Skip to main content

Full text of "The Scansion of Mediaeval Hebrew Poetry"

See other formats


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 

Read more about Early Journal Content at 
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 


By B. Halper, Dropsie College. 


One of the essential characteristics of poetic compo- 
sitions, apart from the aesthetic beauty of their contents, 
is the harmonious structure of their outward form. Besides 
polishing every sentence, poets throughout all ages sought 
to arrange syllables in such a way as to obtain a rhythmic 
flow. As formerly poems were composed for recitation, 
they had to be made pleasant to the ear, and hence metre 
and rhyme developed. Incidentally metre and rhyme, 
especially the former, helped to preserve the poems from 
being corrupted in the mouths of the reciters. For in 
most cases when one word is substituted for another 
the metre is disturbed. 

In mediaeval Hebrew poetry, owing to the peculiar 
history of this branch of literature, an accurate knowledge 
of scansion is of vital importance. Composed by writers 
whose aim was to imitate the Arabian poets, the Hebrew 
Diwans were up till recently copied and edited by men 
who had little or no knowledge of Arabic prosody. The 
texts were in many cases corrupted beyond recognition, 
and it is only with the help of the metre that we can 
hope to restore the original lines as they left the hands 
of the author. Thus as an apparatus for textual criticism 
a knowledge of the metres is indispensable. This know- 

vol. iv. 153 M 


ledge, however, must be accurate, otherwise it is apt to 
mislead rather than guide us. It is exasperating to come 
across notes in modern editions of Dlwans, where the editor 
informs us that he emended the text in accordance with 
the metre, when the metre was entirely misunderstood 
by him. 

Alive to the importance of this subject, Jewish writers, 
at different times and from different points of view, made 
many attempts to describe and classify the metres em- 
ployed by the Hebrew poets whose mother tongue was 
Arabic. Foremost among early writers is Sa'adya b. 
Dannan 1 who in his chapte;' on the metres 2 gives an 
exhaustive account of this subject. But although he may 
rightly be called a skilful versifier, and it must likewise 
be admitted that his Hebrew style, despite its slight 
harshness in many places, is quite fluent, he failed to grasp 
the fundamental principles of Arabic prosody. It is cer- 
tainly true that he is versed in the language and literature 
of the Arabs. But it is equally true that he often mis- 
understands his models. To prove this assertion we need 
only mention the fact that Sa'adya in the above-named 
chapter gives the following two verses as Wafir and Kamil 
respectively : 3 

an1ntp o^iip oka d*e>jk n«p (i) 

anna a"oj)D iw-n a^Dsn ■nrn 

^21'ix wd '? nin< nx ^w vria ( 2 ) 

ib»*15 J 1 *"?"? fr?9 n ^j£ 18 y$¥ N,n 

1 The transcription Danan is certainly inaccurate, as in Hebrew it is 
J Hil . On analogy of many proper nouns spelt in this way, as, for instance. 
Hassan, Hajjaj, I think the correct pronunciation is Dannan. 

2 Meleket ha-Sir, ed. Neubauer, 1865. 

3 Op. at., pp. 14, 15. 


He even misunderstands the significations of some of the 
metres. Thus it is well known that Baslt means extended, 
outspread. Yet Sa'adya translates it into Hebrew by DltTB 
simple.*- Now baslt happens to signify ' simple ' in collo- 
quial Arabic and in philosophic terminology, but the very 
compound character of this metre excludes its being called 
by that name. As an apology for Sa'adya it may be 
stated that some Arabian writers committed the same 
error. He renders ramal by ^in sand, 6 confusing rami 
with ramal. 

In recent years this subject was again taken up, and 
in some cases treated more scientifically. But the remark- 
able feature about it is that, although in a matter of this 
nature only one interpretation is possible, unanimity of 
opinion has not been secured. H. Brody, who has 
devoted a great deal of energy to this investigation, and 
has done praiseworthy work in the field of mediaeval 
Hebrew poetry, published a pamphlet entitled Studien zn 
den Dichtungen Jehuda ha-Levis. I. Uber die Metra 
der Versgedichte (Berlin, 1895). In this treatise he gives 
a scientific account of the introduction and development 
of the Arabic metres in the Holy Tongue. Had the 
treatment of the various metres been accurate Brody would 
have left nothing to be desired. Unfortunately, however, 
in many cases his classification of the metres is based on 
opinions which cannot be substantiated, and he was 
therefore driven to resort to anomalous vocalizations. The 
reason of his failure to give a final solution to this simple 
matter must be attributed to his having resorted to theory 

4 Ibid., p. 11. It is unlikely that Sa'adya meant the rare usage of 
JityQ he stretched. 

— T 

6 Op. cit., p. 13. 

M 2 


instead of practice. Instead of studying the Arabic metres 
at first hand, he merely consulted Freytag. Now Freytag's 
Darstellung der arabischen Versktmst, though it was 
excellent for its time (it was published in 1830), has long 
ago become antiquated. Moreover, he follows Sa'adya 
too blindly, and takes over the errors of the latter. It is 
through these circumstances that he is led to state 6 that 
the feet mufctilatun and mutaftfilun are impossible in 
Hebrew, since two moving §wds cannot follow each other. 
As a matter of fact the two consecutive short syllables of 
these two feet are not the rule in Arabic, they may only 
be substituted for a long one. A full description of this 
point will be given later on when Kamil and Wafir will 
be dealt with, and my contention will be proved beyond 
a shadow of a doubt. It is also his opinion that a watid 
mafriik must be excluded, because there can be no §wd 
mobile at the end of the word. But no line in Arabic 
poetry ends in a short syllable, for, as is well known, every 
vowel concluding the line is regarded as long, though it 
may be naturally short. It is thus evident that all the 
forms can be easily transferred to Hebrew, only some 
optional combinations, such as the substitution of two 
short syllables for a long one, could not always be em- 
ployed. Altogether, whenever the Arab poet had the 
option of using a long or short syllable, the Hebrew poet, 
for reasons which will become apparent later on, almost 
invariably preferred to employ a long syllable. This, of 
course, accounts for the fact that has often been observed, 
that the Hebrew metre is less fluctuating than the Arabic. 
If any further proof were needed to show that Brody's 
treatise is inadequate it would be enough to call attention 

6 Metra, p. ai. 


to the fact that two out of the four metres which he 7 
terms tmbestimmte, and states that he cannot trace them, 
are well known in Arabic: (c) is Munsarih, and (d) is 
the shorter Kamil. There are also a few metres in ha- 
Levi's Diwan of which Brody takes no account at all. 

It is my object to give here a concise account of the 
metres employed by mediaeval Hebrew writers. As by 
far the greater bulk of the metres are taken directly from 
Arabic, I shall quote examples from both languages. 
In order to prove my statements, I shall not refer to 
authorities, but shall show the practical application of the 
rules. After all, the concrete embodiment of abstract rules 
gives the reader a better opportunity of judging for himself 
than a mere reference to a famous authority. I shall also 
make use of this occasion to explain some obscure passages 
which have not been hitherto satisfactorily treated, especially 
those on which light may be thrown by quoting parallels 
from Arabic poetry. 

That the Hebrew metres consisting of vowels and 
moving swas were directly borrowed from the Arabs, and 
are the product of a conscious imitation, is a truism which 
needs only to be formally stated to be appreciated. Yet 
the question may be raised how far we can rely on the 
metres of the Arab poets to guide us in analysing the 
poetical creations of their Hebrew followers. For it may 
well be the case that, while attempting to introduce 
a foreign metre into Hebrew, the representatives of the 
Spanish school of Hebrew poets failed to grasp the 
fundamental principles upon which Arabic metres are 
based. Sa'adya b. Dannan ought to be a warning example. 
We in our turn should therefore be wrong in applying 
7 Op. cit., pp. 48, 49. 


the rigid rules governing the Arabic metre to its Hebrew 
offspring. If we wish to understand our poets we must 
take account of their errors. We must look at things from 
their point of view, not from ours. We have to analyse 
the metres as they are, and not as they should have been. 
To this we may reply that whereas later writers failed to 
comprehend the rhythmic flow of Arabic poetry, the earlier 
poets who are responsible for popularizing these metres 
in Hebrew, have fully understood them. Judah ha-Levi 
employed almost every kind of the Arabic metres with 
great skill. And even in his poems — which are not 
numerous at all, and were probably composed at the later 
part of his life, when he repented having followed the 
'customs' of a people whom he detested — which are in 
metres not conforming with the hard and fast rules of the 
Arabs, it is easily seen that they are intentional deviations. 
This is likewise true in the case of Samuel ha-Nagid, 
Ibn Gabirol, Moses Ibn Ezra, Abraham Ibn Ezra, and 

Reproaches have repeatedly been heaped upon the 
representatives of that school of poets for having intro- 
duced into the Holy Tongue a system of foreign metres 
which are against its inherent characteristics, and thereby 
corrupted its purity. The latter part of this complaint 
cannot be substantiated. It may readily be conceded that 
a metre depending on the quantity of the syllables does 
not contribute to render Hebrew poetry rhythmic, since 
that language possesses no short vowels, as far as quantity 
is concerned. To substitute the kva mobile for a short 
vowel is an artificial device which could have made no 
impression on the ear when the poem was recited. It 
may be compared with the method of writing acrostics. 


One may perhaps be led to admire the skill of the writer, 
but the poem is not made sublimer in tone. On the other 
hand, we must repudiate the assertion that the Spanish 
poets corrupted the purity of Hebrew style. As regards 
the rudiments of languages we can safely say that all are 
equal in being against all metres. The tendency of metre 
is to arrange words differently from their natural order. 
Hence it is absurd to consider a particular system of 
metres more suitable for a language than another. Poetry, 
however, likes to be fettered. Pegasus seems to run more 
freely when chained. This may sound paradoxical, but 
is nevertheless true. The very restraint seems to stimulate 
the poetical mind. This is the reason why poetry is more 
polished than prose. There are certain frames of mind 
which can only become active when their area is limited 
and restricted. The Arabs with their characteristic insight 
call poetry manzum ('strung' or 'joined') and prose 
maniur (' scattered '). Oscar Wilde candidly admits that 
a thought often suggested itself to him while in search 
for a rhyme. We have therefore no cause to regret that the 
Spanish school of poets adopted a system of metres which 
apparently deprived them of their freedom. For it 
awakened their energies, and served as an excellent 
stimulus. The result of that activity was gems of thought 
and polished style. 

I shall now proceed to give a short account of the system 
of metres employed by the Arabs. Like those of the 
Greek and Latin poets, the Arabic metres are based on 
the quantity of syllables, that is to say, a number of long 
and short syllables occur regularly according to certain 
rules. Although the earliest Arabic poetic literature 
transmitted to us dates after the Christian era, it is quite 


certain that neither the Greeks nor the Romans influenced 
the Arab poets. For the Arabic metre sprang forth in 
the desert, and the desert-dweller even to-day repels any 
foreign influence. It is, however, a striking feature that 
there are many points of resemblance between the systems 
of prosody of these literatures. Evidently the poetic 
instinct works independently in the same direction among 
different races. The Arab poets divide syllables into two 
classes : (i) those that are long, and (a) those that are 
short. A long syllable is one which contains a long vowel, 
or is closed. To use an English example, go and got 
would be considered long syllables in Arabic. A short 
syllable is one which contains a short vowel and is open. 
A verse is divided into two hemistichs each containing 
a certain amount of feet. A foot consists of a certain 
number of syllables, short and long, disposed in a certain 
order in accordance with the rules appertaining to them 
Feet may be roughly divided into two chief classes : 

(i) Composed of four syllables ; 
(2) Composed of three syllables. 

Feet of four syllables have many variations : 

(a) Diiambus is the one occurring most frequently, and 
its normal form is - «-< - o . 8 Here, as in the Greek and 
Latin literatures, the first syllable may be long, and its 

form is then — <-> . The Arabs call this foot mustaf'ilun. 

This foot may sometimes be reduced to — ^ <-> — or 

— O \J KJ. 

(b) Epitritus tertins, consisting of a spondee and iambus. 
Its normal form is — w . The first syllable may be 

8 These lines are throughout to be read from right to left, as they are 
to be applied to a language written in that direction. 


substituted by two short ones, and the foot is then 
— v — w \j. On account of this peculiarity the Arabs call 
it mutafailun. 

(c) Epitritus primus, consisting of an iambus and 

spondee. Its ordinary form is w. The third syllable 

may be substituted by two short ones, when the foot 
becomes — u w — w . This foot is known as mufailatun. 

(d) Ionicus a minore, consisting of a pyrrhic and 

spondee. Its form is ordinarily <-- w. The first syllable 

may be long, and on this account the Arabs call it 


(e) Antispast, consisting of an iamb and choree, and 

its form is w w. The last syllable is usually long, 

and hence its name is mafallun. It may be changed into 

— W — W. 

Of the feet consisting of three syllables the variations 
are naturally less. 

(a) Brachius, whose normal form is <-> — <->. The third 
syllable may be long, hence it is named fdulun. 

(b) Anapaest, whose normal form is - u u, The first 
syllable may be long, and hence its name is failun. In 
some cases this foot may be substituted by a spondee. 

The hemistichs contain a certain number of these feet. 
A metre may be simple, composed of one class of feet, or 
compound, composed of two kinds of feet. 

According to this principle all syllables in Hebrew are 
long, for, as is well known, in that language no short vowel 
can be open unless it has the accent. Then, according 
to the conception of mediaeval grammarians, accented 
syllables are long as far as metre is concerned. The 
short syllables, therefore, would have to be furnished by the 
light vowels which are extremely rare, and would scarcely 


supply anything like a sufficient number. The swa mobile, 
accordingly, had to be raised to its former glory, and 
given the rank of a short vowel. Arabic-speaking Jews 
probably did not fail to notice that many moving swds 
in Hebrew correspond to short vowels in Arabic, and this 
influenced them to adopt this principle. The light vowel, 
together with the swa which follows it, was in some cases 
regarded as a long syllable, as if the Swa were quiescent. 
But there are some passages, I think, in which the poets 
intended them to be considered as two short syllables. 
This fact has hitherto been overlooked. But if this obvious 
suggestion be adopted, the number of cases where a Swa 
mobile was taken for a quiescent would be appreciably 
reduced. There are a few rare instances in which the 
light vowel was employed as a long syllable and the swa 
as a short one. These details will be pointed out later 
on when occasion arises. 

Apart from the latitude which was allowed to the Arab 
poets in forming their verses, they frequently made use of 
poetical licence. Now that our dictionaries and grammars 
are based on those poetical monuments, we are not in posi- 
tion to realize the extent to which the Arabs took liberty 
with their language. But it is quite certain that many 
forms owe their existence to the fact that the poet wanted 
a certain arrangement of syllables. Thus xiyubligu did not 
fit the metre, he simply changed it into yuballigu. Although 
lexicographers are careful to register all possible forms, 
a reader of poetry often comes across conjugations which 
are not to be found in the standard lexica. Then the 
broken plurals could not all have been used in practical 
life, and there can be no doubt that the multiplicity of 
forms existing of one noun is due to this circumstance. 


When the first. syllable of rtiatun (broken plural of rain) 
had to be long, the poet simply made the word sound 
ni'atun. There are also numerous cases where a hamza 
was changed into wasla. The Hebrew poets, however, were 
less fortunately situated. They wrote in a dead language 
which had long become stereotyped. They were fettered 
by a masorah and a very limited vocabulary. Some of 
them were even too timid to use a word in the plural if its 
singular alone occurred in the Bible. It is only in ex- 
tremely rare instances that we meet a post-biblical word 
or expression. They were afraid of being branded as 
ignoramuses or corrupters of the Holy Tongue. The poet 
wants to say "OB?, but the metre requires a short syllable at 
the beginning of the word. It is quite natural for him to 
use the Pu"al and say "OD? on the analogy of *af and "OE> 
and Isa. 19. 4. But a learned grammarian comes along 
with his concordance and conclusively proves that "QD? does 
not exist in the Old Testament. Our admiration is aroused 
when we see that in spite of these discouraging circum- 
stances the Spanish poets produced sublime verses which 
contain no flaws whatsoever. It makes us think with regret 
of the wealth which these gifted poets would have bestowed 
on the Hebrew language had it been alive and capable of 
being stretched. The poetic licence that the Hebrew poets 
sometimes allowed themselves was to regard a swa mobile 
as qidescent and vice versa, and to use a word in a form 
which does not occur in the Bible. They were, of course, 
severely reproached for these aberrations by later gram- 

It has justly been observed that the metre in Hebrew is 
more fixed, that is to say, it offers less option than the 
Arabic. The reason of this phenomenon is not far to seek. 


We have seen that the Arabs often used a short syllable 
instead of a long one. This is an option which the Hebrew 
poets could easily dispense with, for in Arabic the short 
and long syllables occur in equal proportion, whereas hvas 
must occur much more rarely than all the other vowels put 
together. Then the substitution of two short syllables for 
a long one is only possible in the case of a light vowel and 
the had which follows it. The Hebrew poets, therefore, 
preferred to use as many long syllables as possible. We 
shall see later on that two important metres which are 
extremely frequent in Arabic occur rarely in Hebrew, 
simply because too many short syllables are required. 
Then there is also a psychological reason which will 
account for the fact why the Hebrew poet preferred to 
adhere to a fixed form. The imitator likes to observe as 
strictly as possible all restrictions imposed on him. It is 
his desire to out-Herod Herod. The Arabian poet had 
sufficient confidence in himself, and realized that nothing 
will be detracted from his value if he breaks the monotony 
of having all feet identical in form. But the Hebrew poet 
hesitated in resorting to variations, lest his skill as a 
versifier should be questioned. Nevertheless deviations 
do exist in Hebrew poetry. Editors were unnecessarily 
driven to resort to anomalous vocalizations and even to 
emendations. Brody 9 recognizes this fact, but makes 
wrong use of it. He is right in quoting Freytag that mus- 

taf'ilun (— \-> ) may become mufailun (- w — w), and yet 

be cannot see his way to vocalize "0'BTI? and fTina instead of 
*MB"rv and nina which is in the middle of a sentence. 10 On 
the other hand, in his edition of Judah ha-Levi's Dlwan 11 

9 Metm, p. 7. 10 Op. cit., p. 39. Cp. below, p. ao6. 

11 See especially II, notes, p. 12a. 


he observes that mustafihm may have a &wa mobile pre- 
fixed to it, that is to say, it may become — <~> ' v. This 

assertion is utterly groundless, for the Arabs never per- 
mitted that. No poet is recorded to have allowed himself to 
make use of this privilege of giving mustafihm more than 
four syllables. The fundamental rules of Arabic prosody 
permit the shortening of a foot or the substitution of two 
short syllables for a long one, but never tolerate the 
addition of a superfluous syllable. Should such a principle 
be adopted, we could take almost any piece of prose and 
make it agree with any metre we choose. I shall revert to 
this point when treating of each metre individually. 


Writers on prosody in Hebrew and Arabic are at 
variance as to the arrangement of the several metres. 
This subject was approached by Oriental writers from 
peculiar points of view, and every one gave the order 
which suited his preconceived theory. It is, however, a 
minor question which need not detain us here. I prefer to 
adopt the order which is in accordance with the relative 
frequency the metres occur in Arabic. 

Sixteen metres are recognized by the Arabs, four of 
which may be called favourites which are seldom, if ever, 
absent from any Dlwan. These four are : Tawil, Baslt, 
Kamil, and Wafir. 

1. Tawil, long. 

(a) This is a compound metre consisting of four feet in 
each hemistich, two of which feet are mafailun, and the 
other two are fa'ulun, occurring alternately. Its normal 
form is 

— v — w I O — \j\ WW — w — kj — ww — W| WivU — w 


The length of the line gives the poet an opportunity of 
developing his thought, for Arabs are careful to have a 
complete idea in every line. This is probably the reason 
why every poet sang in this metre. The blind poet Bassar 
has the following beautiful line : n 

' If turbid drinks you never tasted, thirst you did endure, 
For on our earth no man exists whose drinks are always 

In this form the minimum number of short syllables in 
each hemistich is five, and it is therefore no wonder that 
it is extremely rare in Hebrew. Judah ha-Levi has the 
following piyyut: 13 , 

injQ3i -rrin ikt ya\ ''273 

vhii\ ^ris??? Wpo ninri &) 

^jniDa rtfe'san nrpyn Tfxi 

mwi tpybo nt?r£t? t?x aipb 

It should be noticed here that whenever the choice lay 
between a long and a short syllable, the former was 
invariably employed by the Hebrew poet. 

(6) Sometimes the Arabian poets made the fourth foot 
identical in form with the second, especially in the rhyme- 
bearing hemistich. This is necessary when a ridf 'that 
which rides behind', is introduced. 14 Its form is then 

12 Kitabu-l-Agatil, III, p. 28, 1. 7. 13 Harkavy'sed., II, p. 98. 

14 This is a technical name given to one of the letters of prolongation 
N, 1, ', when it immediately precedes the raw!, or unchangeable part of the 
rhyme. See Wright, Arabic Grammar, II, p. 353. 


w|U — w [ ^ | w — \j I] u | w — w | >_> I G" — w 

Mustim b. al-Walld has the following line : 13 

~ o w-o fi op* 0,0 -0-0 .. ^ - 

'She of bewitching eyes, unskilled in the enchanter's art, 
She clings to me in private, openly keeps me apart.' 

This mode, reducing the number of short syllables to 
four, is more frequently met with in Hebrew, and almost 
every poet of note attempted it. Ha-Nagid has several 
short poems in this metre : 10 

'huh ox 1 ? l ni3:6 by ty) cota 
wrqe>n ^ nxs apia 'BBtfn noa 

One at once notices in njab, Arabic badr 'full moon', 
which is always used figuratively of a beautiful woman. 

Ibn Gabirol has the following poem which is abounding 
in Arabic expressions and colouring : 17 

W333 TICK '30 W 'T? 

'nyoi? naps rab a^aK* *^w 

Ibn Gabirol's addressing himself to his friends reminds 
us of halilayya, with which many Arabic poets begin their 
love poems. The description of a sleepless night (line 3 ff.) 
is also borrowed from the Arabs. 13 I suspect that TiOino in 
line 5 b has a different meaning from what it has in the 
Bible. My tumult or confusion hardly suits the context. 

16 Dtwan, ed. De Goeje, p. 37. 10 Harkavy's ed., p. 8. 

17 Dukes, p. 46; Brody, p. 26. ,8 Cp. below, p. 200. 


It is not unlikely that the poet intended this word to stand 
for Arabic humum ' cares '. The cooing dove (line 6) is also 
a favourite theme with the Arabs. Read Wwa in line 8 a, 
for there is no sense in saying, my life and my security is in 
my soul ivhich I love. A glaring Arabism is ??01 in line 10 a. 
It represents walillahi, which is often met with in such 
expressions as lilldhi anta ' How admirable are you !' When 
an Arab quotes a pithy saying he usually introduces it by 
walillahi man kala or walillahi darru man kala ' How 
admirable is he who says !' The line in question ought to 
be rendered, 'How admirable are my friends, my beloved !', 
&c, reading "'ptt'n. Brody, who misses this point, vocalizes 
b#?) in vain, which is against the tenor of the poem. The 
picture of the beloved departing on camels (line n ff.) is 
to be found in many kasidas. Compare especially the 
Mu'allaka of 'Amr b. Kultum : 19 

' I recalled youth's love and yearned for days gone by, 
When at eventide I saw their camels hie.' 

For line 15, "01 rnnp 733 O^sri, compare the words of 
Tauba b. Humayyir : 2!) 

' May the rain from the pure morning clouds give you to 
drink ! ' 

r- - . '". u I T — ** 

The famous poem, ??n tWD f\M ns^ by Judah ha-Levi, 
is also in this metre, although all editors without exception 

19 Lyall, Ten Ancient Aralic Pcems, p. m, 1. 15. 
w Ncldeke, Delictus, p. 5, 1. 8. 


failed to recognize it, and hence many corruptions arose. 
Harkavy, 21 in addition to the printed errors, suggests a few 
more which would conceal the real metre beyond recog- 
nition. Brody thinks that its metre is 

that is to say, a variation of Baslt. The objection to this 
scansion is obvious, for the poem begins with a hwa mobile. 
In his notes 22 Brody meets this objection with the assertion 
that the prefixed §wds are not to be counted. As I re- 
marked above, 23 no mustaf'ilun can have more than four 
syllables. In Brody's edition there are in five lines thirteen 
superfluous fauas occurring at the beginning of words, and 
can by no stretch of imagination be taken as quiescent 
ones. Now in order to make this poem agree with this 
mode of Tawll seven more hvds are required. A proper 
analysis of each line individually will at once show that 
these hvds were there originally, but were lost through the 
failure of editors and copyists to recognize the metre. It is 
a significant fact that all the acrostic-bearing words are 
quite in order, and have moving hvds. It is only some of 
the other words which had no ' fence ' to protect them that 
suffered corruption. Now out of the seven moving hvds 
that have to be restored, two can be obtained by altering 
the vocalization. Read nnjp (line la), deriving it from 
rfnjpn (Jer. 48. 38) and i? (line 1 b), referring it to *|ta ns; ( 
which is masculine. In line 4 b we should read *["tiff\ , thus 
preparing for the climax Oi). Editors have sometimes 
found it necessary to offer such anomalous punctuations 
as *M3 in line 3a; but I think that originally it was 
Bnefon 13? (cp. OW ejas, Job 39. 13). Copyists probably 

21 Notes, I, p. 170. 22 n , p. 122. » p. 165. 



considered it their duty to emend it in accordance with 
the familiar expression occurring in the Bible. Such 
biblical reminiscences are to be found very frequently in 
editions in which the metre is disregarded. My suggestion 
is rendered more certain by the fact that variant readings 
of this word are recorded by editors. It has been my 
observation that the majority of corruptions arose under 
such circumstances. Whenever the poet for exigences of 
metre was compelled to deviate from familiar phrases, the 
copyists were ready to correct the error. *3 in line 3 a 
should be read "f^ or ft}Q. Here again the copyist knew 
of Ps. 77. 12. Similarly ?ja J'K ?]aj>ip (line 4 a) is a reminis- 
cence of Jer. 8. 19, and the poet probably wrote =1? P$? 
T5^?. 2 * For TINBO read nKBB INC. Here again editors 
record variations. The whole poem should run as follows : 

an -]2»? nnp ban vmra sju na? 

anyo nNBD *tkd^b>w nspaa 1? 
Dip maw nya "loaa wi ttort 

ann t^n ijui. ^? ffc Ttf 2 ? 

ny ontsfen f)aa by , 3:w »»* 

anyrn TpDj •'nyana nyix 

Dipna Dxi -jiabp *ja px Dao tpfifTt 

any. 1 ; eoto *ab ^an oyoi 

It should be observed that in line 3 a iy will be better 
understood when we know that it has the meaning of 
Arabic hatta with the subjunctive=' in order that'. 25 

x Cp. 'H'nE'X, Eccles. n. 17. 

S5 See Wright, Arabic Grammar, II, p. 29. 


Those who object to this mode of treating the text 
should bear in mind that a careful study of this poem will 
convince any one that a regular metre was intended by the 
poet, and this is the only one possible. 

For another probable variation of Tawll, see below, 
p. 183. 

a. Basit, outspread. 

(a) This is also a compound metre which may be 
regarded as a companion to Tawll. The number of its 
syllables is the same as that of the latter. It is composed 
of mustafilun and fd'ilun, occurring alternately, and its 
normal form is 

— u Uj -<j — "O — wU — w — Ujj — wUj — \j — w| — v^wj — \j — sZ7 

The following line is attributed to Abu Adhina. 20 

- *^ - 9 ■" o O-o 9 9 v) *■ 9 rf - 

' A man attains not ev'ry day his quest, 
Nor 's he allowed with fate's gift to be blest.' 

The minimum number of short syllables is four in each 
hemistich, and this excludes its becoming a favourite with 
the Hebrew poets. Nevertheless almost every poet 
attempted to master it. Ha-Nagid has the following poem 
which has been transmitted in a very corrupt form. The 
metre is, however, quite evident : 27 

_„Vil- '- - !-, v '_j_ J _ I, 
n^ns' tfpw* Dia ^ Djjfoai ^aaa 

26 Abul Feda, Historia anteislamtca, ed. Fleischer, p. 124, 1. 8. 

27 Harkavy's ed., p. 107. 

N 2 


Ha-Levi has the following piyyut : 28 

"•Jfctfa toe* |iiK ijB spiels oi>n 

It should be observed that the hua quiescent of no??] 
is regarded here as mobile. 

(b) Arabian poets allow the rhyme-bearing fetilun to 
be reduced to a spondee. This is necessitated by the 
introduction of a rid/. 29 Hassan b. Thabit has a lyrical 
poem in this metre : 30 

O -^- * * % M o- , l '°*°^ 

-4-5 —Li i) LAjl ^jA-iJ JLJIJ 
jJW ejJoJJl Jj-ol is*-*-?- J-i-*—^ 

'And wealth doth overwhelm those men who are not staid, 
As sweeps away a flood the stems that are decayed.' 

In this form the metre is more frequently met with in 
Hebrew. Abraham Ibn Ezra has the following lines : 31 
^ ty nip prn rn -rate -na 
tann 5)1^ ny liaaa injn? n»b> 
fivp ^ rfyv: vruax niD iid? -ipr6 

Ha-Levi's panegyric to Solomon b. Farosal is in this 
metre : 32 

tv*: — •• : : 't; • t 

T V t; t : T -; T J V 

28 Harkavy, II, p. 141. 29 See above, p. 166. 

80 Diwan, ed. Hirschfeld, p. 6g, no. 159, 1. 7. 

31 Egers's ed., p. 10. S2 Brody, I, p. 14. 


Like most of the poems of this class, the present one 
is full of Arabic colouring and imagery. It is true that 
D*BE>3 npga occurs in Nahum 3. 4, but the sense required 
here is not quite identical with that of the Bible. Here 
it is an enchantress of a different kind, and reminds us 
of the Arabic sakir(un). Then D , *i^3 does not mean 
wandering in the biblical sense, but corresponds to Arabic 
hajr(tin) or kijrdnitm) ' departure, separation '. This is 
the signification of this word in the majority of cases in 
mediaeval Hebrew poetry, where it is not the T$a who 
suffers, but the one left behind. The first verse of our 
poem is to be found, in one form or another, in almost 
every Arabic Dlwan. Muslim b. al-Walld says : 33 

9 ~ - 9 -~ * «-~*-0 

' Her absence is near, and her company is far,' that is 
to say, her separation lasts longer than her friendship. 

Line % a is almost a literal translation of another verse 
by Muslim b. al-Walid. A lady vaunting her beauty 
says : 34 

9 O- - 9 - 9 0-0 9 O » -o ^ 

„o-o t-o' 9 o - o <*\ 

LA;JtJl) U_5*£l C*-**J iO-^-H 

' I'm the illuminating sun when it doth shine, 
But I am never known to set or decline.' 

The same lady says : 35 


„0-» 9 » 9 o " 

c*~2~ J^" Ws-k ^ij 0-**Jj 

' I desire no perfume but my own,' 
and this explains line a b. 

33 Dlwan, p. 155, 1. 11. " Op. at, p. 153, 1. 8. s5 /. c, 1. 11 b. 


Brody 36 has rightly pointed out that lines y, 8 are 
extremely difficult. The expression nna 'cai is quite 
familiar in Hebrew as well as in Arabic from which it 
is borrowed, and calls for no special explanation. But 
I think if we try to get the Arabic expressions which these 
lines represent, we may obtain a satisfactory solution. 
One often meets some such sentence as nafsl fida leilin 
yufakku bihi asiruki ' I would offer my soul as a ransom 
for the night in which your prisoner is released '. Now 
when Arabian poets talk of releasing this kind of prisoner 
their expressions may be interpreted in two ways : 

(i) The poet wishes to become indifferent to the object 
of his affection, that is to say, his desire is to be freed 
from her bewitching influence. We often hear the dis- 
appointed lover arguing logically with his captivator: 
'Either requite me with your love, or give me back my 
heart of which you deprived me.' And cruel Venus 
mockingly replies : ' neither of your requests will be 

(2) On other occasions the poet in asking for freedom 
wishes to have his desire gratified. That our poet here 
intends to convey the latter meaning is quite evident from 
the following lines in which love's revelry is artistically 
described. The greatest difficulty lies in line 7 b : 

To my mind ->an has been misunderstood. It does not 
denote region or district, but rope, and together with O^ono 
represents Arabic hablu-l-hawa 'the rope of love', which 
one meets so often. Ibn Gabirol translates this phrase by 
^an^n van, 3 ? which renders my suggestion more certain, 

36 Notes, p. 26. 8 ' Dukes' ed., p. 42, no. 30, 1. 9 a. 


since the Hebrew poets used all these synonyms. '""J"]^ 1 ? 
is very likely a metaphor for body, which in prose would 
be "ion. The hemistich should accordingly be rendered: 
Only the body is in the rope of her love. The poet thus 
explains what kind of freedom he is asking for. Let the 
soul alone be free from anxiety, but the body should still 
be enslaved by her love. He then goes on depicting the 
riotous revelry for which he is longing : 

wxa roni Dnx nst? rt'xoi line* 
nnao Dtan »m nae ms 

t vt : - • -t : T * 

This verse has also been misunderstood. Brody 38 
remarks mm znsb w 'bm , . . on .-uann ^p ni>xn ownrr 
jt-bd Kin is *ie>n iijn /ijo KW n>a pjiva. Let the reader 
try to construe this verse, and see how far the difficulty 
is solved. It seems to me that O^IJO is here used meta- 
phorically for lips. It should be noted that in the Bible 
this word has no fixed meaning. There is no doubt that 
it is identical with Arabic majd(un) ' honour, glory ', and 
very likely has the signification of choice, excellence. The 
transition to the metaphor is, therefore, not very far. ""SI 
(line 8 b) is not construct state, depending on Diart, but 
is with pronominal suffix 1st person singular. The line 
should be translated : 

' To quaff and drain her rosy lips, and sate 
My mouth and lips with her mouth's honeycomb — 
Her lips will serve as cup ! ' 

In the note quoted above, Brody proceeds : nanJD nvn 
its i\rb do pnni> njnv v?m ma m ws>: rowix nx bbrb 
nax nay iaa did '•a (136 nx j*s 'd) n">-\ b>"d31. Here again 

38 I, notes, p. 26. 


Brody misses the point. His note to the latter poem 

runs : 39 

nstt *is'y "aa D*b 19 

na bx na ftaw *j« 

in iptw i^a noil I'm Dttn is wmi nsj? nx rreni "p^y w n^ 
(dim "i&w pin torn nay ^)hjdi (nay »aa n^x pnn)ijn nnx nN. 
I wonder whether the reader could follow this description. 
In this case also "ti = my mouth, and the verse should be 
rendered : 

' Watch my mouth is a cup in an antelope's mouth, 
My wine and his lip, — mouth to mouth ! ' 

What a beautiful metaphor, and how befogged and 
mutilated it emerges from a faulty construing ! Those 
who are acquainted with the figurative expressions of the 
Arabs will, I hope, have no objection to my explanation. 
An Arab says ' the swimmer' (sabih) and means the camel 
which is the ship of the desert. He says ' the fettered ' 
[mukayyad) and means the foot. Ha-Levi himself speaks 
of apples and means the breasts. 40 Sometimes Hebrew 
poets say pomegranates instead of apples. 41 Of course 
this metaphor is also borrowed from the Arabs. Muslim 
b. al-Walld says : 42 

U-9-N cl^-f 0~^ mW-^-' ' P)J 

fi 0% 

' Love's pomegranates did youth plant 
Upon their chests, adorned with bones.' 

39 Op. cit., p. 223. 

40 See op. cit., II, p. 18, and still more explicit II, p. 6. 

41 Moses ibn Ezra, Tarsis, p. 36, no. 106, and also ha-Levi, I, p. 100. 

42 Dtwan, p. 149, 1. 30. 


Ya'kob b. El'azar has the metaphor of the cup and 
lip in more explicit terms : 43 

,T»n nab" ipb iw ^o 

To return to our poem. In line 17 a, both Brody and 
Harkavy punctuate naa which is impossible, since this word 
is intransitive in Hebrew, and it certainly cannot be taken 
as predicate to y^P) which is plural. It is therefore 
necessary to vocalize naa. This expression can only be 
properly understood with the help of Arabic. In the 
Bible the extinction of one's light is supposed to be a curse, 
as is plainly seen from Prov. 30. 20 ; Job 1 8. 4, and a few 
more passages. But the expression in this poem repre- 
sents Arabic ahmada niranaha or at/a a nirdnaha 'he 
quenched her flames'. The woe-begone lover often com- 
plains fi fu'adl nar(un) leisa laha humud(un) ' within my 
heart there is a flame which does not abate', or 'is not 
extinguished '. When his hope is realized he may say 
uhmida-n-nar(u) or utft a-n-nar(u). The expression of 
Canticles 8. 7 is not to the point here, for in that verse 
there is no reference to the realization of love's hope. 

In this metre is also ha-Levi's famous poem : 44 

tj;tdn Dibe*} ^phfn $rj f& 

asbf iD'm jtesw rnioi d;d 

Al-Harizi evidently had this poem in mind when he 

is Brody and Albrecht, Sa'ar ha-Slr, p. 163. 
44 Brody's ed., II, p. 155. 


wrote the following lines which have the same metre and 
approximately the same rhyme : 45 

?p-it»>s« mete noxn waj "ap 

1 V; - : t; - v v; t • : v '■„• 

inajo nibfi i^na Dib^ 

'jE'w "r?b wn i>KBjn; PI qua ring 

I^Iwd 'pa fM pneqi pin ix 

(<?) But even in this form several short syllables are 
necessary for each line. Hence a further step was taken. 
On analogy of the rhyme-bearing anapaest, all anapaests 
were reduced to spondees, and the following variation of 
Baslt was the result : 

__|_ w __|__|_„__||__|_ w __|__|_ v ,__ 

As only four short syllables are to be employed in 
a line of eight feet, the Arabian poets could not very well 
compose their poems in this metre, and writers do not 
classify it among the proper metres. In Hebrew, however, 
it could be employed with advantage. Moses Ibn Ezra 
in his TarUh has the following lines : 46 

•px f6 nVri artNPi bin *by aix ^3 

» • t v: - : v -: t 

ps K»3rfm ma tj? msob &?b 

The following poem by Ibn Gabirol is in this metre : m 
nnfen o?p by rha\f fsa 
n»a»B» nbyo ix n»ax nsn 

t* •* : t;v t*; : 

\sm pine* si'dnpi no »n sriJK nix-a 

46 TafykemonT, p. 140. 4G p. 24. 

47 Dukes' ed., p. 35. See also Geiger, Salomo Gabirol, pp. 67, 134. 


In Dukes' edition the metre is in some places not clear. 
But one can in most cases restore the text without difficulty. 
In the above lines by had to be changed into "&$( and the 
hemistichs differently divided. In *|bKn (line a a) I recog- 
nize an Arabism. It does not signify gather, but combine 
like Arabic jamda. The whole line should be rendered : 
When seeing a living-dead man she combines mirth and 
weeping, and like dew upon a rose is a tear upon her cheek. 
For the rest of the poem a few more, emendations are 
necessary to make it agree with the metre and yield good 
sense. In line 4 a delete b of nSvnb. The preceding word 
iiDn can scarcely be right. Perhaps some such word as 
nsiri should be read. Delete 1 of nici (line 5 b). In line 
6 delete 1 of nnwi and paoi. In line 7 b 1 of inw should 
be deleted. Either punctuate ri?^K> or read njaitr in line 9 a. 
Vocalize njBJH (line 10 b), not as in the text. In line 13 a 
read roan instead of ■'Dan, and in the second hemistich point 
nvatf, not n;?E\ The end of line 14 is obviously corrupt, but 
the scansion is quite correct. nwiKI should perhaps be 
divided into njn riKl. But the whole line is doubtful. 48 #" 
»z£# o/*r£ divided they would form two companies with two 
souls, but you are the moving spirit of both is too much cir- 
cumlocution iot you are the moving spirit of mankind. In 
line 15 rivns is against grammar. Tie (abstract noun, as in 
Prov. 1. aa) probably was the original reading which was 
afterwards substituted by the more usual word. Sense 
and metre demand the reading 'teas pin* *3 instead of the 
vague niD3 £so (line 15 b). The b of npb (line 17 a) should 
be deleted, for the imperative is obviously intended. In 
line 18 a, 1 of oStO should be deleted. 

48 See, however, Geiger, p. 134. 


To the same metre belongs ha-Levi's panegyric 
addressed to Moses Ibn Ezra: 49 

*b* 5>»5>? nsao wa inn jn 
'Ma *wb> liiia not£ jot nno 

I have advisedly omitted to punctuate *ikb>. All editions 
take it to be "M? *7&? ?■£?*; but I am not sure whether 
"WP flesh would not suit the context better. The assigning 
of the metre of this poem, it should be observed, is open to 
dispute. Brody thinks the metre is 

This metre is quite unknown, and I do not think there 
is sufficient ground for making this poem an exception, 
and the insertion of the short syllable is scarcely justifiable. 
This poem consists of twenty lines, thirteen out of which 
have swds, moving and resting, in the middle of words, and 
whichever way we scan the lines grammatical rules are 
violated. The real objection to my method of scansion 
is the fact that five lines contain favas at the beginning 
of words. But this objection is only apparent. In line 3 
read btonb instead of bton b$ (cp. Jer. 22. 10). In line 6 
delete 1 of ti&\. Editors have found great difficulty in 
line 7. Some manuscripts have ijn which cannot be made 
to fit the supposed metre. Brody has a long note on this 
line, but all the explanations he quotes are admittedly 
inadequate. The difficulty is, however, only created by 
editors who are compelled to read nj3 to get an extra 
short syllable, and I think the proper reading is 
"six ye nya neon Vijfe> ty 

The night of his hair (i. e. his black hair) covers him 
49 Brody's ed., I. p. 122. 


from my eye (lit. is a covering for my eye). For the meta- 
phorical use of nais cp. rrixVin (Eccles. 12. 3). Brody reads 
■>£ftf iJS »ip, and tacitly admits that it is difficult. All 
kinds of fanciful explanations were suggested for '•aft. The 
way I vocalize it we have in ">y3 an instance of — w — o. 
In line 8 read Jfitpx instead of nyatpK. In line 10 read Til 
instead of THJ. Line 12 has again caused difficulty. 
Harkavy has 1y$1, which is against either way of scanning. 
It has therefore been suggested to read \?gl, and numerous 
suggestions were offered, all of which are, to say the least, 
far-fetched. But the simplest thing is to read fb^ } and this 
clause should be rendered : About him are the choice words 
of my mouth. Another obscure line is 13. If iy be cor- 
rect, the only possible vocalization is ISP will be hostile, or 
vex (cp. Isa. 11. 13). Harkavy's "IV is certainly unintelligible, 
and Brody's T5P is inadequate, for this word denotes will be 
in straits. Whichever way we take IV the b of ?J? is indis- 
pensable. I therefore think that we should delete ?I, and 
translate the line : If Time is hostile I shall take refuge in 
his company. According to this conception iN in the follow- 
ing line will have to be emended to ON . If, however, the 
conception of this line as given by commentators is right, 
the best reading would be 

iirpna nVno "fry: join dk 

In line 15 "isai seems to have a meaning slightly different 
from that in the Bible. For it evidently signifies here 
persist, and it is not unlikely that the mediaeval writers 
were influenced by Arabic asarra 'he persisted, perse- 
vered'. Instead of W, which is a biblical reminiscence, 
read BM. 

The two Arabic lines at the end of this poem have been 


slightly misunderstood by Harkavy and Brody. The 
former translates ^n by pun and the latter by abn. Now in 
Arabic hal introduces an open question, so that both ren- 
derings are inaccurate. Brody is nearer the right meaning 
of the second line, which Harkavy renders very vaguely. 
The proper reading, however, is that of Geiger : bn = he 
dwelled. The lines are to be translated : 
' Ye at the van of knowledge, glory, and high rank, 
Turn to the rear when Musa dwelleth in your midst.' 

The poet has skilfully used the two antitheses kaid ' a 
leader ' and half(un) ' rear '. Halla may be conditional ; 
but we may also render : for Musa dwelleth, &c. 

As these lines stand they agree with neither method of 
scansion. But in few poems do the Arabic lines agree with 
the metre. This is certainly due to the fact that many of 
the copyists knew no Arabic, and paid no attention to the 
scansion of the lines, and hence many corruptions arose. 
These lines, however, are straightforward, and the first one 
is an ordinary Baslt, and this makes my scansion more 
probable. In the second line we should have to alter the 
arrangement of the words to make it fit the ordinary Baslt. 
But the following reading which involves insignificant 
changes may be suggested to make the lines agree with the 
metre of the rest of the poem : 

epefort babibti obybx rnap & 
f\bi bit jhikb idio 033 bn 

The singular jttlKS would refer to each one separately. 

It should be noted here that Sa'adya b. Dannan so gives 
this metre as the third Tawll. In this case we have to 
assume that fdulun became a spondee, which is at first 

60 Melehet ha-Slr, p. n. 


sight hardly possible. The Arabs allow in Tawll to drop 
the first short syllable only. This is sometimes the case at 
the beginning of the first line of the poem. For then the 
audience may not notice the omission, and the rhythmic 
beat is not thereby disturbed. In the middle of the line, 
however, it is inadmissible. Some writers on prosody 
record rare exceptions where later poets allow themselves 
to drop the short syllable at the beginning of a hemistich 
in the middle of a poem. But I think in these cases a new 
poem was intended. For those who arranged the Dlwans 
sometimes made two separate poems run into one, when 
both had the same rhyme and metre. Rosin 51 follows 
Sa'adya, and observes that Freytag (p. 170) supports this 
rule. But Freytag does nothing of the kind. He explicitly 
states that only at the beginning of the hemistich is one 
allowed to omit the short syllable. However, Sa'adya may 
not be altogether wrong. There are a few poems apparently 
in this metre which have caused some difficulty, as some 
of the supposed mastafilun feet are prefixed by §wd mobile. 
Brody asserts that this is permissible, but I have sufficiently 
explained above that this view is untenable. At first sight 
one is inclined to resort to emendations, and it is my 
opinion that if emendations are justified in any field, 
mediaeval Hebrew poetry should come in for a large share. 
In some cases this may be effected with more success. 
But there are many poems in which these unnecessary swas 
are present, and no emendations are justified. Moreover, 
it is remarkable that this peculiarity of having a superfluous 
Swd seldom occurs in Kamil or Wafir, but is usually present 
in this kind of metre. I am therefore inclined to believe 
that the Hebrew poets proceeded, without justification, of 

51 Reime und Gedichte, p. 8. 


course, to reduce the fa'itlun foot to a spondee, or better 
still, to make the first syllable of fa'fihm optional. The 
scansion would then be 

w |__ W |___ w |__( w) ||___„|__ M | |__ (w) 

When no short syllable occurred at all this secondary Tawil 
coincided with the secondary Basit, and this accounts for 
Sa'adya's opinion. But since there is more justification for 
reducing fctilim to a spondee, I prefer to assign all poems 
which have no superfluous syllables to the Basit metre. 
Those, however, in which superfluous syllables occur must 
be regarded as belonging to a secondary Tawil. Ibn 
Gabirol's following poem is in this metre : 52 

nsj^oi rta'y nriK> top nst v? 

na^ ik» ma ran -rita -van 

TT : TT T - : - T 

.HEfto -to lopp rvn? nn 

For obvious reasons I cannot assign the poems *fa na^ 
(above p. 170) and rbinf fB3 (above p. 178) to this metre. 
The former, as was shown above, can by proper analysis be 
made to be a regular Tawil, and in the latter many of the 
emendations are necessitated by the context independently 
of scansion. 

3. Kamil, perfect. 

This metre is composed of Epitritus tertius repeated 
three times in each hemistich. For the first syllable of 
every foot two short ones may be substituted, which are 
regarded as equal to a long one, and it is from this circum- 
stance that its name is derived. For it has more compulsory 
long syllables than any other metre with the exception of 

52 Dukes' ed., p. 40 ; Sa'ar ka-Sir, p. 38. 


Wafir, which is its companion. It is of very frequent 
occurrence in Arabic, and has several well-established 
variations. Hebrew poets could handle it with great 
facility, as the number of compulsory short syllables does 
not exceed three in a hemistich containing nine long 
syllables, and it is therefore no wonder that one meets it 
very frequently. Jewish writers on prosody, however, have 
confused it with Rajaz, and Kamil is entirely excluded. 
Sa'adya b. Dannan also takes all the Hebrew poems in 
Kamil as Rajaz, and ' invents ' a new combination which he 
imagines to be Kamil. 53 The reason of this misunder- 
standing is not far to seek. In Arabic the Kamil is at once 
recognized by the peculiarity referred to above. But in 
Hebrew it is impossible to have two moving swas in suc- 
cession, and hence mustafilun and mutafailun coincide. 
A light vowel with its hwa mobile, which very likely repre- 
sents w w was taken as — . But a careful study of Arabic 
prosody will prove beyond doubt that the view to exclude 
Kamil and make Rajaz occur often is untenable. The 
latter metre in Arabic is mostly of two feet, and allows of 
so many variations in each foot that many writers on prosody 
deny it the rank of a developed metre. It is no doubt the 
connecting-link between the rhymed prose known as saf 
and the metres which were developed later on. Its close 
affinity with rhymed prose is quite evident from the fact 
that one and the same poem may have hemistichs of two 
and three feet. Altogether it is to be employed for extem- 
pore poetry. Whenever I read a poem in Rajaz I have 
the same impression as when reading the short Kur'anic 
outbursts. Ample illustration of the above remarks will 
be found later on when Rajaz will be treated of. On the 

63 See above, p. 154. 


other hand, Kamil is a metre of full dignity, and is a 
favourite with all Arabian poets. Two of the Mu'allakat 
— those of Labld and 'Antara — are written in it. It is, 
therefore, quite impossible that the Hebrew poets should 
adopt Rajaz, and leave Kamil out. Furthermore, there are 
metres of this type which have no equivalent in Rajaz, but 
are well-known forms of Kamil. 5 * 

(a) Its normal form is 

_ w _S^|_ w ._k<w|_ l _,_v ! ^|j_ w _v = A ! 

Labld's Mtiallaka is in this metre : 55 

«- r — r — o . — ' — q— r o — ^ <* 

- » .. - - — 

-}~~s' 9 * - - — s*t ) 

' I stood beseeching them, but what avails it to beseech 

Those deaf and stolid things ? not to be fathomed is their 

speech ! ' 

Moses Ibn Ezra has the following lines in his Tarsis: 66 

to? wxb fvn 'lite -0 

n?j)n its jtqji nriB' 

mi ty nten inp njrnai 

This is a favourite picture in Arabic love poems and 
panegyrics, "inp obviously represents badr ' full moon '. 

Judah ha-Levi has expressed his belief in the immortality 
of the Jewish race in this metre : 57 

54 See below under d. 

65 Lyall, Ten Ancient Arabi: Poems, p. 6g, 1. 24. 

66 p. 30. 

57 Brody, II, 307. 


mom n^> n^^i ota nipn 

ai>£ nib ninte «.3ij an 

wis; s6} '•ia obiyb ni^nj* 

Ha-Nagid has the following lines : 58 
nnxrin *6bo nsyb nabo 

^j? Tinwan na-iK Tab msy 
n-iiynn njnnn ^n feny 

Al-Harizi in his Mahberet Itiel says : 59 
wett* bs bjn m ma tacts' 
va^K !>a ei'nn; bx sjki *w?n 

T ; T T ■• -; T 

vanix Di^ 'oa 'a nayai few 

Almost every poet of note attempted this metre. At 
the same time, to make it more extensively employed, it 
was advisable to elide a short syllable. The Arabian poets 
often allow the rhyme-bearing foot to consist of three long 
syllables. This must be the case when a rid/ 60 is required. 

(b) Its form is then 

Vi=i[_ w _M^|_ w _!^|j_ w _S=i!=i|_ w _^|_ v _ / _i = !w 

Muslim b. al-Walid has his long panegyric addressed to 
Zeid b. Muslim in this metre : 61 

68 Harkavy's ed., p. 8. 

69 Chenery's ed., p. is. 

60 See above, p. 166. 

61 Dijodn, p. 165, 1. 83. 



9 9 „ I — ■ v * * O_o .fi 

" - 2 * o-o * o* - si — • 

' I see calamities from me do not depart, 
I am their goal — they aim their arrows at my heart.' 

Ha- Levi has the following poem in this metre : M 
risM?' ^etei rm9$ "fan 

pint? f 3 }D?n ski; niyn 
nat3iK> ^nym ni\n ins 

Like the majority of all other love poems this one 
contains many Arabisms. To begin with, nil?! cannot have 
here the signification it has in the Bible, for the question 
would not be to the point. When a man is young there is 
no need for him to long for youth ; it is just when he grows 
aged that he is inspired with yearnings for the days that 
are no more. Brody in his notes approximately gets the 
right sense when he remarks that youth with all its plea- 
sures is meant here. But the meaning of the line becomes 
infinitely clearer when we know that nnp^ here is a trans- 
lation of Arabic siba(n), which denotes ' youth ' as well as 
' ardour of youthly love '. The question then becomes 
forceful : Do you long for youthly love after your lock grew 
grey? najnp also has a slightly different meaning from 
what it has in the Bible. It means here ' forelock', and 
represents Arabic duabat(un). One often comes across 
such an expression as §amila-l-ma&ibu-d-dawdiba ' hoariness 
covered the forelocks '. 

In Dukes' edition of Ibn Gabirol's Dlwan (p. ao) there 

62 Brody's ed., I, p. 129. 


is a contemplative poem full of sublime reflections which 
is in this metre, but owing to its deplorably corrupt state 
there are many obscure passages which require radical 
changes in order to yield a suitable meaning and fit the 
metre. S. D. Luzzatto and Senior Sachs 63 have made 
attempts to explain some of the difficulties, but have ad- 
mittedly left a great deal to be desired. Geiger has a free 
translation, and some notes in his treatise on Ibn Gabirol. 64 
My conception of the poem differs fundamentally from that 
of these scholars: 

T* *t;t t : * 

irm? Darin nnjx "nebi 

Even in these four hemistichs some slight changes are 
necessary. Dukes has ">b\tt, and Luzzatto has emended it 
to y5§ drops. Senior Sachs in Osar Hokmah 65 reads w , 
but withdrew his suggestion in Ha-Zophe Le-Ha-Maggid, M 
because he found in a book of a mediaeval grammarian that 
v*K here is to be understood in the sense of \?v. This, of 
course, makes no difference in regard to the sense. As 
Ibn Gabirol certainly wanted to say if, 1?K is better than ylN. 
For no one would attach any weight to the reading of the 
manuscript used by Dukes. Instead of VtDljiDTP and WJJ 
(line a a) Dukes has lnononi and 1JJ53. 

The poet declares that he does not weep now. His 
tears tarry, because they are weary of flowing, and seeing 

63 Professor Israel Davidson kindly supplied me with a list of places 
where some of Ibn Gabirol's poems were published. The poem in question 
was commented upon in Ha-Maggld, III, pp. 146, 150 ; Ha-Zophe le-Ha- 
MaggJd, VI, pp. 253, 276 ; Osar Hokmah, II, p. 35. 

64 PP- 45. las - 65 "I; P- 35- 66 VI, p. 276. 


a sigh which follows them they retreat (lit. flee). Had he 
wept his cheeks would have been saturated with tears, and 
would have produced sprouts. This exaggerating metaphor 
occurs in Muslim b. al-Walld's Diwan : 67 

">- - - - - O-o 9 o f . — 3 

' Upon my cheek from weeping grass did grow 
Luxuriant upon my liver was love's bough.' 

The poet's silence, however, must not be taken for 
a sign that his grief is abated, or that he is unable to weep. 
He goes on to describe a night which he spent in weeping : 

to ^!*3 *j»j> n« Vb 

\r$> raaia nTfpj? Tofe*? 

This, I think, is the right reading instead of nTDJ? TW3. 
The meaning accordingly is : The night in which my eyes 
shed tears was such as if the stars were at a standstill 
(lit. took silence) like a hireling of standing (or waiting ; 
cp. Arabic wakafd). The description of a sleepless night 
is a great favourite with Arabian kasida writers who wish 
to attract the attention of the audience by recounting their 
grief. The poet usually complains that the night appeared 
to be everlasting. When treating of Wafir I shall quote 
one or two passages illustrating this figure of speech. The 
pronominal suffix of V33U refers to b"b, and not, as Dukes 
thinks, to »j*j>, which is feminine plural. 

The poet's thoughts now turn to describe the stars : 

nngiD iort totnn & 
W inns rtan too 

t* t-j - •.*••; ' - T 

67 No. 64, p. si8, 1. 3. 


The printed edition has lDtn\ and Dukes rightly remarks 
that it should be changed into 1»n\ On account of the 
metre I prefixed » to ^riN. 

W> |3 DDi33 nj? "a:6 vpft 
vfo naa Dp tat'oai £:p 

t -t : 't - : • : : • 

This is a very difficult verse, and many emendations 
were suggested, none of which can claim any degree of 
probability. The MS. vocalizes VpJ!) which, I think, is 
right. The last word of this verse in the text is )np\> which is 
against the metre, and was emended to Vi|J. Now this word 
hardly gives any sense. Luzzatto's reading f|3 . . . . VjW 
is ingenious, but does not suit the context. I therefore 
suggest the reading vo. It is true that the verb is not 
biblical, but the same objection applies to Vijj. 08 It seems 
to me that )?, which in the Bible means a pedestal, is used 
here for the peg of the tent. The poet thus describes the 
departure of his friends. The verse should be translated : 
My heart grew sick (lit. was dislocated) when the pedestal 
was raised, (out of grief) lest they be destroyed, and in a 
while my clothes grew wet. This translation is strengthened 
by the following line : Had not the morning star made 
my grief to cease, my neighbours would have swum in tears. 
In this line '3 is to be deleted as dittography. The 
following four lines form a digression of a contemplative 
character which is quite usual in any Arabic kaslda. 
The sense is straightforward, and all emendations are quite 
unnecessary, except that in line 9 a ^5j is to be read instead 
of 7]1 on account of the metre. Line 10 : 

t • — : * t : t 

68 For 7 can only be dropped when it has a swd. 


should be compared with the line of Hisam b. 'Ukba 
al-Adawi: 69 

> - t B ° * - ^o^o-c "' •* o «" - 

? B * * f o a <3 -a fi ~ o-t P r*$. 

' The death of him whose deeds were noble was announced ; 
he left 
No peer ; the tidings wellnigh caused the mountains to 
be cleft.' 

Line n contains an Arabism, and hence was misunder- 
stood by the above-mentioned scholars, insi represents 
Arabic katala which means besides 'he killed', (i) 'he 
inspired with vehement love ' ; (a) ' he tempered wine with 
water '. 

The purport of the line then is they killed their lover 
or beloved (by inspiring him with love) with a goblet into 
which they poured wine (lit. the blood of the brother of 
'Aner, i.e. Eskol, a cluster of grapes) which was not killed; 
that is to say, was not mixed with water. He goes on : 
Then they breathed into him (or better read me) the spirit 
of God, and fanned the flame of my heart with the scent 
of perfume. This is, of course, a description of his 
vehement love. See above, p. 177. But in the midst 
of his intoxicating ecstasy the poet's heart forebodes evil, 
and he feels that all is vanity. He asks : 

:->bN3 'roS* nVrp insb-hd 
ino ^sdi p&n iptyn sw 

Luzzatto unnecessarily emends IN^TD to WOV. The poet 
then explains the cause of his grief, and says : For lovers 

59 Hatnasa, p. 369. 


when seeing men (i.e. openly) laugh, but in their tent they 
weep and cry bitterly. The unintelligible D^pxal is obviously 
a misprint for DjnNtrfl. This foot is -u-uu, and there 
is no need to say that the poet violated the rules of 

Another contemplative poem by Ibn Gabirol in this 
metre is the following : 70 

nbxStf vm »b>b3 nrvn 6 

v v = : • : - t : t 

rfeijJ tin nW rmn &6 

v v : t :*■ t • t 

f*)j[K neai [ring] 6p yx 

On reading this poem one cannot help being struck 
by the fact that the underlying idea is Imru'-l-Keis's 
famous lines : n 

- f rfO- 5 ~o £ «- o & o-- 

w- 5 * °'| ~ ° * . - » -^J - 

J-jfi..© A.S>**J . ( .k.-a>! l^.I.XJa 

- o s •» m*9 °-o - o .. o_c 9 of o** 

jJLi*! Jjj*JI xss»Jl d,Jj Jjj 

' Little wealth — which I seek not — would have sufficed me, 
Had my search been in life's lower plane ; 
But the object of my quest is highest glory, 
And my like high glory will attain.' 

The idea of the second line is to be found in the 
continuation of Ibn Gabirol's poem, line 17: 

n?8! bvtm bbm i^n by 
nSoln n^n by nete Btea 

70 Dukes' ed., p. 37. 7I Al-Fakhrl, ed. Derenbourg, p. 52. 


In line 6 b read rfana nnra. 
In line 8 a read nnnte instead of nnrva. 
Instead of Itya (line 9 b) which gives no sense, and does 
not scan, read some such word as "iyb. 72 

In line 1 1 a Tiatop gives no sense and no metre. Read 
w>n.i "aoj '■nb \rb r\m 

Line 12 a is obscure, but the metre will be restored, 
and perhaps the difficulty removed, if we read rUTO-isa 
instead of ncita, and BWl!> instead of )wrb. I wonder 
whether nolN does not mean here body. Cp. above, p. 175. 
The meaning would then be : If she does not make me 
master of her body, the earth will not destroy my love. 
HDIX would accordingly be used in two different senses, 
as is usually the case when a word is repeated twice in 
the same line. The following line would then be quite 
simple. The poet complains that he is not appreciated, 
and the loftiness which his soul attained is unknown to 
the outer world. His soul is despised (I take ''B'W, line 7 a, 
as object) by one whose good opinion he is especially 
anxious to acquire. What is meant by she is an open 
question. Instead of "\p» (line 15 b) we should certainly 
read nni£, which is more idiomatic and restores the metre ; 
rp"Y"i (line 16 a) should be rrn, and in the second hemistich 
ncN ( which gives no sense and does not scan, should be 
changed into vnDK. 

Al-Harizi has the following satire: 73 

nope* t6 rtDDitf ^ana m^a 

yisip nssD iab <a ny 

aim epD D\a It nyjj b 

yaib i!?"iri ni-inn ix 

72 See, however, Geiger, p. 131. 7S Tafykemoni, p. 171. 


The Arabian poets in their satires {Mja') often accuse 
their .enemies of niggardliness, which is considered as the 
greatest vice. Buhl(un) or lu'm{un) ' meanness ' is usually 
the contrary of kirama(twi) ' nobility ', ' liberality '. It is 
with them a favourite idea to say that if baseness or greed 
assumed human form, it would look like their enemy. 
Hassan b. Thabit in satirizing Jidam says : 74 

lisJlj j^fDlJ ,j.i.!l y\ J J>\ 

o - * - P o^o - o^. c ~s „ 

' Seest thou not that meanness, treachery, and guilt 
From Mu'ein to 'Aid— their habitation built.' 

On rare occasions the Hebrew poets allowed the short 
syllable to be elided in the last foot of both hemistichs. 

Ha-Nagid's famous ' stammering girl ' has this pecu- 
liarity : 75 

roa rmi i)>]) ■ax n»« 

tt n: ••< • : 
t f : t t V : v * 

Di3;i3 -riND-riK n»!> nrD3 

* t * v t t : t i * 

nyJpn iikd nes nby; ito 

This poem has been translated and explained by 
Kaempf, Geiger, Egers, and Lagarde. 76 Only the last- 
named got hold of the right sense. The other scholars 
attempted to read into these humoristic lines some ideas 
and allusions which are alien to them. I should, however, 
like to add a few remarks. The last line has not been 
adequately explained. Lagarde, with his characteristic 
contempt for anything Jewish, dismisses this line with the 

74 Dlwan, ed. Hirschfeld, p. 82, no. 190, 1. 1. 76 Sa'ar ha-Str, p. 33. 

76 See Mittheilungen, III, pp. 28 ff., where all translations are quoted. 


remark : 77 ' Schwerlich iibersetze ich njiD der Deutung dieses 
" Dichiers " gemass.' But it seems to me that ruiD in line 5 a 
represents the Arabic root saga ' was good, pleasant, con- 
venient'. Thus when the girl wanted to say surd ('be 
gone'), she said suga, which made the poet think of the 
Arabic word, and hence he hastened to her who was 
'fenced in like a lily'. For if ruiD should have the same 
meaning in both hemistichs, there could be no reason why 
the writer should have hastened to her. The joke must 
have suggested itself to ha-Nagid when he heard a foreigner 
talk Arabic. Gain and Ra are often confused by Europeans. 
The French r sounds like the Arabic Gain, and Arabs tell 
many an anecdote similar to ha-Nagid's. 

(c) Another kind of this metre consists in having the 
third foot of both hemistichs reduced to a spondee or 
anapaest. Hassan b. Thabit has a few kasldas in this 
metre : 78 


- e~c --• - * *5 - 0-0 * * - 

' I swore your discourse never to forget 
As he who thirsts thinks of the taste of wine.' 

In this form the Hebrew poets had still greater facility 
in forming their lines, and it is frequently met with. 
Ha-Levi addressed the following lines to a friend : 79 

ns^jp rijn'ij lira nobfe> 

ns^ri nn dki ng"n 
nastac ten ciy£>2 nn 

" Op. at, p. 33. ' s Diwan, p. n, no. 8, 1. 25. ' 9 Brody, I, 181. 


Al-Harizi in his 'Andk has the following epigram : 80 

7f»fe> nan \-\& ^'ya naty 

DVna ifixV ntm b$& 

atnn &61 raa ro3E>n tk 

dV> tb &6} ^!? niaianpits 

It should be observed that nmann in the last line has 
a meaning which is not quite identical with that of the 
Bible. It does not denote perversities, but vicissitudes, 
and stands for Arabic suruf(un) ' vicissitudes, turnings ', 
which occurs often followed by dahr ' time '. 

Ha-Nagid has a long poem in this metre : 81 

r6p -ib>n ai$ rpb atnn 

$>sc ^ tt; a'"n n? 

This mode of Kamil has been given by Jewish writers 
as Sari'. There is no real objection to this explanation, 
and it is hard to say which metre the poet had in mind, 
since the contracted Sari' would have the same appearance. 

(d) What is known as the shorter Kamil, is a metre 
consisting of four feet instead of six in each line. The 
rhyme-bearing hemistich has an extra long syllable added 
to it. The last foot is then said to be muraffal ' having 
a train '. Its form is : 

_|_ w _S=!^|_ w _i^||_ w -Mi=i|_ w _W 

The following lines are attributed to Sureih b. 'Imran: 82 

80 Brody's ed., no. 222. 81 Harkavy's ed, p. 20. 

82 NSldeke, Delectus, p. 4. 


" * - - ~ 

-- 51 B> «t-0 ■* O ^ 

jL^jJLJl ~«J| w Vj-^-J 

' Let noble people be your friends ; 

When you sought not their love in vain, 
Drink from their cup, and do not fear, 
Though deadly poison it contain.' 

Ha-Levi has the following poem : 83 
crayr\ -in *ib tibti 

' t s t t - ' ; t 

e^k nna?o s|dm ija 
onap -ran ^a w 

Brody 84 classes this metre among the unbestimmle ! 
For a fifth probable kind of this metre, see below, 
p. 230. 

4. Wafir, ample. 

This metre may be called a companion to Kamil. Its 
feet are built on the same principle, that is to say, they 
must have three long syllables and a short one. The order 
of the syllables, however, is different. It is the Epitritus 

primus ( J)> but for the third syllable two short ones 

may be substituted. A hemistich is composed of such 
three feet, and is always catalectic. 

Its form is 

__ w |_J=A = ;_^|_^_ w ||__ w |_!^i_ w |_! = A =i _ w 
83 Brody's ed., II, p. 159. ei Metro, p. 49 d. 


The Arabian poets employed it very frequently, and 
their Hebrew followers have a special predilection for it. 
Obviously the fact that the short syllable is to be at the 
beginning of the foot makes it an easy task for the Hebrew 
poet. For it is usually possible to add 1 or one of the 
letters of a"bs2 whenever a swd mobile is required. The 
greater bulk of ha-Nagid's poems are in this metre, and it 
is quite a favourite with most poets. Brody and Rosin, 
however, exclude Wafir entirely, and call this metre Hazaj. 
They follow Sa'adya b. Dannan in this respect. Kaempf 
is the only one who considers it possible to call this metre 
Wafir, but even he seems to be undecided. A comparison 
of the Hazaj and Wafir in Arabic will make it certain that 
the Hebrew metre with which we are now dealing is the 
latter. To begin with, the Hazaj never occurs with six 
feet. The Arab grammarians with their harmonizing 
tendency have suggested that in theory six feet may be 
allowed, but in practice only four should be employed. 
But no mention is made of this imaginary form being 
catalectic, a form which it always assumes in Hebrew. 
Moreover, the Hazaj in any form is so rare in Arabic that 
many Diwans do not contain it at all, and this fact in itself 
ought to be sufficient to reject the hypothesis that the 
Hebrew poets employed it exceedingly frequently. On 
the other hand Wafir is composed of six feet, and is always 
catalectic, as in Hebrew. It is one of the most dignified 
metres in Arabic, and Hebrew poets were right in intro- 
ducing it into their language. Here again, as in the case 
of Kamil, in many instances a light vowel with its hwa 
mobile will be regarded as two short syllables. 


'Amr b. Kulthum's Mitallaka is in this metre : 85 

- — - M---0 - O - 

•» o '■" -» -"O^ Out •* $ 

' Thou who art departing, stand awhile and wait ! 
We'll converse in truth before we separate.' 

The famous poem of Joseph b. Hasdai is in this metre: 86 
noxyi ;ta rnoa in •oxbn 

now nna )ij?p -cno ^a 

Goldziher in his brilliant article in JQR., XIV, 734, 
has sufficiently explained this figure of speech of herding 
the stars. He quotes, among others, a poem by Hassan b. 
Thabit's sister, in which she describes her inability to sleep 
owing to some grief. But Hassan b. Thabit himself has a 
poem which, I think, will most appropriately illustrate this 
metaphor : 87 

\jy*l ^1 J-£-»J <Si]y>> piJ 

W**J {_^- s ~ py*\ <^iy 

y u\ 

^sp> ixr cJ>p 1.4I. yk 1S1 

l-*JaJ j.^***-" J^*\ , c*r*' c 


Lyall, p 

1. no, 1. 9. 
p. 27. 

86 Sa'a 


'While at Hamman my night seemed long without an end, 
As though the foremost stars to set did not intend ; 
Until they disappeared I tended heaven's hordes — 
Till then no sleep for me ! — as though they were my wards ; 
And when star after star was vanishing from sight 
My eye kept vigil on the last star of the night.' 

It should be added that rcCa in Arabic has the meaning 
of tending the flock (as in Hebrew) as well as watching in 
general. The expression rcta-n-nnjuma has been taken 
literally by Lane. Other scholars take it as a figure of 
speech in which the stars are represented as sheep. 88 
Whatever the explanation is, the meaning is quite clear. 
The Arabs were so familiar with this metaphor that a 
contemporary of Harun al-Rasid (Muslim b. al-Walid) 
says : 89 

^a3i»j ^£JJL«9 cj]j-*' jjLeJO 

' If my feelings were like yours I should not spend the night 
A companion of sadness, comrade to the stars.' 

Ha-Nagid has the following poem : 90 

njina iww i^nd Troth 

T * ** s * v ■*: •* T T ; 

rryuD b*fr\ ybxo »bii 

Line 7 of this poem requires explanation. It is 
n^nsi two rw?? win nto 
nynp ty ^iaa ry rm 

Brody in his notes to ha-Levi's Diwan (I, p. 164) quotes 
numerous passages from other poets who made use of this 

88 See Noldeke, Delectus, p. 94, note to lines 2, 3. 

89 Diwan, p. 148, 1. 6. 9 ° Brody's ed., p. 108. 


expression. But the explanation remains incomplete till 
we know the purport of this metaphor in Arabic, whence it 
was directly borrowed. As is well known, it is the custom 
in the Orient to insert stibium into the eyes. When a person 
spends the night without sleep he is said to have inserted 
night's stibium, that is to say, the blackness of the night, 
into his eyes. Both the Lisan al-'Arab 91 and Taj al- 
'Arus 92 attribute the following statement to Abu Amr : 

!ju»jl J-JUl J-S-SO ^J^-S iLlx jl IjjLu 8.1J j-b~*> J-ay-1 Jl-fi~> 

Jl*J^)IS SwkJwoJ J-*JJ1 ^U-"> J-KJSS9 J§"'> CS' 

It is said of a man who spends the night without sleep, 
travelling or working, he made the night his stibium, that 
is to say, he zvatches, and puts the blackness of night into his 
eyes like stibium. 

Ibn Gabirol also has many poems in this metre : os 

linos if< sriix ruwrni 

in«o vrafa; bib -1^3 

In Dukes' edition the metre is in some places corrupt, 
and Egers 9 * has restored the text. In line ia b I recognize 
in VVJ] an Arabism which is very frequent in mediaeval 
Hebrew poetry. Whether Hebrew Tt is to be connected 
with Arabic durra{tun) is a matter open to dispute, and 
philologists are not unanimous on this point. But it is quite 
certain that in the Bible "n denotes a kind of stone, whereas 
here it obviously means a pearl = Arabic durra{tun). Ibn 

91 IV, p. 75- M II, p. 312. 

83 Dukes' ed., p. 22. u Zunz, Jubehchrift, p. 197. 


Gabirol's line reminds us of Al-Hariri's grammatical treatise 
durratu-l-gawwas ' the Pearl of the Diver '. 

Line 37 has been misunderstood by Egers, who was 
therefore driven to emend it violently : 

133^ W ^JflO TDN1 

Dukes has rightly pointed out that the poet has 1 Kings 
20. 41 in mind. We should accordingly translate : / shall 
remove the coverings from my hearts (i.e. perhaps, mind's) 
eyes when nights covering is suspended. See the preceding 
line and Goldziher's explanation of it in JQR., XIV, 720. 
I have rendered "ibn by covering in accordance with the 
opinion of the best authorities. It is, however, likely that 
Ibn Gabirol took it to be identical with "1BK ashes. Should 
this view be right we should have here the metaphor of ?y JflB 
which was explained above. A similar expression occurs 
in the Dlwan of Muslim b. al-Walid : 9S 

■ o »-o 

L»3l.) , 'J^'j- ^.SNi-Jl . -^l-JD i£jL*_9 

» Jp. O_o o «- o - 

' He was communing with the stars until 
The ashy night deprived his eyes of sleep.' 

Egers suggests the following reading : 

After my explanation of the text there is no need to 
comment on this emendation. 

95 Dlwan, p. 59, 1. 3. 

P 3 


Ha-Levi has the following poem : 9C 

Here nnjtt prpbably represents Arabic sibain) (see above, 
p. 1 88). About Q»j6p (line 14 a) it should be remarked that 
although rwaon 'yi'p occurs in Mishnic Hebrew in the sense 
of 'sails', it is doubtful whether ha-Levi would have used 
it in that signification without ru%D, had he not been 
influenced by Arabic kiru(ri). 

Moses Ibn Ezra in his Tarsi's has the following lines : 9T 

sidmi innaina nSaj vr 

*t : * ; J - ; t • ■ •; 

nasi at; K>iJN nawrt 
f)D33i vbN ikd rbn nnb 

•t:*: t« : tt : 

The ' quarreller ' here is 'adiliiiti) ' reviler ', who plays 
such an important part in Arabic love poems. For ma air 
compare the line of Simma b. 'Abdullah : 98 

\~~ '- °*\ - ' ° T ° Z~ \~ 

' When I recall the days which at Hima I spent 
I turn unto my liver, fearing it be rent.' 

Al-Harizi in his Tahkemoni has the following epigram: " 

nam jit^a Tiao •jj'-d; 
naisn avifo tna Tibi 

T t: - * v; TT »T ; 

Qi-iiK* pan i£?k B'si *|i> 
nawn Bten »an|» kw 

86 Harkavy's ed., I, p. 28. 9 ' p. 61. 

9S Hamasa, p. 540 ; Delectus, p. 16. S9 186. 


t *: - ' tt- - •• t : 

can? •q-j'in ■oa tap '•n* 

Goldziher in his article referred to above 100 fully ex- 
plained the figure of speech in which a liberal man is com- 
pared to the rain and rivers. It should be added that in 
Arabic almost all words denoting moisture signify liberality. 
One need only mention jada ' was kind, liberal ', ' rained 
abundantly ' ; nada(n) ' moisture ', ' liberality ' ; gamr{un) 
' deep ', ' copious rain ', ' liberal, generous ', especially in the 
expression gamru-l-hulki. 

tot? in line 3 a is an Arabism. It corresponds to Arabic 

sara = ' he returned, became '. Biblical Hebrew would have 

required Vn followed by b, and this would have disturbed 

the metre. 

5. Haflf, light. 

This is a compound metre composed of failatun and 

mustafilun in the following way : 

v^w — w — w w 


Keis b. Darlh says : 

19 vt- Q*r Of O , , „ 

JZynS (C-i^ l -j fc -J^ •*•* ^J^i 

' I spent the night with care as the companion of my bed, 
And since you from my side departed copious tears I shed.' 

Ha-Nagid has the following poem : 102 
miaa sfsbo na!» nj nai 

t: *t: -; t j - 

-rfo *b\ nnnsta ruETi 

: t : • *t:t - T" j- 

rmn pn &6a ik'e> wvai) niJ 

100 70-ff -, XIV, 724. 1M NOldeke, Delectus, p. 6. loz Harkavy's ed. , p. 7. 


Ha-Levi's poem which Brody quotes 103 should be thus 

vocalized : , 

nna *MB , n; annf i? 

sjeto* nam rpa psben 

• *: t : t -T-:- 

cisi-1*' naa|> »efc runes tn 

'•• : tt - * '; t : t : t 

nnuD &y fy mna nV 
pjaiy* $g ng "ib>k i>K wi 

Al-Harizi in his Mahberet Itiel has this poem : 104 
r? njn ntoa -ibj£ "i>ttfe> 
n?nx oinea r6y» "sba tib 

T -.,-: t ; t : - • * : *t 

mini $«a niDnJ> rrny Jo 

The metre and rhyme of the original is imitated not 
unsuccessfully. In line i b I emended mnx of Chenery's 
edition into WIN on account of the rhyme and sense. 

Moses Ibn Ezra also has a few lines in this metre: 105 
niae Tfttaa ?!>bm "vaan 
n»3iy omsa oeh d1< 

* t=; --tit: t : 

Here also QBh = Arabic j5r<2. See above, p. 305. 

Abraham Ibn Ezra has the following line : 106 

nayb iw iba bo Jinn as 

*:t; •: - : t * -1 -~ 

♦■none ma ncrx na^na 

J" 3 Metro, p. 39. 1M p. 87. 

106 Tarsi's, p. 21. 10li Egers's ed., p. 6. 


6. Sari', swift. 

(a) This metre, which occurs now and again in Arabic, 
is composed of two feet of the mustaf'ilun kind with its 
incidental variations, followed by — u — in each hemistich. 
Its normal form is 

— \y — j — ^ — ^ 1 — w — w[j — \j — !~ w — ^ — ^ — ^ 

Abu-l-'Atahiya has the following line : 107 

- - <J>- U S, J - " 

— *a~o t)~~ 9 *» — * so* 
LaiJI 8wJuC 5 Li L>. L..I 

' A man resolves to act, but fate 
All his intentions will frustrate.' 

It is not frequent in Hebrew, but ha-Levi has the 
following lines in this metre : 103 

mj«nni> rby *6i rb&3 

t-j- : • : t ; t : t : t 

Al-Harizi in his Mahberet Itiel has this poem : 109 

yh m$\ b±> 'ja u«a 
anK" pai !>afe> '3a f>a af> 

b»D pojfl tfjU TIN 'ITTEi 

a'jtg: va% aa rante to 

TIS (line % a) stands for Arabic aha, which is often used 
in the sense of possessor. This usage occurs quite frequently 
in this literature, especially in the phrase 1D1D TIN. 

(b) The Arabs allow sometimes, for reasons which were 

107 Diwan, Beyrout ed., p. 4. los Harkavy's ed., p. 60. 

109 p. 60. 


explained above, 110 the rhyme-bearing foot to be reduced 
to a spondee. The form of the metre is then 

— — — \J — w — W — C/ i — \J — — \J — w — KJ — O 

. Ill 

*0 *•- B> o-c 

AbQ-l-'Atahiya says : ] 

' He who seeks glory shall remain in it ; 
And lo ! man's glory is his piety.' 

Al-Harizi in Mahberet Itiel says : U2 
»js'y niy nrron H»b 

..... -" T ; ... 

For another possible variation of Sari' see above, p. 197. 

7. Matakarib, tripping. 
(a) This is a simple metre composed oifdulun repeated 
four times in each hemistich. Its normal form is 

Muslim b. al-Walid has this line : 113 

' My friends, I see in love no cause for shame, 
I cast off all restraint, withhold your blame ! ' 

110 See above, p. 166. m Dtwan, p. 12. 

112 p. 60. lls Dtwan, p. 151. 


Ha-Levi has this poem composed on a wedding : 114 

D'wa fr$\ niEnn w6o 

wiping ^jpa rrax '•ONnfj 

ar^x? -roni a'O'HJ ixji) 

awju n ? ^?? ^rans a 1 ? foi> 

The Hebrew poets in employing this metre were fond 
of dividing the hemistichs into two equal parts, usually 
rhyming with each other. Instances may be found in 
ha-Levi's Dlwan (Brody's edition, II, p. 184), and in the 
following poem by al-Harizi : us 

a^no nab? a^n a^na 

B'jrtrn arrhj? btysrfn iw 

As in the case of Tawll, the Arabian poets allow the 
first short syllable of the first line to be omitted. 

Al-Hutei'a has this poem : 116 

_ .JLvJI i'ljuC i>^9 ^jI -Ja-cl 

- * „ - - - o -~o~o ro ^ 

' Ibn Kurt gave munificent gifts 
On the morning we met at Suleim.' 

Moses Ibn Ezra made use of this privilege in this 

verse: 117 

inon pbna anon pbn 

Wxp Dp}a a^inr? oat# 
vbja navTfi v^a pa y;3i 
^*y? n^j nnwn a: 151 

114 Brody's ed., II, p. 14. «« Tafrkemoni, p. 15. 

116 Dlwan, ed. Goldziher, p. 181. "' TariW, p. 25. 


(b) Very often this metre is catalectic, especially in the 
rhymeless hemistich. The ridf lli usually causes the second 
hemistich to remain acatalectic. 

Hassan Ibn Thabit glorifies his clan in this metre : 119 
o'-o' »-/- , -» - 

o-o-.- p / * ' n ^ 

' We inherited their dwelling-places 
After them, and ceased not to be masters.' 

In Hebrew this metre is rare. Ha- Levi has also this 
variation : 120 

8. Ramal, running. 

{a) This metre consists of the foot faildtun repeated 
three times in each hemistich. It is usually catalectic, and 
its normal form is 

— \j \j ^j £7 I u v — w C7 I ^J O u u 

Abu-l-'Atahiya has this line : m 

° > ' °-° O — 0-} ' , O - O - 

' Whoever lives grows old, and he whose hair is hoary 
Will die ; and fate cares not whom it is overtaking.' 

Ha-Levi expressed his trust in God in this metre : m 
nn>n» r\b w '•n !>k *£ ^s 

nDnn ia ddni dto dn 

118 See above, p. 166. 119 Dlwan, p. 12, 1. 20. ]M Harkavy, II, p. 5. 
121 Dlwan, p. 39. I22 Harkavy, II, p. 147. 


Abraham Ibn Ezra has this line : 123 

nyyhnx tfy\ naa rasety 

(b) Sometimes this metre is of two feet in each hemistich, 
and is then 

Abu-1-Atahiya says : Ui 

' O slave, how long will you barter rectitude for error ? ' 
Ha- Levi has a piyyut in this metre : 125 

T$8) l^PI ^ iW ^09 '? nag: 

Abraham Ibn Ezra has this prayer : 120 

ynym jik jifoe>j> bs pa; ^nni3? ^ 

9. Munsarih,^^;?^-. 

This is a compound metre closely resembling Baslt. 
Its normal form is 

u w — w — wj — J — w — v7j — Wwl — *-» — C7j — I — u — C7 

Muslim b. al-Walld says : 127 

'JJo- * o -o-g o * - - - 

SJJuil £}-*-*J U irfjj J~J 

' My sleep has fled, my eye kept on inviting it, 
Out of anxiety, but sleeplessness did visit me.' 

12s Egers's ed., p. i. 124 Diwan, p. 86, 1. 10. m Harkavy, II, p. 5. 
126 Egers's ed., p. 50. m Dlwan, p. ai8, 1. 4. 


This metre hardly ever occurs in Hebrew. Ha-Levi, 
however, has a few poems in it : 128 

'f'nanrn •>:?£> }>j5 '•atJ'Wi np 

Brody, who classes this metre among the unbestimmte, 
thinks it possible that ha-Levi was its 'inventor'. 129 Sa'adya 
b. Dannan gives Munsarih as one of the four metres which 
do not exist in Hebrew. 130 At the end follows the contra- 
dictory remark : ' I saw some poems in this metre in our 
literature.' However, Sa'adya's Munsarih does not agree 
with that employed by the Arabs. 

10. Hazaj, trilling. 
This metre is antispastic, and consists of two feet in each 
hemistich. Its normal form is 

. _ W \ X. \J 

Al-Find says : 131 

' There may be safety in evil, when doing good brings no 

In Hebrew this metre is not rarely met with. Ha-Nagid 
has the following dirge : 132 

? njovw -i3j3iDi n^a nisan 

njiajn axab nv "b no itrjpa hh> 

Brody in his notes to ha-Nagid's Diwan remarks that 
n^B denotes device. But such a signification is unwar- 
ranted. It is likely that njpyS represents Arabic kada{un)^ 

" 8 Harkavy, I, p. 57. m Metra, p. 48 c. 

130 Meleket ha-Str, p. 17. 131 Hamasa, p. u. 132 Brody's ed., p. 1. 


which means decision, mastery, power. We should therefore 
translate : Is there any poiver against death ? 

Ibn Gabirol has a love poem in a light vein in this 
metre : 133 

sxn bs\ sbn i>32 nsp ^an -nan 

.... T . T T . v t • ; * * ~ ; 

nnnw tan? wrn ^nte yjnp 

This is the ordinary description one meets in most 

poems of this type in which the poet's hope is realized. 

Line 7 

nxy? D5JD wjM-n D-nn? in 1 ; nafarn 

is almost a literal translation of Muslim b. al-Walld's line : 131 

' The eyes were closed (lit. quiet) and every watchman slept.' 
Moses Ibn Ezra in his Tarsis has these lines : 185 

: ;•• s "J * T ; 

epx as?D3 \cm "\f$ 

The expression D"i3 r)3 as a designation for wine is 
frequent in Arabic. Muslim b. al-Walld says : 130 

i^jiLo *r»J! cuUj ^ Ij^jj 

' Pure wine from the daughters of the vineyard.' 

Ha-Levi has the following lines addressed to a friend : 1ST 

vb]3 -\m roan pnK 

vbrrtro vfapo fiat? 

183 Dukes' ed., p. 55. I34 Dftvdn, p. 149, no. 23, I. 26 b. 

135 P- 34- 1S6 Diwan, p. 178, no. 32, 1. 19 a. 

137 Brody's ed., I, p. 38. 


ii. Mujtatt, compntated. 

This metre does not occur in the Dlwans of the earlier 
poets, and some scholars are of opinion that al-Halll in- 
vented it. It is a compound metre consisting of Mustaf'ilun 
and fa ilatun. Its form is 

Ww! ^ W J| — ~ W W [ W W 

The following line is quoted in al-Fakhrl : 13S 

a s - - - -■ 2 » ° - - ~ o of 

' Go and tell the Caliph : be not rash, 
Something which you like not comes to you.' 

Hebrew poets employed this metre now and again. 
Ha-Nagid has this epigram : 139 

one* & dvie* &6 cnrn »>k nb»j>: m 
cinx onrn rriatyb nnob Ijvd? hs> 

Ibn Gabirol's famous poem, written on his leaving 
Saragossa, is in this metre : 140 

■-aha kj>2 iru 

There are in the printed edition a few corruptions which 
should be rectified. In line a a read ^b. \Mt (in line 13 b) 
should be changed perhaps to *}$] (cp. Ps. 25. 16) or , #j>. 
Instead of "bao (line 17 a) read *«)»|. Read OrojN (line 28 a) 
instead of QrMfK. Line 31 is corrupt. It should perhaps 
be read 

138 Derenbourg's ed., p. 64. 139 Harkavy, p. 128. 

i« Dukes, p. 1. 


MM5» (line 33 a) should be Wty]. In line 35 b transfer 1 of 
■•com to »Jt? in accordance with Hebrew idiom and metre. 
Insert ns before ^isn (line 37 b). In line 42 insert y before 
'Ja, and n's after ii3B>. Instead of S>» (line 43) read ".bg. 
Change ifo (line 44 a) into n|Tg. Delete 3 of D1SD3 (line 50 b). 

Al-Harizi has in his Tahkemoni the following lines: 141 

ded) nw vina lym »y^a an ?jj 
odci iini nbe* "sail e>K bs ixfiP 

•• j ~: t : ™ ; • ■ t : ■• r ~: ; 't 

The following poem by ha- Levi is interesting, as it shows 
the optional use of the long syllable at the beginning of the 
line: 142 

1^53 -i$b n*n riK t&n yhx pc; 

nab nipixo -itaa ai»ro y^tsn ?pn 

12. Madid, extended, 
(a) This metre is built on the same principle as Tawll 
and Basit, but occurs very rarely. It is compound, and 
consists olfailatun, and fa ilun occurring alternately. Its 
form when of four feet is 

-uul w w I — uu| <_> C^ II — uu J u v|- w O j \-iX3 

A woman bewailing the death of her son says : 143 
eLL> eLi-c v-^-*° '^^ ic-A* *z*<J 

'Would that my heart controlled its grief; would in your 
My soul as victim to relentless fate were led.' 

141 p. 51. " 2 Harkavy, I, p. 56. 

1<s Hamasa, p. 415. 


Hebrew poets had little reason to employ this metre ; it 
is hard to wield, and since it was not popular with the 
masters, it had no attraction for the followers. Ha-Levi has 
the following piyyut, 144 which may be taken for a Madid. 
Ftfilun, as is sometimes the case with Basit, was reduced to 
a spondee, and the metre then became 

— J w w | 1 — w *-'|| — I w w | — | ^ w 

rvro: ^a Dy w?t: ^jgE'.'; 

(b) More frequently this metre is of three feet. It is then 

w w J — w w [ w w I] w w — w w | w w 

Abu-l-'Atahiya has the following line : 145 

' Lo, you are dwelling in fate's vale, 
When death strikes you, it will not fail.' 

Abraham Ibn Ezra, who seems to have a special fondness 
for rare metres, has the following piyyut : 146 
niiini) »a nnnax »s nnofe>K 

: • t : : v * t ■ : v 

Ha-Levi's following poem seems to be in this metre. 147 
The concluding line of each strophe, however, is quite 

pg^iab onto -jn -peas penni> n^yri i$ \n n^y; 

111 Brody, III, p. 13. i« Di w „n, p. 28, 1. 9. 

118 Egers's ed., p. 20. "7 Brody's ed., II, p. 320. 


Modern Jewish writers on prosody make no mention of 
this metre in either form. Even Rosin, 148 who is supposed 
to give an exhaustive account of Ibn Ezra's metres, makes 
no allusion to this one. Sa'adya b. Dannan, however, after 
enumerating twelve metres which Hebrew poets employed, 
gives four more which, he asserts, are only to be found in 
Arabic poetry. 149 Madid is one of the four, and is quite 
accurate. It is the second kind, and is as follows : 150 

TB* 1J3 n\3| JHK "3313 

rrynan 1V33 javi •gety 

After this follows the statement : I have seen very few 
poems in this metre in Hebrew. This flatly contradicts the 
assertion of the preceding page, and can hardly be right. 
Perhaps he meant to say in Arabic. 

13. Mutadarik, continuous, or Mutadarak, supplied. 
This metre is rare in Arabic. Its normal form is 

— ^ CM — ^v^j — u u — w CM — ww! — w C< — uu J — wC? 

In Hebrew it is certainly rarer than in Arabic. Kaempf m 
quotes the following line by Ibn Gabirol : 

1153J 13s •>3 ipton npy]« 
nora ^np ^ii rpy$ 

The Arabs allow some of the feet to be reduced to 
spondees. In this secondary form ha-Levi has a piyyut, 152 

lie Reime und Gedichte des Abraham ibn Ezra, 1885. 
I<9 Meleket ha-Sir, p. 15. "« Op. at., p. 16. 

551 Die ersten Makamen, p. 44. I62 Harkavy, II, p. 128. 



in which failun and fdlun occur alternately. It may, 
therefore, be regarded as a Matadarik. 

Svrp_ 3iD3 jr'N 5>3 xrnn? 
\mp x DBn Via} nim i>K 

This piyyut may also be regarded as a secondary Haflf, 
in which the last foot was shortened. 

14. Rajaz, trembling. 

I place this metre last, not because it is the least 
frequent, but because of its peculiar character. It hardly 
possesses all the features of the other metres, which show 
a high stage of development. There is no doubt that 
it forms the connecting-link between rhymed prose (saf) 
and the other metres. It is chiefly employed, as was 
remarked above, 153 in extempore lines. The only stipu- 
lation of this metre seems to be that the lines should 
be approximately equal, and this differentiates it from 
rhymed prose. Its feet are composed of four syllables, 
and almost every possible combination is permitted. It 
is also catalectic. As a rule the hemistichs rhyme with 
each other. Furthermore, a poet is even allowed to have 
a hemistich of three feet and one of two feet in the same 
poem. This is best illustrated by the following lines by 
one of al-Find's daughters, who was inciting her clan to 
waee war : 154 

' War, war, war, war ! the fire was kindled and it raged ! 
The mounts were filled with it ! 
How fair in midday splendour are the shaven heads ! ' 

168 p. 185. 1M Noldeke, Delectus, .p. 47. 

Lcj Lc. Lc« 

, S » - > - 


In Hebrew to this metre may be assigned all those 
poems which have approximately equal lines, not exceeding 
twelve syllables in each hemistich, and do not fit in any 
other metre. In Egers's edition of Ibn Ezra's Dlwan there 
is a poem which bears the superscription nmxt bap) he 
composed a poem in the Rajaz form. Egers remarks 
that it is incorrect, and refers to Kaempf. Of course 
it does not agree with Kaempf 's Rajaz. But then 
Kaempf's Rajaz is really a Kamil, 155 and Ibn Ezra's poem 
is perhaps as good a Rajaz as any other. It is as 
follows: 156 

»3ofc> Vnpn »n hv6 rnlN 
a«ra ngbrj "6nD wmo 

Ha-Nagid has the following epigram which is cata- 
lectic : 157 

■6 an?*n neta nag 

nn^n iah na/in 

nbto '3K DJJ3 

Perhaps to the same metre belongs the following piyyut 
by ha-Levi : 158 

nix ntnj "risu b$ ^pjista -irr 

i]?»n t]!?n -ib>k ayn 
^e%; In njj Viafe> 

ri^a son iapj? w 
iix TbK» s -iia 5>y ns aha l^y 

155 See above, p. 185. 166 p. 6. 

157 Harkavy's ed., p. 150. 158 Harkavy, II, p. 9. 

Q 2 


We thus have an example of dimeter and trimeter 
hemistichs in one and the same poem. 

There is another class of poems which could be assigned 
to the shorter Kamil without the muraffal, but may also 
belong to the Rajaz. That variation of Kamil is as 
follows : 

Hassan b. Thabit praising his family says : 159 

o~ /* 0* o ,0-0 «• t - * t o 1 o~ 

iSXMj ^OJ XSSS*J I ^J mm. 4-l.W Ul (JB ».g,B,4.1l 

' Who provide food when the years of drought afflict the 
land ' (literally : become fixed). 

Owing to the rarity of this metre in Arabic I hesitated 
in assigning the Hebrew poems to it, and am prepared 
to yield to the opinion of other scholars to consider them 
as Rajaz, in spite of their uniformity. Ibn Gabirol has a 
long poem in this metre : 160 

meo ii^os noft •otfri mini tb-nn 

Moses Ibn Ezra in Tarsls says : 161 

i?J5 Dipt? c«ai> tj-ity ay!> t]rtan.3 *n.ty 

izb qidVt Dm K133 tos canto -lire nm 

1 • •• ; • t : • t ' 1 t - 

The fourteen metres enumerated above are directly 
borrowed from the Arabs, and practically cover the field 
of Arabic prosody. The remaining two metres mentioned 
by Arabian grammarians are Mudari' and Muktadab. 
They are extremely rare in Arabic, and as far as now have 

159 Dizvan, ?. 77, no. 176, 1. 5. «° Dukes' ed., p. 4. «' p. 16. 


not been found prior to the grammarian al-Halll, who is 
the first to name them. It is the opinion of some scholars 
that these two metres together with Mujtatt were invented 
by that grammarian. 102 Sa'adya b. Dannan gives a Mudari', 
which, however, does not correspond to the Arabic. The 
normal form of the latter is 

_|_„_i=;|-_ w ||_|_ w _w|-_ v _, 

But Sa'adya's line is as follows : 163 

i?D3 t rmai bwo byibz 
ih by tif vht psi *nK p^nm 

Brody agrees with Sa'adya, and considers the following 
riddle as a Mudari' : 164 

a:nn m Dan ^nn m -jan 

aaitf ibny \bj> nft nv sxpni 

It is, however, quite obvious that this is no Mudari' 
which in Hebrew would have usually been, if we divided 
the feet differently : 

^ w 

If a poem whose syllables are thus disposed be found, 
it would be a Mudari', but those given above cannot claim 
to be recognized as belonging to this metre. 

The metre given by Sa'adya as Muktadab on p. 16 
does not resemble the one bearing that name in Arabic. 

Apart from these fourteen metres the Hebrew poets 
employed other combinations of short and long syllables. 
There is no need to give an exhaustive account of all of 
the variations. The Dlwans of ha-Levi and Abraham Ibn 

162 See Wright, Arabic Grammar, II, p. 368, Rem. 

163 Meleket ha-Su; p. 13. ls4 Metra, p. 40. 


Ezra abound in instances. Some of these combinations 
closely resemble the regular metres. It should be noted 
that also the later Arabian poets have employed similar 
metres, and it is hard to say whether the Hebrew poets 
were the inventors of these combinations or followed Arabic 
models. This is, however, a minor point, since those 
variations were not sanctioned by the grammarians. 

A few of the most typical of these variations may be 
given here. 

(a) The following 165 is a combination of Madid and 
Baslt : 

■ V V — V ■ 

najjl "W 'nsii i^np m nay: 

{b) A Tawll with the second fciidun omitted may be 
recognized in the following prayer by ha-Levi : 10C 

mpn tdk niir *ia ditv 

The concluding line of each strophe is half a Tawll : 

t t : • - : 

As given in Harkavy's text there are numerous corrup- 
tions, but in his notes m he quotes variations which give 
better sense, and scan accurately. Thus instead of H133 
(line 17 a) which is substantiated by the rhyme and sense, 
his text has roj. In line a a vocalize ntoni? (cp. Isa. 30. 18). 

Abraham Ibn Ezra also has a prayer in a similar 
metre : 1CS 

106 Ha-Levi, ed. Harkavy, II, p. 6. lc0 Op. cit., p. 19. 

167 II, p. 170. M« Egers's ed., p. 56. 


(c) In a similar metre with the addition of an iamb 
at the end of the hemistich is the following contemplative 
poem by ha-Levi : 1M 

raatyn •>n»b rvnb? pra rue* 

(<3?) Ha-Levi also has a shortened Baslt : 170 

(*) Ha-Nagid has a peculiar kind of Baslt in which the 
mustafilun feet succeed each other without interruption : in 

ara>j ^d? tnpn i>b?? -iste ins 
etaV 1 sna inpn h\m\\ -inp 

(/) A peculiar kind of the shorter is the follow- 
ing in which the second foot of the first hemistich has 
a muraffal : 172 

nixi »tt rnx 
(g-) A shortened Hafif is ha-Levi's piyyut : 173 

haj \Wa by 
QiiT 'nxsa am caaiD 

(/«) A lengthened Munsarih is the following piyyut by 

ha-Levi • 

■oWy r6« nay yjn; n? 

109 Harkavy, II, p. 149. 17 ° Ibid.,11, p. 126. ln Sa'arha-Slr, p. 33. 
"2 Harkavy, II, p. 27. " 3 C»A at., p. 18. "< /4»</., p. 118. 


In the same metre is the poem beginning ^ TT'''. 175 
(z) Abraham Ibn Ezra has a shortened Munsarih : 176 
nana Jjbd nam n»n 

t*: t • t-;t~ 

na>b& nnjn nms ao 

t • : t< - t t-: 'v ■ 

See also ha-Levi's poem beginning D3^l ^ DD. 177 

In addition to these metres there exist some in which 
the moving swas are entirely neglected. In some cases 
the poets skilfully avoided those swas. Sa'adya b. Dannan 
calls it the vowel metre ("JFUFin iru). A good deal of the 
poems composed in this kind of metre have seven syllables 
in each hemistich. This remarkably coincides with the 
Syriac metre. For the majority of the Homilies (memre) 
of the Syriac writers have seven vowels in each hemistich. 
This, however, does not signify that the Syrians influenced 
the Hebrew poets. Syriac poetry is chiefly christological, 
and its dull, unattractive tone would scarcely appeal to the 
keen imagination of the Hebrew poets. 

There is no need to quote examples of this metre. 
Every reader can ascertain it for himself. 

In conclusion, I should like to point out that editors 
in publishing Dlwans would do well to divide the hemistichs 
into feet. The custom prevailing now is to prefix a row 
of curved and straight lines which make the reader no 
wiser, since those lines can be supplied by himself without 

175 Ibid., p. 15. 176 Egers's ed., p. 24. 177 Brady's ed., II, p. 25c.