix zets mix= ich seize mich) — 
this is undoubtedly due to Slavic influence. There are 
many other levellings and analogical developments that 
have taken place in Judeo-German. 

Several interesting special developments that have 
taken place are : a gerund of adverbial force in -dig, which 
can be formed from any verb by suffixing this syllable 
to the infinitive (e.g. er vent le/ndig 'he cries while 
running ' ; these forms in -ndig are doubtless based on 
M.H.G. participial forms in -ende, perhaps influenced by 
lebendic ' alive ') ; a monosyllabic abstract noun which can 
be formed from any verb and which is used in phrases like 
er git a smik 'he gives a smell, he smells (momen- 
taneously) ' ; the transfer of most neuter nouns to the 
feminine gender (e. g. di hoiz ' the house '< M.H.G. daz Ms; 



236 THE JEWISH QUARTERLY REVIEW 

a similar development has taken place in Lithuanian, in 
which old neuters have generally become masculines, e. g. 
drklas m. ' plough ' as contrasted with Latin aratrum and 
Greek aporpov). 

In phonology two great revolutions have taken place 
in Judeo-German. In the first place, the quantitative 
vocalic differences that are so important in modern German 
(contrast siech with sich, schal with Schall, Sohn with Sonne, 
Musse with muss) are not found in Judeo-German. All 
accented vowels are of practically uniform length — approxi- 
mately midway in quantity between the German long and 
short vowels ; the quality of * and u is that of the German 
long i and u, in other words close. Thus, the vowel of 
Judeo-German zix 'himself is pronounced like that of 
German siech, as far as quality is concerned, but with 
a shorter quantity (yet not so short as in German sick); 
correspondingly with Judeo-German u. Judeo-German o is 
in quality identical with the German in voll ; there are 
two ^-vowels, an open e (as in German Mensch) and a close 
e (as in German geben, barring quantity) ; a does not differ 
in quality from the normal German a. We might put the 
matter thus : there are no long i, u, 0, a, e in Judeo-German. 
This radical difference in phonetic basis between Judeo- 
German and modern standard German I am inclined to 
explain by Slavic influence (the same lack of quantitative 
differences in accented vowels obtains in Russian and 
Polish ; thus, Russian accented i is medium in quantity 
between German I and i). 

The second phonetic revolution referred to is the rise 
of final voiced stops and spirants. In Middle High German 
and its modern representatives a voiced (lenis) stop or spirant 
becomes voiceless (fortis) when final (M.H.G. tages, tac; 



JUDEO-GERMAN PHONOLOGY — SAPIR 237 

modern German Todes, Tod, i.e. tot). In Judeo-German, 
however, a final sonant is not pronounced as surd, but 
preserves its sonant character ; thus, zSgn ' to say ' : ix zdg 
' I say '. I do not believe that sonants when final have 
really remained sonant. I prefer to explain the phenomenon 
by analogy. Original M.H.G. tac (= tak) tages tage was 
levelled to tag tages tage ; when final -e later dropped, the 
g could no longer become surd, hence we have Judeo- 
German tog corresponding to Modern German tak (or tax) — 
tags (or tays). Similarly, veg ' road' < M.H.G. wee by 
analogy with weges w'e'ge (but modern German vex 1 — vegss 
or vejds). That this explanation is correct is indicated by 
such words as op ' away '< M.H.G. abe, where no paradig- 
matic levelling could take place and where final b became 
/, according to regular German phonetic law ; cf. also avik 
'away' (= German weg) as adverb with veg 'road' as 
noun (the adverb was not associated with the noun, hence 
suffered no levelling). In any event, the great frequency 
of final voiced stops and spirants in Judeo-German is a 
feature that is entirely foreign to the main body of German 
dialects but is paralleled within Germanic by English and 
Swedish. 

In the following is given in brief the development in 
Judeo-German of the Middle High German vowels and 
consonants, no claim of absolute completeness of treatment 
being made. The main lines of change must suffice. 

Vowels. 

1. M.H.G. a. 

a. In closed syllables it remained unchanged: gast< M.H.G. 

gast ; vald< M.H.G. wait (ivald-) ; gam ' goose ' < M.H.G. 

gans; Aart<M.K.G. hart; az<M.K.G. als ; drfot< 

M.H.G. arbeit, arebeit ; ba/d <M.H.G. balde ; land< 



238 THE JEWISH QUARTERLY REVIEW 

M.H.G. Ian/ (land-); ka/z< M.H.G. na/s ; ga>;g< M.H.G. 
ganc {gang-); nax/<M.H.G. naht. In open syllables 
followed by x (originally geminated, O.H.G. -hh-) it also 
remained, as in modern German : mdxn < M.H.G. machen ; 
Idxn <M.H.G. lachen. 

b. In originally open syllables (in some cases now secondarily 
closed) it became lengthened to a (cf. modern German 
a< a in open syllables), which, falling in with original a, 
developed to open 0: hdbn< M.H.G. haben; jdgn< M.H.G. 
jagen ; vdgn ' waggon ' < M.H.G. wagen ; op < M.H.G. abe ; 
ndmn < M.H.G. name, namen ; fd/r< M.H.G. vater. Many 
cases of o<a in originally closed syllables are readily ex- 
plained by paradigmatic analogy : tog< tac (cf. (age). 
Original tac tdge, pi. /age first developed to tac /age, /age, 
then, with consonantic levelling, to tag tage, tage ; when 
d>o, this series became tag tSge, tdge; vocalic levelling 
gave tog tdge, tdge ; dropping of final unaccented -e would 
have reduced these forms to tog tog, tog, to avoid which 
umlaut as characteristic of noun plurals came in by 
analogy ; as final result we have to-day nom. tog, dat.-acc. 
tog, pi. teg. Other examples of analogical o<a in closed 
syllables are: hot< M.H.G. stat ; groz<M.tt.G. gras. 
In certain words a became lengthened before r to a. even 
in closed syllables ; this a also resulted in 0: gor< M.H.G. 
gar (cf. modern German gar) ; bort< M.H.G. bar/; bdrvds 
' barefoot '< M.H.G. barvuoz,; tor, /ors/ ' (he) dares, (you) 
dare'< M.H.G. /ar, /ars/. More difficult to explain are 
dos< M.H.G. daz, and vos< M.H.G. waz,; perhaps these 
forms arose in combinations like daz, is/ (originally sylla- 
bified, before 'fester Einsatz' developed before is/, as 
dazfst)>ddz, ist>dos iz. 

c. Cases of e< a are probably only apparent. meg{= modern 
German mag) is probably not directly developed from 
M.H.G. mac, but is due to analogy of 1st and 3rd person 
plural present indicative and infinitive migen (upper 
German) >Judeo-German megn (see 4. below), ken (= 



JUDEO-GERMAN PHONOLOGY — SAPIR 239 

modern German kann) is similarly not directly developed 
from M.H.G. kan (>Judeo-German parallel form kon, see 
b above), but is due to analogy of kennen 'to know'> 
Judeo-German Mnn. 

2. M.H.G. d. 

a. This sound regularly became 0, which is in no respect 

phonetically different from o< M.H.G. or M.H.G. a in 
open syllables: on 'without '< M.H.G. dne ; do< M.H.G. 
da; nox<M.H.G. ndch; hor 'hair' < M.H.G. Mr; jor< 
M.H.G. jar; mol< M.H.G. mdl ; hot 'has', host 'hast', 
hot '(ye) have'<M.H.G. hdt, hast, hdt; Mo 'blue'< 
M.H.G. bid ; gro ' gray '< M.H.G. grd; lo < let ! ' < M.H.G. 
Id; gir6tn< M.H.G. gerdten; nont 'near' < M.H.G. ndhent; 
mon< M.H.G. man, mdhen 'Mohn'. Note that Judeo- 
German sometimes preserves as reflex of M.H.G. d where 
modern German has shortened d to a (contrast Judeo- 
German nox with modern German nach; host, hot with 
hast, hat). 

b. It is shortened to a (as in modern German) before xt: gtddxt 

< M.H.G. geddht; glbrdxt< M.H.G. gebrdht. 

c. In vu ' where ', d of M.H.G. wd, after being labialized to 6 (cf. 

modern German wo), became still further labialized to u. 

3. M.H.G. e\ 

a. This sound normally remained as open e : erd< M.H.G. 

i'rde ; ber 'bear'<M.H.G. b'er ; w#<M.H.G w'irlt; llext 

< M.H.G. sleht; feld< M.H.G. vilt {veld-) ; helfn < M.H.G. 
he'lfen ; zeks< M.H.G. s'ehs. It is to be particularly noted 
that e in open syllables did not, as in most dialects, 
lengthen to e (> Judeo-German e), but remained open e : 
Ubn< M.H.G. li'ben (contrast modern German leben, i.e. 
lebn); b/zm < M.H.G. beseme 'Besen'; nemn< M.H.G 
ne'men; Mzn< M.H.G. le'sen ; be'tn 'to ask for' (= modern 
German bitten) < M.H.G. b'e'ten 'bitten (um Almosen)'; 
gebn< M.H.G. geben. 

b. M.H.G. -ehe- regularly contracted to e (not, as in modern 



240 THE JEWISH QUARTERLY REVIEW 

German, to e>e): (sen 'ten'< M.H.G. z'e'hen ; zen 'to 
see'<M.H.G. se'hen; fow'father-in-law'<M.H.G. sweher. 

c. Before r plus consonant, <?' regularly became broadened to a 

(cf. English farm< Middle English ferrri) : barg 'hill, 
mountain '< M.H.G. b'e'rc {berg-); harts< M.H.G. he'rze ; 
fdrtsn 'to break wind '< M.H.G. vi'rzen ; vdrfn< M.H.G. 
we'rfen ; ltdrbn< M.H.G. ste'rben ; varg (e.g. grinvarg 
'green stuff, vegetation ')< M.H.G. w'erch, were (modern 
German Werg ' tow '). e remains, however, in erd ' earth ' 
<M.H.G. irde. 

d. e' appears as i'm biln 'to bark '< M.H.G. be'llen. This may 

be due to i of M.H.G. singular present indicative bille, 
billest, billet, though ordinarily e is generalized in Judeo- 
German (cf. helft = modern German hilft). 

4. M.H.G. e (i- umlaut of a). 

a. In originally closed syllables this sound fell in, as in modern 

German, with e< M.H.G. e\ Examples of e<e are : end< 
M.H.G. ende ; be'sr < M.H.G. bez&r ; mens< M.H.G. 
mensche; s ve'nt six ' it depends ' (=es wendet sich)< M.H.G. 
wenden; epl ' apple '< M.H.G. epfel (plural of apfel, but 
also used as singular ; cf. Kluge's remark : ' in Schwaben, 
der Schweiz und der Oberpfalz ist das plurale 4#^Singular- 
form geworden'); sme'kn 'to smell '< M.H.G. smecken 'to 
taste, to smell ' (Kluge remarks : ' die Bedeutung "riechen " 
wahren das Alemannische und Baierische, auch das 
Hessische teilweise '). 

b. M.H.G. eke, like e'he, contracted to e : trer ' tear '< M.H.G. 

treher (singularized plural of traher ; modern German 
Thrdne is similarly originally plural, M.H.G. trehene, of 
M.H.G. trahen). 

c. e, like e, seems to have been broadened to a before r plus 

consonant in drbas ' pea '< M.H.G. erweiz, (modern German 
Erbse); parallel M.H.G. arweiz would probably have 
resulted in *6rb3s rather than arias (see 1. b above). Note 
ford ' horse ' < M.H.G. pfiirt (pfdrd-). 



JUDEO-GERMAN PHONOLOGY — SAPIR 24! 

d. e is preserved as e (close quality as in French tie') before yg, 
t)k: breygn 'to bring '<M.H.G. brengen (Middle German 
dialectic form of bringen ; cf. also Old Saxon brengean < 
*brangjan)\ <ffy£«< M.H.G. dinken ; zix be'ykn 'to long 
for', cf. M.H.G. binge (alongside of bange) 'Angst, Sorge'. 
e<i also appears in open syllables: he'bn 'to lift '< M.H.G. 
heben, he/en; ket (plural ketn)< M.H.G. kiten 'Kette'; 
tseln< M.H.G. ziln; enikl ' grandson '< M.H.G. enikel, 
ininkel ; e'dl< M.H.G. idel ; he'vn ' yeast '< M.H.G. hive. 

5. M.H.G. e. 

a. This sound, while losing its length, retained its quality as 

close?; hen 'to stand '< M.H.G. stin; gen < M.H.G. gen ; 
ine kM.H.G. sne ; vetig 'pain'< M.H.G. witac 'leiblicher 
Schmerz, Leiden, Krankheit ' (literally ' woe-day ') ; e'dr 
'rather, sooner' (with inorganic -d-)< M.H.G. er. Before 
final r, e is followed by glide » : zhr ' very '< M.H.G. sere. 

b. It becomes broadened to open e before r in : mer ' more '< 

M.H.G. mer; erlt< M.H.G. Irst. 

6. M.H.G. a (/-umlaut of a). 

a. This sound fell in completely with e. Examples of e< M.H.G. 

a are: lver< M.H.G. sware ; ver< M.H.G. ware (1st and 
3rd person preterite subjunctive of sin) ; het< M.H.G. hate 
' hatte ' ; gire'innis ' capable person, wohlgeratene Person ' 
(ret-< M.H.G. rat-, cf. gerate ' Rat, Uberlegung '). 

b. a has become i in : gix 'quick '< M.H.G. gahe (gex, which 

would be normally expected, is also found). 

7. M.H.G. i. 

a. As in modern German, M.H.G. / has normally remained : 

zix < M.H.G. sich ; gifinn < M.H.G. (ge)finden ; iz < 
M.H.G. ist; 6lind<M.H.G. Mint (blind-); ./»< M.H.G ■. 
visch. 

b. In bdrm 'pear'<M.H.G. &V(genitive birn) and karl 'cherry' 

< M.H.G. kirse, this sound seems, like e\ to have become 
a before r plus consonant (see 3 c). Is a in these words 
VOL. VI. R 



242 THE JEWISH QUARTERLY REVIEW 

due to parallel dialectic e' (cf. Anglo-Saxon feru : O.H.G. 
bira ; Lat. cerasum : O.H.G. kirsd) ? 
c. em ' him ' < M.H.G. im(e) is probably developed from parallel 
Middle German dialectic em[e). 

8. M.H.G. f. 

a. As in modern German, M.H.G. i regularly became diph- 

thongized to ai: taix 'lake, creek '< M.H.G. tick 'pond'; 
zait ' side '< M.H.G. site; tsait< M.H.G. zit ; drai< 
M.K.G.drf; vdih '(short) while '< M.H.G. wile ; main< 
M.H.G. min. 

b. In git ' gives ', gist ' givest ', i is shortened from i (M.H.G. 

git, gist), rather than directly derived from i of gibet, gibest. 

9. M.H.G. 0. 

a. In closed syllables remained: dort< M.H.G. dort; oks< 

M.H.G. ohse; fol< M.H.G. vol (voll-); mdrgn<U.H.G. 
morgen; ort<M.H.G. ort. 

b. It has become u vafun < von. u oizun ' Sohn ', zun ' Sonne ', 

and Mmn 'kommen' is not derived from original 0, but 
goes back to u (see 1 1 a). 

c. In originally open syllables became lengthened, as in 

modern German, to 0, which then, falling in with original 
long 0, developed to e (see 10a) : eb 'ob'<M.H.G. obe ; 
e'vit ' stove '< M.H.G. oven; (bn< M.H.G. oben ; fe'gl< 
M.H.G. vogel; hezn ' trousers '< M.H.G. hosen. In words 
where of close and of open syllables varied paradigma- 
tically, older 0: e (0 : 0) was levelled out to e (0) : kef< 
M.H.G. hof(hoves). It is not clear why we have e, instead 
of 0, in ke'kr ' hunchback '< M.H.G. hocker (perhaps < 
parallel *hoker with ungeminated k ; cf. parallel hoger). 

10. M.H.G. 6. 

a. This sound regularly became e, probably through transitional 
stages oi>6i>ei. Examples are: gres< M.H.G. grtiz,; 
sen ' already '< M.H.G. schdn{e) ; hex< M.H.G. Mch; bret 
< M.H.G. brSt; ret< M.H.G. r6t ; azi<M.H.G.als^; lez 



JUDEO-GERMAN PHONOLOGY — SAPIR 243 

<M.H.G. 16s; re < M.H.G. ro. Before final r glide 9 
intervenes : (sr 'ear'<M.H.G. ore. 

u. M.H.G. u. 

a. It normally remains as u: un ' and '< M.H.G. unde; stub< 

M.H.G. stube ; tsugg< M.H.G. zunge ; Inur ' daughter-in- 
law^ M.H.G. snur; zun ' son '< M.H.G. sun (modern 
German Sohn is specifically Middle German, M.H.G. son) ; 
zun ' sun '< M.H.G. sunne (modern German Sonne is 
specifically Middle German) ; kiimn < M.H.G. kumen 
(variant of komen, probably extended by analogy from 
singular of present indicative kume, kumest, kumet) ; zdmr 
' summer' < M.H.G. sumer; trukn 'dry'<M.H.G. trucken; 
riikn ' to shove '< M.H.G. rucken (parallel to riicken) ; hunt 

< M.H.G. hunt{hund-). 

b. M.H.G. u seems to have become i, probably via », in um 

zist 'um sonst'< M.H.G. umbe sust. 

c. Before r plus consonant u is broadened to in v6rtsl ' root ' 

< M.H.G. wurzel (cf. Middle German worz for wurz 
'plant, root'), also before final r in nor 'only' < M.H.G. 
nur. 

12. M.H.G. a. 

a. Diphthongization has taken place, as in modern German, but 

to oi (probably through ui, which seems to be found in 
some Judeo-German dialects), not au. Examples are : 
oif< M.H.G. 6f; koiz< M.H.G. Ms; moiz< M.H.G. mils; 
moil <M.H.G. mill 'Maul'; kloiz 'Talmudic school'< 
M.H.G. kldse ' abgeschlossene Wohnung ' ; hoit< M.H.G. 
Mt ; toiznd< M.H.G. tdsent {tusend-) ; boi»n< M.H.G. 
bdwen. Glide 3 appears after oi before final r ; zohr< 
M.H.G. sHr ; poter ' peasant ' < M.H.G. bur. 

b. Before x plus consonant it is shortened to u in mir duxt ' it 

seems to me '< M.H.G. dilhte (preterite oidunken, diinken) ; 
cf. d>a before x plus consonant (see 2 b above). 

c. M.H.G. (2 has become a in farzdmn 'to miss, neglect '< 

M.H.G. versdmen. No reason that is apparent can be 
given for this singular change. 

R 2 



244 THE JEWISH QUARTERLY REVIEW 

d. M.H.G. it has apparently become at in : klaibn ' to gather ' 
<M.H.G. kluben 'pfliicken, stiickweise ablesen, auflesen ' 
(> Modern German klauben). This is hard to understand 
phonologically. With its strong participle giklibn, it looks 
remarkably as though developed from M.H.G. kliben, past 
participle gekliben 'anhangen, Wurzel fassen und gedeihen', 
though there are semantic difficulties here. Perhaps 
*kloibn<klt1ben and klaibn <kliben became confused in 

o o 

one form. 

13. M.H.G. u. 

a. Ordinarily U was unrounded and thus fell in completely with 

original /: mil< M.H.G. mill 'mill'; ibr< M.H.G. iiber ; 
tin 'sons'< M.H.G. siine; &«/£•< M.H.G. kiinh (kiinig-) ; 
unmighx<M..Yi.G. unmilgdich; hintl, diminutive of hunt 
' dog '< M.H.G. hunt (hund-) ; lign ' lie ' (subst.) < M.H.G. 
lugen, liigene. 

b. It became velarized to u in : fuh ' fulness' < M.H.G. viille ; 

kuln 'to kiss '< M.H.G. kiissen (perhaps by analogy of ku$ 
' kiss ') ; fufisn ' fifteen ' < M.H.G. vunfze'hen ; fuftsig ' fifty ' 
< M.H.G. funfzic {fiittfzig-) (cf. M.H.G. vunf, vumf as 
parallel forms of vtinf, vumf). 

c. Before final r and before rr it became broadened to a (cf. 

3 c', 4 c, 7 b) in : far< M.H.G. viir ; dar ' thin ' < M.H.G. 
dilrre. 

14. M.H.G. iu. 

This sound (pronounced u) represents older diphthongal iu and 
U as /-umlaut of u. In Judeo-German it became unrounded 
to t, which, falling in with original ;, became diphthongized 
to ai. Examples are: kdizr ' houses '< M.H.G. hiuser ; 
tnaiz ' mice ' <M.H.G. miuse ; nai l new '< M.H.G. niuwe ; 
aix<M.H.G.iuck; dhr< M.H.G. iuwer ; Wa/<M.EG. 
biuchel, diminutive of bUch 'Bauch'; lail< M.H.G. liute. 

15. M.H.G. 0. 

a. As with other umlaut vowels, was unrounded to e, thus 
falling together with original e\ Examples of e < M.H.G. 



JUDEO-GERMAN PHONOLOGY — SAPIR 245 

are : rekl< M.H.G. rSckel, diminutive of roc {rock-) ' coat ' ; 
he'rnr< M.H.G. homer, plural of horn; gike'xts 'something 
cooked ' < *gekochtes (such forms seem to be based on sub- 
stantivized neuter past participles in -tes, e.g. gekochtes, 
influenced by neuter collectives in ge- . . -e with umlaut, 
e.g. gehorne). 
b. In el ' oil' < M.H.G. Si, Sle it seems that M.H.G. resulted 
in e instead of expected e. However, el may go back to 
parallel M.H.G. ol, ole according to 9 c. 

16. M.H.G. a (0). 

a. This sound became unrounded to e, thus falling together 

with original e, whence Judeo-German e: sen< M.H.G. 
schosne; fle'tsn< M.H.G. vlcetzen (causative of vliezpn); le'zn 
' to take in money ' ( < ' to release value '?)< M.H.G Ixsen ; 
tre'stn< M.H.G trmsten. 

b. It is broadened to e before r(ci. 5 b) : he'rn< M.H.G. hceren. 

c. In certain comparatives e< M.H.G. a developed to e instead 

of e without apparent phonetic reason : gre'sr ' larger ' < 
M.H.G. grosser; se'nr <M.H.G. schcener ; he'xr ' higher '< 
M.H.G. hoeher. In he'xr open e may be phonetically 
explained as due to shortening of 6 to before x (which 
had been introduced into comparative from positive h6ch ; 
*harher, instead of hasher, >*hbcher> he'xr); cf. ia (last 
sentence), 2 b, 12 b. The combined influence of he'xr and 
such e-comparatives as le'ygr ' longer ' (in which e regularly 
developed from e, /-umlaut of a) may have served to 
establish a category of ^-comparatives, which analogically 
displaced the phonetically justified comparatives *gre'sr, 
*senr. The change thus effected is functionally useful, 
inasmuch as a difference of form is established between the 
comparative and the inflected positive (nominative mascu- 
line singular) : a gre'sr man ' ein grosser Mann ', but er iz 
gre'sr ' er ist grosser ' (modern German schbner corresponds 
to both lenr and se'nr). That this change of a? to e is not 
phonetic, but analogic in character, is further indicated by 
the parallel kle'nr <kleiner (but positive klen<klein). 



24 6 THE JEWISH QUARTERLY REVIEW 

17. M.H.G. UO. 

a. This diphthong was monophthongized to u and, there being 

no quantitative differences in Judeo-German accented vowels, 
fell together with original u : sux< M.H.G. schuoh ; mutr< 
M.H.G. muoter; bux< M.H.G. buoch; stul< M.H.G. stuol ; 
ku<M.H.G.kuo; brudr < M.H.G. bruoder; tsu< M.H.G. 
zuo ; fus< M.H.G. vuoz,. 

b. In ton 'to do '< M.H.G. tuon it appears as 0. This is 

probably due to the analogy of the participle gitSn < M.H.G. 
getdn (the ablaut uo-d, Judeo-German u-o, is isolated and 
therefore easily levelled out). 

18. M.H.G. tie. 

This diphthong, which serves as /-umlaut of uo, became un- 
rounded to ie and, falling together with original ie, became 
monophthongized to i (it is also possible that tie first became 
monophthongized to it and then unrounded to /'): grin< 
M.H.G. griiene ; ki< M.H.G. kiieje ' cows ' ; mid< M.H.G. 
mtiede ; kil< M.H.G. kiiele, kttel ; bixl< M.H.G. btiechel, 
diminutive of buoch. 

19. MH.G. ie. 

a. As uo, when monophthongized, fell together with u, so ie, 

after being monophthongized, fell together with original 
*: lixt< M.H.G. lieht ; tif< M.H.G. fief '; flign < M.H.G. 
fliegen; bign< M.H.G. biegen; hir< M.H.G. hier ; fir< 
M.H.G. vier. 

b. It became broadened to e before r plus consonant in e'rgits 

' somewhere '< M.H.G. iergen{t), ne'rgdts ' nowhere '< 
M.H.G. niergen(t). Contrast i-< M.H.G. ie- in imr< 
M.H.G. iemer and itst(r) ' now '< M.H.G. iezent. 

c. ze ' they '< M.H.G. sie (but zi ' she '< M.H.G. sie) is perhaps 

best explained as secondarily lengthened from M.H.G. se, 
proclitic form of sie. 

20. M.H.G. «'. 

a. This was not preserved as diphthong ai, as in modern German, 
but was monophthongized to* (probably via e) : hesn< 



JUDEO-GERMAN PHONOLOGY — SAPIR 247 

M.H.G. heizpn; ?«< M.H.G. ein ; bret<M.H.G. breit ; 
A*w< M.H.G. heim (note also Judeo-German adverb ahem 
' nach Hause ') ; e ' egg ' < M.H.G. ei ; edm ' son-in-law ' 
< M.H.G. eidem ; kkn< M.H.G. kleine, klein ; menn< 
M.H.G. meinen ; Mln < M.H.G. heilen ; ren < M.H.G. rein. 
b. It appears as e in imr ' pail '< M.H.G. eimer, eimber (cf. 
M.H.G. parallel form ember), e of kle'nr ' smaller ' is best 
explained as due to analogy (see explanation of gre'sr and 
se'nr in 16 c). 

2r. M.H.G. ou. 
This diphthong early became monophthongized to (cf. ei>e, 
see 20 a) and was further developed, together with original 0, 
to e (probably via oi>di>ei): ban < M.H.G. bourn ; eg< 
M.H.G. ouge ; ke'fn< M.H.G. koufen ; steb< M.H.G. stoup 
(stoub-); rex< M.H.G. rouch. 

22. M.H.G. eu, ou. 

This diphthong also became Judeo-German e (perhaps via >e ; 
or via bi>ei, cf. M.H.G. vrbide as variant of vroude) : /red 
' joy ', Fre'dd ' Joy ' (girl's name) < M.H.G. vroude, vreude ; 
he 'hay '< M.H.G. hbuwe, hbu : leb ' lion '< M.H.G. louwe 
(parallel to /ewe, which would have developed to *leb). 

We thus see that the original rich vocalism of Middle 
High German has been greatly simplified in Judeo-German 
by unrounding rounded vowels (u>i,u>i>i, iie>ie>i, b> 
e,S>e>e), by obliterating quantitative vocalic differences 
{l<ie and i both give i; u<uo and u both give u ; these 
secondary I and u are of course to be carefully kept apart 
from original M.H.G. i and u, which did not fall together 
with them because they had already become diphthongized 
when ie became 1 and no became u), and by monophthong- 
izing of diphthongs (ei>e>e, ou>d>oi>bi>ei>e). In 
particular e is, at least in the Lithuanian dialect, the reflex 
of no less than eight distinct vowels and diphthongs : e (in 



248 THE JEWISH QUARTERLY REVIEW 

open syllables), e, ei, en (ou), ce, d, ou, and o (in open syllables). 
Similarly, i goes back to i, ii, ie, and He ; to (in closed 
syllables), a, and a (in open syllables). Many words that in 
Middle High German are phonetically distinct have, in 
Judeo-German, fallen together owing to the operation of 
the phonetic laws we have sketched. Thus, bret corre- 
sponds to modern German breit and Brot; sen to scion 
and scion ; sten to steien and Stein ; nox to nod and nach ; 
e'gn to eigen and Augen. 

Unaccented M.H.G. e has generally dropped in absolute 
finality; examples of this have incidentally occurred in 
the discussion of the accented vowels. Where unaccented 
-e is preserved (as ' Murmelvokal ' -3), it is generally due 
to a functional, not a phonetic, reason (e.g.giit? lait' good 
people ' and a gxlta toxtr 'a good daughter', in which -9, as 
adjectival ending, indicates respectively plurality and 
feminine gender. Unaccented M.H.G. e unites with follow- 
ing tautosyllabic /, m, n, and r to form sonantic (syllabic) 
/, m, it, and r. In unaccented syllables and when after 
vowels or when followed by one or more stop or spirant 
consonants M.H.G. e appears as 3 (e.g. ergats< M.H.G. 
iergent; di3r< M.H.G. iuwer 'your'). Unaccented M.H.G. 
e sometimes disappears in other than final position. Thus, 
regularly in participial -et after all consonants, including 
d and /, -det and -tet contracting to -t (e.g. givdrt< M.H.G. 
gewartet ; giiit< M.H.G. geiiktet; gire'K M.H.G. geredet) ; 
similarly, -est of second person singular and -et of third 
person singular and second person plural present indicative 
(and imperative) regularly become -st and -/, -det and -tet 
contracting to -/ (e.g. du vdrtst< M.H.G. du war test ; er ret, 
gifint< M.H.G. e'r redet,gefindet: ir hit< M.H.G. ir iiietet; 
second person plural imperative ret< M.H.G. redet). Such 



JUDEO-GERMAN PHONOLOGY — SAP1R 249 

syncopated forms go back in part to M.H.G. originals (e.g. 
M.H.G. vint ' finds' alongside of vindet; getraht alongside 
of getrahtet). 

Other unaccented vowels than are also found, though 
rather less frequently than in modern German. They occur 
chiefly in secondarily accented syllables. Examples of 
suffixed elements with vowel not dulled to 3 are : -ik (e. g. 
kinik< M.H.G. kiinic) ; -is< M.H.G. -isch (e.g. miis ' ugly') ; 
-nis< M.H.G. -nisse (e.g. gfrdtnnis); -uyg (e.g. menuyg< 
M.H.G. meinunge) ; -ket< M.H.G. -keit (e.g. gutsket ' good- 
ness ', grdsket ' greatness '). Diminutive -tin appears in 
Judeo-German as secondarily accented -le, preceding 
M.H.G. -e- being developed to -a- (e.g. kindale< M.H.G. 
kindetin ; these diminutives in -ale imply a loving or 
caressing attitude, whereas forms in -sl % -I are simply diminu- 
tive). M.H.G. -licit regularly appears as -hx (e.g.fre'l3x< 
M.H.G. vrcelich). Full vowels of unaccented syllables which 
have no definite significance as word-forming elements tend 
more frequently than in modern German to be dulled to 3 (e. g. 
drb$s 'pea '< M.H.G. enveiz„ anveh; drbrt ' work' < M.H.G. 
arbeit). This is true even in cases where the unaccented 
vowel is the stem vowel of the second member of a com- 
pound, provided the analysis of the compound is not felt 
as obvious (e.g. bSrvss ' barefoot '< M.H.G. barvuoz, : kimpat 
' confinement after childbirth ' < M.H.G. kintbitte). An 
example of extreme reduction, in which not only an un- 
accented diphthong but also the consonant following it is 
lost, is knobl ' garlic ' < M.H.G. knobelouch. 

M.H.G. e standing in a syllable immediately preceding 
the accent seems regularly to develop to a: ba-< M.H.G. 
be- (e.g. baklogn< M.H.G. beklagen) ; far- < M.H.G. ver- 
(e. g. farbrdnij. < M.H.G. verbr'ennen) ; ar- < M.H.G. her (e.g. 



250 THE JEWISH QUARTERLY REVIEW 

ar- in local adverbs — aruntr, arlbr, arois, and others) ; ant- 
<M.H.G. en(f)- (e.g. antke"gn<M.li.G. engegcn(e)\ antlefn 
<M.H.G. entloufen); «-< M.H.G. en- in adverbs (e.g. avik 
<M.H.G. en-wec 'away'; alter <M.H.G. en-her 'hither'; 
akin ' thither ' ; ahdm ' towards home '). Accented ent-, 
however, remains : tntfarn ' to answer '<M.H.G. entwiirten 
(parallel to antwurten). Unaccented M.H.G. bt also de- 
veloped to ba (e.g. ba mir 'bei mir'< M.H.G. bt mir) ; 
unaccented M.H.G. uf became of (e.g. af a bdyk 'on a 
bench ') ; M.H.G. unaccented vor developed to far, thus 
falling together with M.H.G. ver- and viir (see 13 c ; far< 
viir very likely also developed in unaccented position) (e.g. 
farbdi ' vorbei ' ; far jorn ' years ago '< M.H.G. vor jdren ; 
far tog ' before daybreak ' ; fartsditns ' long ago '). M.H.G. 
ein as article, which always stands in proclitic position, has 
become a (before consonants), an (before vowels) ; as 
numeral ' one ', however, it develops to en (see 20 a). 
M.H.G. zer-, ze- appears as tsu- (e.g. tsurisn< M.H.G. 
zerrissen) ; this correspondence, however, is undoubtedly 
not purely phonetic in character, as parallel M.H.G. zur-, 
zu- is found in Middle German dialects. M.H.G. verb 
prefix er- appears in Judeo-German as dr-\ cf. parallel 
M.H.G. der-. M.H.G. ge- appears as gi- with short open 
i (e.g. gimdxt< M.H.G. gemacht: gzziint< M.H.G. gesuni) ; 
it is barely possible that this gi- goes back to O.H.G. gi-. 
Proclitic man, in its indefinite sense, becomes reduced to 
mn (e.g. mn ment ' man meint '). 

The whole Judeo-German vowel scheme thus reduces 
itself to six full vowels : a, o, i, u, e, e ; a ' Murmelvokal ' 3 
(also i) ; and two diphthongs : at, oi. 



judeo-german phonology — sapir 25 1 

Consonants. 

The Middle High German consonants have undergone 
less sweeping changes than the vowels. The most im- 
portant innovation has already been mentioned : the 
generalization of a paradigmatic final stem sonant, the 
otherwise constant interchange in German dialects between 
final surd and medial sonant being thus obliterated in 
Judeo-German. The comparatively few consonant changes 
that it has suffered will be noted under the various con- 
sonants. The chief points of general application are these : — 
The stops exist in two strictly differentiated series as surds 
and sonants ; there is no amalgamation of the two into one 
group of 'voiceless mediae', as in many Middle German 
dialects, nor has the sonant lost any of its resonant quality ; 
the surds and sonants are as clearly set against each other 
as in English. The distinction that obtains in modern 
German between guttural x (after back vowels) and palatal 
x? (after palatal vowels, r, and /) is absent in Judeo-German ; 
the guttural x (as in German Bach) is used in all positions 
{thus, to German schlecht corresponds Judeo-German slext 
with x as in Dutch slecht and as in Swiss dialects). The 
pronunciation of r differs in different parts of the Judeo- 
German area. While the trilled tongue-tip r, which may 
be due to Slavic influence, is found in Southern Russia, 
the uvular r (r grassey^) prevails in the Lithuanian dialect ; 
it is pronounced with considerable vigour, but is not 
markedly trilled, hence is probably better defined as voiced 
velar spirant (y). This uvular r and the frequency of 
guttural x serve to give Judeo-German a characteristic 
guttural acoustic effect. In our consideration of the con- 
sonants we begin with the semivowels. 



252 THE JEWISH QUARTERLY REVIEW 

1. M.H.G. j. 

a. It is generally preserved as j (y of English young) : jujg< 

M.H.G. junc (Jung-); jor < M.H.G. jar ; /<%-«<M.H.G. 
jagen. 

b. Where it served as glide consonant in M.H.G. between pre- 

ceding palatal vowel and following unaccented e (as in 
kileje, miieje, sajeri) it has dropped in Judeo-German 
(together with final -e) : ki ' cows '< M.H.G. ktieje. 

c. It has dropped initially before Judeo-German i (M.H.G. it) : 

iygl< M.H.G. jiingd(in) ; id ' Jew '< M.H.G. jiide (parallel 
tojude). It is interesting to note that i-<ji- requires a as 
preceding article : aid 1 ein Jude ' (not an id). 

2. M.H.G. w. 

a. This sound, where preserved, became dento-labial v: vald< 

M.H.G. ivalt (wald-); tsve< M.H.G. zwei ; sver 'heavy' 
< M.H.G. sincere; svax< M.H.G. swach ; kve'ln 'to well 
up, swell (with joy) ' < M.H.G. qui'llen (i. e. kwellen) ; vort< 
M.H.G. wort. 

b. It appears as /after / in: e'ntfirn 'to answer '< M.H.G. 

antwurien. 

c. After / and r it became stopped to b, as in modern German : 

arfos 'pea'<M.H.G.<i>zw5; farb 'colour'< M.H.G. varwe. 

d. Between vowels (but not after ^-vowels) w seems, as in 

Swabian dialects (cf. also German hiebetK M.H.G. hieiven), 
to have become b: leb ' lion '< M.H.G. le'we, louwe ; (big 
< M.H.G. ewic (ewig-) ; iybr 'ginger '< M.H.G. ingeiver 
(cf. M.H.G. variants ingeber, imbe'r). 

e. It is syncopated between «-vowel and following vowel : 

b6i»n 'to build* < M.H.G. biiwen ; ahr< M.H.G. iuwer. 

3. M.H.G. /. 

a. Normally it remains : land< M.H.G. lant {land-) ; layg< 

MH.G. lane (lang-) ; laixt < M.H.G. licht ; als ' all ' < 
M.H.G. allez,; fdlti< M.H.G. vallen; gold<M..H.Q. golt 
{gold-). 

b. It has been syncopated before an accented syllable in : az 

'that, when'<M.H.G. als; aze ' so '< M.H.G. also. 



JUDEO-GERMAN PHONOLOGY — SAPIR 253 

4. M.H.G. r. 

a. As we have seen, it became uvular in pronunciation : ret< 

M.H.G. rot; re'gn< M.H.G. r'e'gen; rex< M.H.G. vouch; ber 
' bear '< M.H.G. ber; gor< M.H.G. gar; he'rn< M.H.G. 
kceren. 

b. In mdtrn 'to torment '< M.H.G. martem /-has been synco- 

pated by dissimilation from r of -em. In forms of ve'rn < 
M.H.G. werden r is syncopated before final -t and -st: du 
vest<du tvirst, er vet<er wirt (e of vest and vet is analo- 
gical), ir vet<ir werdet. 

5. M.H.G. n. 

a. This sound normally remains, also in infinitive ending -en : 

ftai< M.H.G. niuwe ; nit 'not '< M.H.G. met (variant of 
nieht, niht); noz< M.H.G. nase ; nas< M.H.G. naz,; ken 
< M.H.G. kan and kenne; he'lfn< M.H.G. he'lfen ; zint< 
M.H.G. sint' since'. 

b. In «'« as indefinite article n has remained only before 

vowels, otherwise it is syncopated : a man ' ein Mann ' 
but an Sks 'ein Ochs' (cf. English a, an). Wrong 
division has produced, e.g., nam ' nurse' (M.H.G. ein' amme 
> an am > a nam) ; not ' awl ' (M.H.G. ein' die > an oI> 
a not), n has been syncopated also in : le'btdik ' alive ' < 
M.H.G. lebendic ; fufisn < M.H.G. viinfze'ken, fuftsig< 
M.H.G. vunfzic. It is barely possible that fuftsn and 
fuftsig have been remodelled, by analogy oifinf ' five ', from 
etymologically justified */ux- < *fux- < *fuyx- < Indoger- 
manic *J>r/kw-(cf. Swabian/w/jz<F 'fifteen'; seeW. Streitberg, 

o 

Urgermaniscke Grammatik, 1900, p. 111). 

c. In M.H.G. ne'ben 'near' n- has become dissimilated to /: 

Ubn. 

o 

d. It is assimilated before/ to m in kimpst< M.H.G. kintbette ; 

vaimprhx diminutive plural < M.H.G. wtn-ber. 

6. M.H.G. m. 

This consonant seems to have remained in all cases : medl 
' girl' < M.H.G. meidel; man< M.H.G. man (mann-); mos 



254 THE JEWISH QUARTERLY REVIEW 

< M.H.G. mdz,; mir< M.H.G. mir; Mmn<M.H.G.kumen; 
hem <M.H.G. heim. It is particularly noteworthy that un- 
accented -em has not been weakened to -n as in modern 

O 

German : be'zm ' switch used in rubbing down in sweat-bath ' 

< M.H.G. beseme (cf. German Besen) • f6dm < M.H.G. 
vadem (cf. German Faden); bedm'Xoit, attic '< M.H.G. 
bodem (cf. German Bodeii). In zamd ' sand ' m, as we have 
seen, is more archaic than n (M.H.G. sant, sand-). 

7. M.H.G. y (written n). 

This sound, which occurs only before g and k, has been pre- 
served in all cases: gtgdt;gn< M.H.G. gegangen ; juyg< 
M.H.G. junc (/ung-) ; iggl< M.H.G. jungel(fn) ; ddykn< 
M.H.G. danken ; deykn< M.H.G. denken. 

8. M.H.G. z>,/(Urgermanisch/) and -ff-, -/- (Urgermanisch / ). 

a. As in other modern German dialects, these two etymo- 

logically distinct sounds fell together in Judeo-German, 
except for intervocalic -v- (see b) : f6tr<, M.H.G. voter; 
fil< M.H.G. vil; fedr< M.H.G. veder ; fe'tr 'uncle' < 
M.H.G. zvter ' Vatersbruder ' ; far-< M.H.G. ver- ; U6fn< 
M.H.G. sM/en; tif< M.H.G. tie/; Mfn <M.M.G. helfen; 
sar/KM.'R.G. scharf ; dorf '< M.H.G. dorf '; dfn< M.H.G. 
of en; hef< M.H.G. hof ; vo/f<MM.G. wolf. 

b. Medially before vowels M.H.G. v appears as v (voiced 

dento-labial identical with v< M.H.G. w, see 2 a): e'vn 
' stove '< M.H.G. oven; taivl ' devil' < M.H.G. tiuvel ; 
hevn ' yeast '< M.H.G. hive; bSrvis ' barefoot '< M.H.G. 
barvuoz,. Mbr ' oats ' goes back to M. H.G. ftaber, not haver 
(see ro a). 

9. M.H.G. pf{ph). 

a. Initially ^yhas become simplified to/: funt< M.H.G. pfunt ; 

ferd< M.H.G. pfert(pferd-); ixf/eg'I was wont to' (present 
in form, but imperfect in meaning) < M.H.G. pflege ; fan 
' pan '< M.H.G. pfanne {fatnkuxn ' Pfannkuchen ' is pro- 
bably made over by analogy of fain <fin). 

b. Medially and finally it lost its spirantal element and became 



JUDEO-GERMAN PHONOLOGY — SAPIR 255 

/: &?/< M.H.G. kopj ; £/<?/«< M.H.G. klopfen ; epl< 
M.H.G. epfel, apfel; liupn 'to shove' (er itupt tinir'he 
eggs on ')< M.H.G. stupfen 'stechend stossen, antreiben'. 

10. M.H.G. b. 

a. Normally b is preserved (as voiced lenis); it occurs also 

finally (< M.H.G. -/), probably by analogy of medial -b-: 
bret <M.H.G. breit ; bret< M.H.G. brSt ; bai,ba<M.H.G. 
bi; barg<M.U.G. bire {berg-); k'bn< M.H.G. leben ; le'bn 

< M.H.G. neben ; h6br ' oats ' < M.H.G. haber (of which 
to/w>modern German Hafer is variant); tsibl 'onion' 

< M.H.G. zibolle (variant of swibolle, zwibei) ; stub < M. H.G. 
stube ; $teb< M.H.G. siotip (stoub-). 

b. In 6vnt ' evening '< M.H.G. dbent, M.H.G. b has become 

spirantized to v ; also in hdrwn, see 18 c. For M.H.G. 
medial bilabial spirant 5, from older -b-, in Middle German 
dialects see V. Michels, Mittelhochdeittsches Elementarbuch, 
1900, § 159. 

c. M.H.G. -mb- has, as in modern German, become assimilated 

to -mm->-m- : kam < M.H.G. kamp (kamb-) ; urn < M.H.G. 
umbe; lam, diminutive //male < M.H.G. lamp {lamb-), 
diminutive lembelin. 

d. In a number of words M.H.G. b appears as p. This is 

intelligible where final -b developed to -/ and was not 
levelled out by analogy of medial -b- : zip ' sieve ' < M.H.G. 
sip {sib-); op< M.H.G. abe, ab {arSp ' herab ' ; as verb 
prefix before participial gi-, op- appears as 6-: bgiton 
' abgetan '). Less easily explained are certain examples 
of initial and medial /: potor ' peasant '< M.H.G. bur; 
putr ' butter ' < M.H.G. buter ; gdpl ' fork ' < M.H.G. gabel ; 
kkpn 'to be stuck to '< M.H.G. kleben; vaimpr-{hx)< 
M.H.G. winber. In estimating these and similar develop- 
ments {Kd, k<g) it must be remembered that Judeo- 
German knows no ' voiceless lenis ' stops, but only fully 
voiced lenis stops (corresponding to Upper German voice- 
less lenis) and unaspirated voiceless fortis stops (corre- 
sponding to Upper German voiceless fortis). 



256 THE JEWISH QUARTERLY REVIEW 

11. M.H.G. p. 

This sound regularly remains : parle'n <M.H.G. persdn (for 
a<e see 3c); ^7«< M.H.G. spilen; spruyg <M.H.G. 
sprunc {sprung-). 

12. M.H.G. s, -ss- and -?-, -5?-. 

a. Initial and medial j (except before voiceless consonants) 

became voiced to 2 (this includes also final -s when alter- 
nating with medial -s-) ; zun < M.H.G. sun 'son' and 
sunne 'sun'; zogn< M.H.G. sagen ; ze'gr 'clock '< M.H.G. 
seiger ; ziixn < M.H.G. suochen; az < M.H.G. ah, alse ; 
aze'< M.H.G. alsd ; u 1 izr < M.H. G. unser ; kez i cheese '< 
M.H.G. kcese; £/<&«< M.H.G. Mdsen ; bios ' breath '< 
M.H.G. bids {bids-) 'Hauch'; groz< M.H.G. gras {gras-). 
Medial ungeminated -z,- has also developed to 2 in : lozn 
'to let '< M.H.G. Idzpn. Judeo-German 2 in muz 'must' 
may be similarly developed from medial ungeminated -5- 
(M.H.G. muoz,: muezpn > muz : miizn with generalized 
vocalism of muoz, and medial -2- of muez/n) or, perhaps 
less likely, from medial -s- of preterite muose (later super- 
seded by analogical muoste). More often, however, -2,- is 
treated like -zjz,- (see b). 

b. Final -?, medial -z& and (generally) ---, and medial -s- before 

voiceless consonants appear in Judeo-German, as in modern 
German, as voiceless s: ois 'out '< M.H.G. uz,; fus< 
M.H.G. vuoz,; vais< M.H.G. un\; dos< M.H.G. daz,; 
<#»<M.H.G. ez&en; bhr< M.H.G. bezzpr; baisn<M.W.G. 
bizftt ; hesn< M.H.G. heizpn ; nest < M.H.G. n'e'st ; um zist 
< M.H.G. umbe sust ; aosKM.H.G. hast. Judeo-German 
m/stn ' to measure ' (with analogic participle gtmSstn) has 
perhaps resulted from confusion of M.H.G. mezgfn 'messen' 
and western 'den Inhalt messen'. M.H.G. ist > Judeo- 
German iz is due to loss of -t and voicing of s because of 
its frequent use as proclitic {iz probably generalized from 
antevocalic use, e.g. iz a mdn< M.H.G. is{i) ein man). 

c. For some not evident reason medial M.H.G. -s- appears as 



JUDEO-GERMAN PHONOLOGY — SAPIR 257 

Judeo-German -s- instead of -z- in : Mr< M.H.G. keiser ; 
nisn 'to sneeze '<M.H.G. niesen. 

o 

d. M.H.G. -ss- seems to have regularly developed to s, i.e. 

modern German sch (it has thus not, as in modern German, 
fallen together with M.H.G. -z,z,-, -z,-) : kus< M.H.G. kus 
(kuss-) ; kusn (with vocalism of kus ) < M.H.G. kiissen ' to 
kiss;' k&n< M.H.G. kiissen 'pillow'; pisn 'to urinate' 
< pissen. 

e. After r both s and z, appear as s; kars< M.H.G. kirse; 

parsen 'beautiful woman '< M.H.G. person ; erst <M.H.G. 
erst; hirs<MM.G. hirz, 

f. Before /, m, n, w, p, and t initial M.H.G. s developed, as in 

modern German, to s: slext< M.H.G. sleht ; smaisn 'to 
beat ' < M.H.G.smfzen ' streichen, schlagen ' ; sne < M.H.G. 
sne*; lver< M.H.G. swcere ; lpet< M.H.G. spate; sten< 
M.H.G. stein. 

13. M.H.G. sch. 
This sound is regularly preserved asjf: sepn< M.H.G. schepfen; 
sainn< M.H.G. schinen; se'tl 'perruque with evenly parted 
hair worn by orthodox Jewish women ' < M.H.G. scheitel 
'crown of the head, parting of the hair'; misn< M.H.G. 
mischen ; idil< M.H.G. judisch. 

74. M.H.G. 2 and -tz-. 
These affricatives are everywhere preserved as ts : tsen < M.H.G. 
z'e'hen; tson< M.H.G. zan ; tsve< M.H.G. zwei; harts < 
M.H.G. herze; kats< M.H.G. katze. 

15. M.H.G. d. 

a. Normally d is preserved (as voiced lenis) ; it occurs also 
finally (< M.H.G. -t), probably by analogy of medial -d- 
dax< M.H.G. dock; dar ' thin '< M.H.G. diirre ; drai< 
M.H.G. dri; moid < M.H.G. maget (maged-) ; bod < 
M.H.G. bat (bad-) ; feld < M.H.G. ve'lt (ve'ld-) ; ferd < 
M.H.G. pfert (p/erd-) ; 6dr 'vein' < M.H.G. dder. 
Examples of nd< M.H.G. nd (including cases of -nt alter- 
nating with -nd-)< O.H.G. nt are : bindn < M.H.G. binden ; 

VOL. VI. S 



258 THE JEWISH QUARTERLY REVIEW 

Sntsindn < M.H.G. anziinden ; vundrn<M.li.G. wundern ; 
blind <M.H.G. Mint {blind-); land<M.H.G. lant {land-) ; 
rund< M.H.G. runt (rund-) ; kind< M.H.G. kint (kind-) ; 
end < M.H.G. ende. For examples of nt < M.H.G. -nt 
(-nd-) see 15 d below, 
b. M.H.G. r</ appears as rin: ve'rn < w'erden (similarly ix ver< 
M.H.G. ich we'rde, glv6rn< M.H.G. geworden ; -rst and -rt 
of this verb develop to -st, -t, see 4 b). This development 
is not strictly normal, but is probably due to frequently 
proclitic character of w'erden owing to its use as auxiliary 
verb ; contrast ferd< M.H.G. pferd-. Quite parallel to 
this is /< M.H.G- Id in : mdnzbil ' man '< M.H.G. mannes 
bilde (e.g. zwei mannes bilde er da gesach 'da sah er zwei 
Manner', Der Wartburgkrieg, herausgegeben von Karl 
Simrock, 1858, p. 65, 1. 4 of no 37), in which bilde has lost 
its accent (-2- of mdnzbil due to voiced surroundings of 
M.H.G. -s-); contrast accented bild ' picture '< M.H.G. 
bilde. In certain cases nd is assimilated to nn>n (cf. m< 
M.H.G. nib, see 10 c): un ' and '< M.H.G. unde ; frdnn 
' in existence, to be found ' < vorhanden ; gifinn ' to find ' 
(simplex finn not in use) < M.H.G. gevinden; glltdnn< 
M.H.G. gestanden; tson< M.H.G. zant (zand-), but also 
zan. In un we can readily explain n<nd as due to lack 
of accent (cf. r<rd and l<ld above) ; in/rdnn andgihdnn 
it seems very likely that original -ndn regularly developed 
to -nn, internasal -d- becoming completely assimilated (in 
such forms as bindn, gibAndn it is clear that -ndn was 
restored by analogy of forms like ix bind, er bint; note 
that in frdnn, whose connexion with M.H.G. hant (hand-) 
was lost, and giitdnn, with its infinitive and present hen, 
ix Ue, no disturbance by analogical levelling could take 
place). As iox gifinn (also gifunn, ix giffn) and tson (also 
plural tse'nr; diminutive tse'ndl has not original -nd- but 
intrusive -d-, see 15 c below), I would suggest that M.H.G. 
nd of zand- and vinden (which goes back to O.H.G. nd — 
zand, findan — < Urgermanisch np — cf. Gothic tunpus, 



JUDEO-GERMAN PHONOLOGY — SAPIR 259 

finpan) was, at least in some dialects, phonetically distinct 
from M.H.G. nd < O.H.G. nt (thus, O.H.G. findan > 
M.H.G. vinden >]udeo-Germa.n -finn; O.H.G. bintan> 
M.H.G. binden >]\ideo-Germa.a Mndn); in dndr< M.H.G. 
ander< O.H.G. ander, -nd- may have been protected from 
becoming -n- because of following -r (cf. M.H.G. winter < 
O.H.G. wintar as contrasted with winden<winlan). 

c. Between n as stem ending and -/ (-/-) as diminutive ending 

d develops as glide consonant (cf. Gothic timrjan : O.H.G. 
zimbarSn) : be'ndl ' little bone ' < M.H.G. beinel ; fe'ndl 
' little pan ', diminutive of fan < M.H.G. pfanne ; Mndl 
'little cock' < M.H.G. henel; Mndl ' little hen '< M.H.G. 
hiienel ; diminutive plural of nouns in -n- is -ndhx (e.g. 
bendhx ' little bones, fruit pits '). 

d. In certain cases, as we have seen in 15 a, M.H.G. -nt (-nd-) 

and -// {-Id-) developed to -nd and -Id, as would be nor- 
mally expected for Judeo-German. In a large number 
of examples, however, -/ is generalized, replacing -d- also 
medially : giziint (also, e.g., in a gizuntr ' ein gesunder ')< 
M.H.G. gesunt (gesund-); hunt (also, e.g., diminutive hintl; 
contrast Mndl as diminutive of hun ' hen ')< M.H.G. hunt 
(hund-) ; hant (also, e.g., diminutive he'ntl; contrast he'ndl 
as diminutive of hon ' cock ') < M.H.G. hant {hand-) ; vint 
(also, e.g., diminutive vintl)< M.H.G. wint {wind-) ; funt 
< M.H.G. pfunt (pfund-);fraint kMM.G. vriunt {vriund-); 
faint < M.H.G. viant {viand-) ; bunt (also, e.g., diminutive 
bintl) < M.H.G. bunt (bund-); bwtt< M.H.G. dbent (dbend-) ; 
gidu/t<M.H.G. gedu/t, gedulde (but also gedultec); gelt< 
M.H.G. gilt (geld- but also g'e'lt-). I can suggest no definite 
rule for such differences of treatment as blind < M.H.G. 
Mint (blind-)<OM.G. Mint (Mint-) and vint< M.H.G. wint 
(wind-) < O.H.G. wint (wint-). Possibly -nd forms are 
generalized in words where medial -nd- occurs often (e. g. 
blind because supported by inflected Minds and Mindr), but 
-nt forms where medial -nd- either occurs infrequently (thus, 
M.H.G. biindel would not be of frequent enough occurrence 

S a 



260 THE JEWISH QUARTERLY REVIEW 

to influence bunt, hence itself suffers analogical levelling to 
bintl, which can hardly be directly traced to H.G. buniil) 
or has become obsolete in Judeo-German (thus M.H.G. 
hende had to develop, with loss of -e, to Judeo-German 
hend, which could not maintain its -d against singular hant, 
hence itself suffers analogical levelling to hent, which can 
hardly be directly traced to O.H.G. henti) ; bunt and bindn 
appear contradictory, but can be readily explained, as they 
would not be felt to be connected closely enough to 
influence each other. In untr< M..H..G. under<O.H.G. 
untar, -nd- has, as in modern German, again become 
hardened to -nt-, probably because of following -r (cf. 
M.H.G. winter <wintar); similarly hintr< M.H.G. hinder. 
As for Judeo-German gelt as contrasted with/eld, it should 
be noted that O.H.G. has correspondingly gelt but/eld. 
e. Different from these examples of -nt and -It from M.H.G. 
-nd- and -Id- are certain cases of initial t< normal M.H.G. 
d (cf. p<b, iod): taits< M.H.G. diutsck, tiutsch (also 
fartaitsn 'to translate '< M.H.G. diutschen, tiutschen 'auf 
deutsch sagen, erklaren ') ; tuykl< M.H.G. dunkel, tunkel 
(M.H.G. tunkel is normal, hence this example belongs 
rather under M.H.G. t) ; tetl ' date '< M.H.G. datel {tetl 
may be assimilated from *detl; why e instead of expected 
o?). 

1 6. M.H.G. t. 

a. This sound, aside from cases of M.H.G. -/: -d-, has been 

kept in all positions: ton< M.H.G. tuon ; /#«/< M.H.G. 
tumel ' betaubender Schall, Larm '; tel< M.H.G. teil ; k6tr 
' tomcat ' < M.H.G. kater ; vintr< M.H.G. winter ; zint 
' since '< M.H.G. sint ; bet< M.H.G. belte ; ret< M.H.G. 
r6t; gtvdlt<M.H.G.gewalt ; nont ' near '< M.H.G. ndhenl 
(note also Judeo-German comparative ne'ntr). 

b. It is not easy to see why -tl has become -dl in be'rdl, diminu- 

tive of bort ' beard ' < M.H.G. bart. Perhaps original 
*be'rtl was transformed by analogy of diminutives in -ndl 
(see 15 c). 



JUDEO-GERMAN PHONOLOGY — SAPIR 261 

Initial tw-, as in modern German, has developed to tsv-: 
tsvfygn<M.H.G. twingen; tsvdgn 'to wash one's head'< 
M.H.G. twaken, past participle getwagen ; tsvonx< M.H.G. 
twarc (iwarg-) ' Quarkkase' (this word may have been 
directly derived from Slavic, e.g. Polish tvarog, from which 
it was borrowed by M.H.G., in which case Judeo-German 
tsv-<tv-<tw- would have taken place after Judeo-German 
had become isolated from other German dialects ; this, 
however, is rendered very improbable by parallel form 
zwarc in late M.H.G.). 

d. Medial -fte/- has become -p- in: efos ' something '< M.H.G. 

et(e)waz (cf. Latin b < dw in bis, p<tw in postis). — How 
explain rdtvn 'to save'? It is undoubtedly connected 
with M.H.G. and O.H.G. retten < West Germanic *hraddjan 
< Urgermanisch *kradjan, but cannot be directly derived 
from it. Perhaps parallel to *hrai-jan with /-suffix was 
*hract-wan with w-suffix> O.H.G. *{h)ratwan> M.H.G. 
*ratwen, dialectically preserved in Judeo-German as rdtvn. 
In that case -tw->-p- may hold only in normally un- 
accented words. 

e. -st has become -s>-z, because of lack of accent, in: iz< 

M.H.G. ist. Similarly, -rtn has become -rn in : tntfarn 
' to answer '< M.H.G. entwilrten. 

17. M.H.G. h (as spirant), ch. 

As was noted above, no distinction is made in Judeo-German 
between guttural x and palatal x' (as in modern German 
icti), but both are represented by guttural x. This feature 
may be archaic rather than due to levelling. 

a. It is kept in all positions except before s: laixt< M.H.G. 
lihte ; nox < M.H.G. nock; nox < M.H.G. ndch; zix < 
M.H.G. sick; kalx < M.H.G. kalch (parallel to normal 
kalK) < O.H.G. kalch; marx 'marrow' < M.H.G. march 
(parallel to mark); gix ' quick '< M.H.G. gdch ; sux 
'shoe' < M.H.G. schuoch (note analogical plural h'x < 
M.H.G. schuohe) ; hex< M.H.G. hock (note analogical 



262 THE JEWISH QUARTERLY REVIEW 

comparative he'xr: < M.H.G. hasher); bilxr ' more proper ' < 
M.H.G. billich ' gemass, geziemend ' (g of modern German 
billig is secondary in origin). 

b. Before s, as in modern German, it has become k: oks< 

M.H.G. ohse ; vakst ' grows ' < M.H.G. wahset. 

c. Before diminutive -/ nouns ending in / insert x: $p(lxl, 

'plaything', diminutive of $pil< M.H.G. spil ; mailxl, 
diminutive of moil ' mouth '< M.H.G. mtil; Mlxl 'little 
throat, voice ' < M.H.G. kel. I doubt if this -xl is in any 
way connected with modern German diminutive -chen. 

1 8. M.H.G. h (as aspirate). 

a. It is preserved initially : halz< M.H.G. hah; fo«<M.H.G. 

hane ' cock '; hot< M.H.G. hat; ahfn<M..H.G. hin. 

b. Between vowels, as in modern German, it disappears : laisn 

'to lend '< M.H.G. lihen ; nont 'near' < M.H.G. ndhent. 
For M.H.G. She- and -ehe->-e- see 3 b and 4 b of Vowels. 
h has also disappeared mfrdnn 'present ' <vorhanden. 

c. In a few words h is inorganic: hailn 'to hurry '< M.H.G. 

Hen; hdrwn 'to work hard' <arben, areben (Swiss arban, 
Nassau erwi ; see Kluge, Deutsches Etymologisches Worter- 
buch, s.v. Arbeit) with v<b, see 10 b. 

19. M.H.G. g. 

a. Normally g is preserved (as voiced lenis); it occurs also 

finally (< M.H.G. -c), probably by analogy of medial -g- ; 
it has nowhere undergone spirantization to y (as in modern 
German tayi) or / (as in modern German vej9). Examples 
are: gut< M.H.G. guot; £?/' yellow '< M.H.G. ge'l ; zdgn< 
M.H.G. sagen ; ne'gl ' nails '< M.H.G. negele; veg< M.H.G. 
wee (iveg-); karg< M.H.G. karc (karg-). It is preserved 
also after y: sif/gn < M.H.G. singen (contrast modern 
German ziyn) ; juyg < M.H.G. June (Jung-) (contrast 
modern German jug). 

b. In certain words with M.H.G. -c: -g- Judeo-German has 

generalized -k : tsvayk 'tongs, pincers '< M.H.G zwange 
(note retention of w as v in Judeo-German); sok ' juice '< 



JUDEO-GERMAN PHONOLOGY — SAPIR 263 

M.H.G. soc, sog- (parallel to more normal sue, sug-), which, 
however, is more likely borrowed, as indicated by its 
0-vocalism, from Russian sok ' juice ' (Germanic loan-word) 
than directly derived from M.H.G. In nouns and adjectives 
ending in M.H.G. -ic (-ig-) Judeo-German has regularly 
-ik: kinik <M.H.G. kunic (kiinig-) ; hSnik <M.H.G. honic 
(honig-); lebvdik ' alive' < M.H.G. lebendic (le'bendig-). 

c. In a few cases Judeo-German has k< M.H.G. g not alter- 

nating with -c: beykn ' to long for '< M.H.G. bangen ' bange 
werden', binge 'Angst, Sorge'; Mkn 'to look '< M.H.G. 
gucken (here g — k may have become assimilated to k — k). 
Compare t< M.H.G. d(i$e) and/< M.H.G. b (10 d). 

d. In art (e.g. es art mir nit 'it does not concern me, I don't 

care')^ seems to have been syncopated between r and t; 
cf. M.H.G. arget 'macht besorgt, arg'. 

e. g has developed as hiatus-filler in g'srign, past participle of 

sraidn ' to yell '. Possibly r—g as dissimilated product of 
r — r of M.H.G. geschrirn. 

20. M.H.G. k. 
This sound is everywhere preserved: korn< M.H.G. korn ; kez 
' cheese '< M.H.G kcese; kkn < M.H.G. kkin{e); krixn< 
M.H.G. kriechen ; kne'dl 'dumpling '< M.H.G. knodel ; 
hdkn< M.H.G. hacken ; zak< M.H.G. sac (sack-); avik 
'away '< M.H.G. enwe'e (not levelled out to ave'g because 
no longer felt to be connected with veg ' way '). M.H.G. 
qu (i.e. kw) appears as kv: kve'ln 'to bubble with joy '< 
M.H.G. quellen. 

Such, in brief, is the history of the Middle High German 
vowels and consonants in Judeo-German. It will have been 
noticed that the changes in the Judeo-German consonant 
system, when compared with its Middle High German 
prototype, are not as radical as in the case of the vowels 
and that many of the important consonantal developments 



264 THE JEWISH QUARTERLY REVIEW 

are common to modern German. As in the vowel system, 
so also in the consonant system, simplification, though to 
a less degree, has taken place (e.g. M.H.G.^/is represented 
by p or/, according to its position). 

Accent. 
In stress accent no changes have taken place, the stem 
(normally the first) syllable, according to the well-known 
Germanic law of accent, regularly receiving the stress. In 
Ubadik ' alive '<M.H.G. lebendic the accent falls on the 
first syllable, not, as in modern German lebendig, on the 
second ; the lack of stress in the second syllable is probably 
responsible for the syncope of the n. With the Judeo- 
German accent of this word cf. the following from the 
epic of ' Kudrun ' (I, 29) : 

' Si sprach : " s6 riche nieman ist lebendic erkant ".' 

Exceptions to the general law of Germanic accent are 
exceedingly rare. A case in point is svestrkind 'cousin' 
(literally ' sister's (or brother's) child '). 

Hebrew loan-words (Hebrew words are either ultimate 
or, far less frequently, penultimate in accent) accommodate 
themselves so far to the German rule that, if ultimate 
in accent, they throw their stress back to the penultimate 
syllable ; words of more than two syllables, however, can- 
not be accented back of the penult. This sweeping and 
simple law of penultimate accentuation of Hebrew words 
holds, it should be noticed, not merely for such as have 
been incorporated into Judeo-German, but for the present 
pronunciation of Hebrew in general. In the case of 
i.aturalized words a final vowel (whether followed by a 
consonant or not) has, in accordance with the genius of the 



JUDEOGERMAN PHONOLOGY — SAPIR 265 

German language, been weakened to the dull <?. Thus 
Hebrew x*zir ' pig ' > Judeo-German xdzr ; lason ' language' 
> losn ; gannafi ' thief ' >gdn3f; mispdya ' family ' > mispox3. 
In reading Hebrew as such, however, these final vowels are 
not reduced ; the words given above are then pronounced : 
xdzir, ISsen, gdnov, mispoxo. These examples show inci- 
dentally that the Hebrew a and o developed, together 
with the Middle High German d and o, into o and e 
respectively. 

As regards the accentuation of the Slavic (Russian and 
Polish) loan-words, the rule is, on the whole, to keep the 
native accent. It should be noted that such words hold 
relatively the same position in Judeo-German that, e.g., 
French words with un-German accent (such as Position, 
raffiniert) hold in modern German. 

Besides stress accent, a very important factor in the 
pronunciation of Judeo-German is the musical intonation 
of the sentence. In the normal pronunciation of sentences 
there is a very considerable valuation of musical cadence. 
Simple statements, interrogation, surprise, indignation, 
emphatic insistence, irony, and many other moods are 
differentiated by these differences of cadence ; it would be 
possible, indeed, to construct a rather long series of types 
of sentence-cadence for the pronunciation of word groups 
in various emotional keys, some of which would show 
excessively violent rises and falls in pitch. This mobility 
of musical expression gives Judeo-German much of its 
characteristic acoustic effect. The rhetorical effectiveness 
of Judeo-German speech is increased by the use of a large 
number of modal particles (cf. German doch,ja, schon, wohl, 
mal), which are partly Middle High German, partly Slavic, 
and partly Hebrew in origin. Altogether, they neatly hit 



266 THE JEWISH QUARTERLY REVIEW 

off many nuances of mental attitude and despair in many 
cases of adequate translation. 

I trust that I have shown that a thorough investigation 
of the phonology, morphology, and vocabulary of Judeo- 
German will prove abundantly fruitful to students of 
German dialectology.