THE SCRIPT AND TEXT OF IBN QUZMAN'S DIWAN: SOME GIVEAWAY SECRETS

Depending wholly on internal evidence, this study sets out to show that the unique manuscript of Ibn Quzman's Dfwdn, published in facsimile by David de Gunzburg in 1896, is a dictated copy. It tries to show also that the copyist, who has often been blamed for corrections or classicizing emendations of the popular Andalusian text, was not qualified for such a "role". While he seems to have been a first class calligrapher, evidence is provided here that his knowledge of Arabic was not at all on a level with the standard of his calligraphy. He is shown to be responsible for the wide range of textual problems which have puzzled successive editors of the Díwdn. (c) Consejo Superior de Investigaciones Científicas Licencia Creative Commons Reconocimiento 4.0 Internacional (CC BY 4.0) http://al-qantara.revistas.csic.es 314 J. A. ABU-HAIDAR AQ. XX, 1998 In sharp contrast with all this, Ibn Quzman is shown to be an unrivalled master of Arabic and classical Arabic lore. He also seems to have had a remarkable "acculturation" in the Romance language. Apart from citing the evidence for these findings, this paper amounts to an illustration of how they can help towards a better reading and understanding of Ibn Quzmàn's Dlwdn. (c) Consejo Superior de Investigaciones Científicas Licencia Creative Commons Reconocimiento 4.0 Internacional (CC BY 4.0) http://al-qantara.revistas.csic.es

in al-Andalus known to us, and our chief treatise, so to speak, on the nature and physiognomy of this literature.As such the Facsimile imposes a severe discipline on us in reading and interpreting its texts.No alterations of any form with a view to improvement should be made in it unless such emendations are imposed or irrevocably dictated by the rationale of the text itself.As the title of this paper purports to show, both the script and the text of the Facsimile have some giveaway secrets which should be studied and ascertained with the utmost heed.What I refer to as 'secrets' are simply some features of the script and the text which, by their repeated or even consistent recurrence start to serve as guidelines to reading any faulty or uncertain script, and to deciphering and interpreting the diction used by the poet.I hasten to add, however, that what I mean by 'faulty' script will become apparent in the course of this paper, and that I do not refer thereby to the scriptio defectiva like J for 4J or 5 for ^ which, after all, is the hallmark of popular literature.This paper will deal exclusively with the 149 zajals which appear in the St Petersburg manuscript.This is because the 1896 facsimile edition presents these zajals in the form in which they have come down to us, while the other zajals or fragments have been culled in recent times from edited texts like Al-'Itil al-Hdl I of al-Hilli .^ Readers now have at their disposal two painstaking and competent complete editions and translations of the 193 zajals and fragments attributed to Ibn Quzman.I refer to the edition by Emilio García Gómez (1972) in Latin characters, with translations on the facing pages, and the more recent edition by Federico Corriente (1995).Corriente had already published an edition of the Diwdn in Arabic, with transliterations on the facing pages (1980).He followed this 1980 edition with a translation of the Diwdn into Spanish, entitled El Cancionero hispanoárabe, 1984.Corriente's first translation of the 193 zajals and fragments appeared in 1989 under the title Cancionero andalusí, and has now been reprinted in a revised and updated edition (1996).
While I have referred above, and in the title to conclusions relating to the script and text of Ibn Quzman, I have already given hints perhaps that a discussion of such conclusions cannot be meaningful without reference to both the copyist and the poet.These inferences or conclusions will be outlined below, and each of them followed by examples from the text to show AQ.XX, 1998 The lapses on the part of the copyist discussed so far, might well be described as inconsequential.They are easy to detect and to rectify.Other lapses, however, have occasioned serious confusion.
In zojal 15,7 the copyist fails, it would seem, to hear the last radical in the common adverb />^L^ and writes: Make no mistake about Abu Bakr when you try to find his peer Garcia Gómez reads the defective adverb as haiy, and transliterates: La yaglat fî Bû Bakr haiy yatlub la-hu [min] sabih (Garcia Gómez, 1972,1: 82).
In his 1980 edition Corriente, likewise, reads the term in question as hayyan (Corriente, 1980: 114-115).In the 1995 edition he reads the term as hand.
It is significant to note in passing that the copyist writes \ .r.-.il and \ .^^II 'testicles', in the same strophe, instead of the more correct forms (^.r.-.II and ^.^\ II respectively.
To give one or two more analogous examples, the copyist writes ^ j j for I jj or pi J J in 15,1, ^SJ3 for Ijj or P Ijj in 17, 1 and in the same manner, I .tiA for ^JLM in 149,1.
With all these examples of the copyists' incompetence, and the multitude of others still to come, to speak of the copyist as «less than completely at home» in the Andalusian dialect, as Gorton does, is to minimize or completely obscure the textual problems of the Facsimile (Gorton, 1975: 1-2).The way Corriente keeps speaking of the «corrections» or classicizing emendations made by the copyist ^jiti AII ¿-^LiJI rr^-> .^"iseems to be equally misguided, (see e.g.Corriente, 1995: n.5, 23;n.l, 93;n.2, 98, etc.).Rather than attribute any classicizing tendencies to the copyist, we should guard against standardizing any of his schoolboy howlers as either classical or Andalusian Arabic.
In zajal 9, 37 the copyist writes ¿yx^ for what is clearly in the context Lo LLC 'we have not got' : We have not got, my Lord, in the Peninsula (anyone) other than you (to depend on).
Both Garcia Gómez (1972, I: 52) and Corriente (1995: 53) read this as ma'nd, meaning 'we have'.To avoid the resulting anomaly in the sense i.e. 'we have (people) other than you (to depend on)', Garcia Gómez introduces the negative particle /a at the beginning of the verse, and Corriente turns the poet's declamatory statement into a question: ¿Tenemos, señor, en la Península sino a ti? (1996: 75).
Ibn Quzmân does his poetical genius much credit by incorporating into his poetry numerous popular sayings.They enhance the familiar or intimate tone of his compositions.It is a cause for surprise that the copyist gives a most glaring example of his bungled work in the way he writes one of these sayings.
In the context of stressing the fact that little can be achieved without working for it, Ibn Quzmân in zajal 13, 13 uses the popular saying: ^ ¿y^ 4.Jjjuc 1 A HI 11 ^t^ ¿^uj^ «The harvest or produce (lit.the sacks'^) of him who does not water his corn is dependent on the heavens, i. e. (the rain from) the sky».With the reference to 'the sky' this is reminiscent of the French saying aide-toi, le ciel faldera.
Presumably, not understanding what he heard, the copyist writes the first letter of the final word, ^, as part of the word preceding it.He thus writes the preceding verse with a word at the end which makes no sense at all: Man lam yasqi qamhu li-s-samá dula Apparently associating dula with the verb dalla 'to suspend, let hang', he translates: Del cielo pende quien no riega el trigo (1972,1: ll-lli).
Needless to say that with the reading 4.Jjj instead of 4.JJJLC, any translation or interpretation is apt to be wide of the mark.Corriente points out in a footnote to his translation (1996: n.7, p. 84)  When the poet uses recondite or simply unusual terms, the copyist seems to be utterly at a loss as to what to make of them.In zajal 146, 1 the poet refers to excessive longing inflaming the inner corners of the eyes, the 3L0Î, presumably with tears.Hearing the word diiLoF and not knowing what to make of it, the copyist writes instead the meaningless term diil^lJI: Corriente retains the manuscript reading, and explains in a footnote that iltirmq is of Romance origin and that «it means travelling shoes and is applied to wearing these shoes » (1995: n. 1, p. 417).Corriente, unlike Garcia Gómez, does not supply the Romance terms or Romance origins he surmises, so that the reader can verify the validity of these terms.He translates, in keeping with the footnote mentioned above: Gran nostalgia provocó tu partida (1996: 343).This parallels the translation by Garcia Gómez: Pena sin igual movió tu ausencia (1972, II: 723).
Zajal 17 by Ibn Quzmán, a zajal oí seven strophes, is arguably perhaps the most eloquent and readable poem of praise in Arabic literature, and there is no shortage, as we know, of such pompous and flatulent poetry in Arabic.In strophe 5 of zajal 17 the poet tells his Maecenas, whom in fact he does not mention by name, that much as the mention of his name is sweet to the tongue, the genial impression he leaves in people's hearts (souls) is even sweeter.The copyist, however, seemingly hearing a half-word, writes ^ju^yklW ^ d[jt3 instead of (^yail\ ^i é,j¿kj «the impression you leave in people's hearts»: Both Garcia Gómez (1972,1: 92) and Corriente (1995: 74) take the bungled spelling djtJâ to be a misspelling for AJ^ (a post-classical loan word from Persian)^ which Garcia Gómez transliterates as ka 'ak, and Corriente explains as 4-1/1 •^.Both translate the term into Spanish as 'pasteV.There is no need to say, however, how much it trivializes the context to tell a Maecenas that the impression he leaves in people's souls is sweeter than a piece of cake.I have no hesitation in treating the preposition ¿^ in the verses quoted above as added bungling by the copyist, and in trusting the discretion of the poet to say: AQ.XX, 1998 That the copyist had little familiarity perhaps with the term waq' is given further confirmation in zajal 46, 5 where he writes: Ibn Quzman shows an exceptional versatility in the use of rhetorical figures such disjinds 'paronomasia' and tibdq 'antithesis', etc.There is little reason to doubt that in the verses reproduced above, he would have retained the consecutive muid'ama 'balance between phrases', and written what the copyist fails to hear clearly, or to undestand: There is little doubt too that Ibn Quzman would have written ^-^ .t.l and not 1 -% .7tl.Garcia Gómez transliterates the second of the four versicles above as wa-mauqV az-zamir (1972,1: 244), while Corriente reads: ^hí-»jj Mj^j (1995: 160).

ARABIC
The various parts of this discussion, it is hoped, will complement each other in providing a helpful approach to reading the Dmdn of Ibn Quzmán.But they unavoidably overlap inasmuch as examples given in one section could well serve to illustrate the purport of other sections.
Ibn Quzmán, as we know, states in strong terms in the introduction to his Dïwdn that observing the rules of desinential inflection, i'rdb, in the zajal is, like death, an object of abhorrence.While using the tanwîn quite extensively in his zajals, more often than not he gives it the accusative case ending irrespective of grammatical requirements.It appears with the accusative case ending as the subject of a clause or sentence, as the predicate of a particle like ¿,1, and quite often after prepositions, and in possessive constructions.
Examples of these will appear in the course of the discussion, but the way the copyist realizes the tanwîn, and the irregularity or lack of consistency with which he writes it, gives him away as a beginner or raw recruit in writing Arabic, and confirms still further the conclusion that the Facsimile of the Dîwàn reads like a dictated copy.While in the majority of cases he realizes the tanwîn as doubled wowels at the ends of words, he has quite often confused his readers by writing the tanwîn as ¿,1, or even, at times, by writing the double vowels and adding the alif and nun after them.This irregularity could easily suggest two or more copyists instead of one, if it did not occur within the same zajaU and, on occasion, within the same strophe.Thus in 34, 7, for example, we read: jLijL^ I J,,,» ,111 Â-cLiu ^ySk j_à

«The day I visit you is a festive day Every hour of it brings a new joy
The way the tanwîn is realized in the first verse above has proved to be one of the least problematic of cases, although Garcia Gómez reads ¿,1 in the first verse as part of the tanwîn, and transliterates: yauman nazurak andfî 'îd (1972,1: 178).
Other instances, however, of the copyist's uncertainties or gaucheries, have proved to be much more confusing and serious.
In zajal 1, 3 both Garcia Gómez and Corriente have missed the copyist's bungling and both retain and read the ¿,1 he writes for the tanwîn in hutdman as a separate particle.The copyist writes: I waste away between hope and despair I have become a wreck, a shrivelled body AQ.XX, 1998 Garcia Gómez transliterates the second verse as follows: wa-sirtu hutdm in suhitydbis (1972,1: 6).
Despite the poet's predilection, so to speak, for the tanwm.Corriente emends rdjin in the first verse to read rdjT, and attributes the manuscript reading with tanwfn to what he calls 'the classicizing tendencies of the Oriental copyist ' (1995: n. 5, 23).
In zajal 20, strophe 14 the bungling of the copyist, seemingly taking down what is dictated to him, reaches one of those fairly common extreme cases of confusion where he writes: Here, he does not only realize the accusative tanwm as ¿,1, but being totally unaware of the meaning of what he writes, and guided primarily, perhaps, by what he hears phonetically, he joins the ¿) he hears at the end of the accusative tanwm to the following word.The poet is clearly saying: You and I are on a par with each other, and there is no other better than us.
Corriente, apparendy unaware of the use of the tanwm, changes \j^\ into jui^l and reads: J>awl Ljjua.1 ^ despite the stilted classical bent of such an emendation (1995: 85).Garcia Gómez changes the ¿, of the tanwm into cj and transliterates: Id 'uhrà bi-'agwad (1972, I: 108).Needless to say that any translation, with such emendations, is apt to be wide of the mark.
A much greater confusion has, likewise, been caused in 43, 3 by the way the copyist realizes and writes the tanwm.An element of shoddiness in placing the diacritical points, in this case, has not helped either.The copyist writes: What has primarily confused readers in this case again, is that the copyist writes ¿,Là.i for what is obviously l^>^ 'a trap'.The poet says about a loved one: He is straight and shapely, and excellent above all If he is a trap (a snare), by God I did not Bring him about; it is God who shaped him (lit.brought him about).
As a part of the endless arbitrary emendations Corriente introduces in his edition, he reads: vU.JJI bl c.^ (1995: 151).
There is no apparent justification for changing ^j^ into \^^^ and Lajiû into ^".A.«.,j.Equally difficult to explain or to justify, in the third verse, is the way Corriente changes 4.111 into the enigmatic or simply unrecongnizable JJI as he often does in the course of his edition.
As both Garcia Gómez and Corriente miss the term I \ ^ 'a trap; a snare' in the context, there is little point here in giving their respective translations, which could only amount to conjectures.The poet, incidentally, uses the term ^ 'trap' elsewhere in the DTwdn, as in 2, 6; 20, 15; and 69, 12.In the light of what has preceded we may feel justified in emending the numerous cases oiîanwîn realized with ¿,1, or simply with ¿j as in 19, 4: where the ¿, of the îanwïn lüj is seemingly written with the following verb Jl j. Garcia Gómez transliterates: Lam nufakkirfî 'aiyi waqt az-zawdl (1972,1: 98).The reading Jl j llj j is given further credibility by another analogous case in 13,5: I cried Allahu akbar the minute I saw him.
Corriente reads ^^1 U with a hamzat wasl instead of (^U, and adds the third person masculine pronoun to the rhyme word, and the preceding rhyme words, without indicating to the reader how all these are meant to be read in a popular context. In zajal 9, 40 we have an instance of tanw In with a genitive case ending, again written clumsily by the copyist: Should I come (to you) concerning a matter I fear.Garcia Gómez transliterates: ...ft 'amar, in nahsdh (1972, I: 52).
Corriente reads the ¿,1 of the îanwîn as LJI: In the light of the preceding remarks also, we can go on to emend even less obvious cases of tanwin.In 45, 8, for example, Ibn Quzmán says about a benefactor he is praising, that he is not happy with gold unless he sees it spent on others.Unlike the copyist, as we shall see, Ibn Quzmán knew his Arabic well.If he at all meant ¿>-^ to be the subject of his first verse, he would have written j22.Since the subject is his Maecenas (masculine) it becomes imperative to read: The copyist, besides, writes ¿¿^ with a clear dagger a///above it.Was he at all unsure how to write the accusative tanwTn of a noun ending in nilnl Indications have already been made in Part I above that the blundering of the copyist is not restricted to spelling mistakes in writing the tanwm, A lot more goes to show the wide gap between the copyist's "quackish" Arabic and the poet's mastery of the language and its literary lore.The one charge, as already indicated above also, which cannot be made against the copyist is 'classicizing tendencies' on his part in copying Ibn Quzmán's Dïwdn.Whenever the poet's language gets a little learned or abstruse, the copyist has left serious problems for his readers.
In zajal 9, 5 the poet seems to be using Ui-^ in the sense of JdU^I 'making null and void, or of no avail': Mirth is essential for foiling (doing away with) worry.
In 9, 9, however, the copyist writes: Hearing, it would seem, an uncommon term like ¿j:.yï\, and missing the definite article with a sun letter, he sets down only the phonemes he hears.The poet is simply saying that timidity, or, in the context, fear of the peering raq Tb does not allow for lasting bliss (in love): Both Garcia Gómez and Corriente read az'aq, and try to make sense, as best they can, of the bungled reading in their translations.The poet uses the cognate terms az'aq and zaHq in zajals 13, 2 and 18, 1 respectively.
As the following sections try to demonstrate, Ibn Quzman, like al-Mutanabbî before him, could well claim a familiarity with the unusual and unfamiliar of classical Arabic diction -the shawdrid.He could also claim equal familiarity with the language of the Qur'dn.In zajal 38, 3, and in a prologue of adulation, the copyist writes: It is as good as certain that we should read «±JÙ 'to blow', as in silra CXIII, 4.Although Garcia Gómez retains the reading nafas, he appropriately translates: En su inteligencia no sopla Satán (1972,1: 197).
Corriente reads naffas without any evidence to support his reading.He translates: tiene un talento no inspirado por el diablo (1995: 132).
There is little doubt that the poet means to say that the talent, he is praising, is not tainted by evil or the breath of Satan.«A talent not inspired by Satan», as Corriente translates, does not amount to much praise in the context of a panegyric.Corriente's reading and translation, besides, do not at all take account of the preposition <uJx: which appears with the verb, and seems to be sufficient justification for reading ^jl U^ .t. <LJJC. .% Ù\ ^^ «not blown on by the Devil».
This survey can be drawn much longer, if not endlessly.But one final example here should suffice to point out that much as the poet was widely versed in the classical Arabic literary lore, the copyist did not have, perhaps, even a hearsay knowledge of it.
In zajal 121, 1 the poet, using a metaphor in which he draws extensively on classical Arabic lore, tells a loved one who shows him the cold shoulder: If your delay should turn out to be (a steadfast warrior like) Miqdld al-Kindl, my patience (much more steadfast) is 'Amr bin Ma'dî.
Not only does the copyist write 'al-Hindî' for 'al-Kindî ' which Corriente has rectified (1996: n. 2, 302), but for any one able to consult the Facsimile, the copyist had, for all appearances it seems, started by writing ¿^Jf-^ as two words, and, seeing perhaps that they looked quite awkward for a proper name, went back over what he had written and joined up the two syllables.
One aspect of the a^'dl of Ibn Quzmán which has never, perhaps, been studied and properly appreciated, is that the dropping or disregard of i'rdb by the poet, has not presented obstacles to a clear understanding of what he writes.In this light his work constitutes a strong argument that / 'rdb is accidental and not essential to a language.What has presented difficulty, as we have seen, is the blundering copyist who, quite often it would seem, was unable to visualize clearly what he heard in order to be able to write it down clearly.Quite naturally, the wider the gap between poet and copyist in cognizance of Arabic vocabulary or Arabic classical allusions, the more likely it was that serious bungling would occur.

III. IBN QUZMAN DISPLAYS AN UNRIVALLED MASTERY OF ARABIC
Just as the texts of Ibn Quzmán reveal his mastery of classical Arabic, our awareness of the poet's consummate mastery of the language can help us to set right some of the bungled readings in his text.
In zajal 85, 3 the poet addresses the sacrificial ram of the *ïd with what amounts to a serenade, calling him habîbî 'my beloved'.While the verse containing this passionate address, the first in the strophe, presents no problem, the third verse has proved quite enigmatic: Harîfak 'your associate; your companion' which appears in the rhyme position in verse 1, is perhaps an uncommon term, but it is congruent with the tone of endearment in the rest of the verse.If we remember, however, that Ibn Quzman displays throughout his Dîwân a competent knowledge of classical Arabic and pre-Islamic lore, we would be as good as certain that he was aware that the term > .^^refers to the 'heart' or the 'bosom'.He would have been equally aware that just as the term > «j^ -•• refers to a camel's hump or the fat of the hump, it refers also to the hump cut into pieces, and he would have used it metaphorically to refer to juicy cuts of mutton.In the light of the poet's resourcefulness, there is little doubt that the enigmatic terms in verse 3 are é.-¡-j^ and AQSA .U respectively: As verse 4 goes on to say: SdjLçjLâ JLa.JLS «düL^ Jl^ JÙ\ «HOW are your roasted cuts, and the dried strips?», the logical sequence is complete.
Until we get a more convincing reading, I have read >iilllj ^_^' .-iUTS ^JA «do not butt me with the head and shy off» in verse 2, in place of the suggested reading JQTS ^-^V^ .^"^^^ «No te avergüenza huir?» as read by Garcia Gómez and Corriente, and by al-Ahwânî before them.My reading is more in keeping with the text, and with the plea for mercy which follows in the verse.
Yantah is attested in a zajal by Ibn Rashid, an older contemporary of Ibn Quzman, which was first published by S. M. Stern mAl-Andalus in 1951 (XVI, 379-425), and republished in 1974 (Stern, 1974: 192-5).Stern's reading and interpretation of the zajaU however, shows that there was much in it he could not make out, and both the meaning of the term yantah and the significance of its use in the fifth and last strophe of Ibn Rashid's zajal remained as a result unrecognizable in his work.Stern reads sharrah, in strophe 5, i. e. «cut meat into oblong slices» for the purposes of salting and preserving as sarrah, which he interprets as «to comb one's hair».Consequently, instead of understanding mallah, which follows sharrah in the same verse, as meaning 'to salt', he translates it as 'beautifies himself.Instead of what is clearly 'amal 'aldhablu mabzdr mumallah, i. e. «he put up on his rope spiced salted meat» he reads, apparently also with the sanction of H. A. R. Gibb, 'amal 'aid 'aylu mabzdr.Taking the anomalous reading *aylü to be a cognate of 'd'ila 'family', he translates: «prepares for his family spiced and salted meat».
For our immediate purpose, however, it is enough to point out that Stern translates the term kabsh implying 'a hefty ram', as 'a lamb' and he tells his readers that «Many a zajal of Ibn Quzman turns round the lamb and the flour of the * Id... » (1974: 195).Perhaps because the meek lamb is not known to «butt with the head», Stern translates «lays '¿ndl kabshun fa-yantah», as «I have no  (1974: 195).In fact Stern rcaás fa-yuntah (passive), instead of fay ant ah, which would make the verse in question literally mean «I have no ram which could be butted with the head», and hence the seemingly conjectural rendition «I have no lamb to kill».
'Aldhablu is patently clear in the MS text published with the transcription by Stern which H. A. R. Gibb read, and to which «he has contributed a few emendations...» (Stern, 1974: n. 45, 193).Even if 'ayl, should prove to be a lexical item in popular Andalusian, meaning 'family', the preposition 'aid, preceding it in the text, does not in any context mean 'for', to justify translating 'aid 'ayId 'for his family'.The root nataha, likewise, does not semantically suggest either 'kill' or 'slaughter'.More often than not it applies to sheep, rams, bulls and cows acting, and not acted upon.
It is to be regretted, in the light of all this, that Stern's editor did not look carefully through these texts before republication in 1974.
That Ibn Quzman was abreast of pre-Islamic lore is attested in zajal 12, in which festivities seem to be apace, with the poet enjoining the incumbents to make haste with all the preparations.In strophe 5 he urges them saying: The text in this case is clear.Yet Garcia Gómez finds it necessary to change A^JJ-XJC into 'airlkum in his transliteration, and he translates: Corriente retains the facsimile reading, but he translates: ¡Preparad la camella, preparad al hebreo ! (1996: 81).
He goes on to explain in various footnotes to his editions and translations that the 'hebreo' or 'Hebrew' mentioned in the text seems to be a member of a 'troupe' enacting the story of Joseph in Egypt, while the Arabs and the shecamel also mentioned in the zajal, represent the merchants who bought Joseph.
Our awareness that Ibn Quzman was well versed in his pre-Islamic lore, however, gives us the assurance that what he is referring to by ^S^J^JC, as a counterpart to A^l^b, is the j\ ô.nl ju^ which referred to a sturdy camel innured to long travel, or, for that matter, to a she-camel just as well, if not better qualified.Of course the metre of the zajal does not allow for j^ A£J1 ô.itl so the poet resorts to license and says ^S^j^, Is not the zajal, after all, a genre which does not take too much heed of the fine points of syntax or philology?
The two antipodes, the resourcefulness of the poet, and the incompetence of the copyist, can help us to clarify problematic passages with a fair degree of certainty.
Zajal 27 starts with an encomium of a loved one of the poet, or, more accurately, of his fascinating looks, what his hair feels like, and what his chest and his mouth look like.The first two verses of 27, 1 read as follows in the Facsimile: Both fjnfv'\ in the first verse, and ¿i.xl in the second have proved difficult to read and to interpret.
It is as good as certain, however, that with his versatile vocabulary the poet says that the hair of his loved one is, as it were, made of (j <-> A% that is «made of down» or is «as soft as down».The copyist, seemingly unfamiliar with a term like çjr> A\, deemphasizes the last radical and writes ^JJÜUOJ.Likewise, the bosom or chest of the beloved is described as white or, literally, made of ¿1^, implying that it is as white as «cotton» or «the core of a palm tuber».But the copyist, to all appearances, finding yaqaq unfamiliar and inscrutable, «makes a hash of it» and writes the term as JIJG.This made Garcia Gómez transliterate wa-sadru min ta 'annuq, which makes little sense and does not tally with his translation «con pecho para abrazos».García Gómez retains ñamas in the first verse, but translates wa-sha'ru min namas as con pelo cual de hurón v^hich suggests that he takes namas to be a reference to the nims or 'ferret' (Spanish hurón) (1972, I: 146-7).Reading nims, however, would violate both the metre and the rhyme scheme.
Corriente introduces drastic changes in his reading of the problematic verses, without any attempt to justify these changes, or to apprise the reader of them.He changes all three substantives, preceded by the preposition min into verbs.As a result, he finds himself constrained to replace min in each case with the relative pronoun man.What was a problem of vague vocabulary becomes a serious one of halting syntax -man, as the subject of three separate phrases remains in each case without a predicate:^ No question marks appear as in the 1980 edition.With the emendations outlined above, it is suggested that the two problematic verses should read: That the copyist had deemphasized namas, and written namas instead is given added confirmation further on in zajal 27.Strophe 6 in the zajal ^ is one in which the poet is addressing his Maecenas, having finished, so to speak, with the love presagement.Verse 3 in strophe 6 reads as follows in the Facsimile:
Both Garcia Gómez and Corriente deemphasize <J^^y^ in the verse and read i_ij^ twice, and both eliminate ¿,1 without giving any explanation for their emendations.Their elimination of the ¿jl, however, is in all likelihood due to the fact that they missed reading the tanwin realized by the copyist as a separate ¿,1 (i.e.¿,1 e.i^>-uû for U^^u^) as explained in some detail in the preceding section.It is my strong hunch also that the copyist had, unnecessarily again, deemphasized the ^j^ in cj^^o^ preceding the absolute accusative li^^^ and written instead,justashe had written QMAJ in lieu of(j^ A' I in the opening strophe.Short of a more convincing reading and interpretation, we should read: If one like him were to be found (at all), he would give (lit.spend) only scantily.
All this should not give the impression either that the wide gap between the poet's versatility and the copyist's incompetence is restricted to their grasp of Arabic vocabulary.The copyist, as one would expect, is equally at sea whenever the poet resorts to Arabic idiomatic usage, metaphorical usage or, still worse, to the use of figures of speech such as antithesis, paronomasia, etc.
In zajal 20, 4 we read: He reprimanded me, gave me a dressing down I felt embarrassed before him, and was perplexed.
It is as good as certain that the poet, knowing his Arabic well, would have used the correct idiom in the second verse -^ju ^ -kJLui and resorted to the plural of j-i for requirements of rhyme and metre.It is as good as certain too that the copyist replaced the preposition ^^ of the classical idiom, unfamiliar to him, with the preposition ¿^ more familiar in common jargon.In the light of Ibn Quzman's proven familiarity with the classical idiom, editors of his Dîwdn can well reinstate the idiomatice ^ in the second verse above without compunction.
As already indicated, there is a clear testimony in the Dîwdn that Ibn Quzman, living three centuries after such a champion of badV as Abu Tammâm was quite comfortable with the use of the rhetorical figures of speech that Abu Tammâm and other Abbasid poets had popularized, and particularly such figures as tibdq 'antithesis' and jinds 'paronomasia'.Addressing his Maecenas in zajal 19, 14, Ibn Quzmán wishes that he should go on moving from happiness to a greater bliss and happiness, and that he should witness as well the delight of his friend, and the rage of his foe (lit.envious one):

J ft 111 ^ H .]r\ t.C. q .^Juusü J ^O-usUI jjj^j
The poet, having thus indulged his taste for the use of tibdq 'antithesis' -jjj^ and i^^r, 3^Juu3 and jj .n-^, he goes on to indulge it further, and says in the following verse:

^J-^^ J^-^*^ s>-*v3 ù ^ '"^^ '^^^ J
Hope you will put on weight, while he (the envious one) remains as thin as a stick.
The copyist, having in all likelihood missed the niceties of the two cases of tibdq in the preceding verse, goes on to make a travesty of this figure of speech in the latter verse as well.He apparently hears tasmd instead of tasman, and he writes: Garcia Gómez transliterates: wa-'ett tusammà wa-hu bi-hdl al-'ud (1972,1: 102).
The translations, in both cases are, quite naturally, wide of the mark.
The copyist does not fare better when metaphorical language is used.In zajal 141, 1 the poet makes a simple declamatory statement that the heavy burden of the penalty (the fine to be paid) falls on the one who loves: By writing waqr instead of wiqr, and by an added uncertainty in the placing of the diacritical points, the copyist seems to have confused all his readers.Nykl reads: But while reading 4u»l^^l, Nykl seems to identify this term with ^l^^l ^passionate love' as can be seen from his translation: Prodiga tu amor a quien te ama (1933: 436-7).

He translates:
Ahórrate nariguera con quien ama (1996: 334).Ibn Quzmán's knowledge of Arabic seems to have been equally matched by his familiarity with the Qur'anic text.Here too, our awareness of what I have called the two antipodes, the poet's competence, and the copyist's uncertain grasp, can be of help in following the vagaries of the script, and in putting them right.In zajal 38, 3 we have already seen the copyist write ^^^Jù for ^XJLI, a term acquired from sura CXIII, 4. In zajal 93, 3 he writes ¿)^UÍi ^^i for what is clearly in the context a reference to ¿H-^^^ 'the self-exalted and arrogant' of^w'm XXXVIII, 75.
Such ineptitude, or such repeated 'clerical errors' can help us to anticipate and lay bare similar other errors.When the poet says to a beautiful, but uncompliant loved one in zajal 114,4 what seems to read jj>iJI p ÜLa c^Sj we can have little doubt that he is echoing suras III, 185 and LVII, 20, and that we should read in place of the copyist's bungling jjjjül ^Ll-o cLJ»i j, or, in other Qur'anic terms L:»AJI SU^aJi f-1^, the comfort and pleasure of illusion.AQ.XX, 1998 Garcia Gómez emends the faulty term, and transliterates: wa-anta mauqi* al-gumr (1972, II: 586).
Corriente retains the Facsimile reading unchanged, but explains in a footnote that maqd' implies mauqi \ the way Garcia Gómez had read the term (1995: n. 2,352).

IV. OTHER TEXTUAL CORRECTIONS IN THE LIGHT OF THE PRECEDING GUIDELINES
What I refer to as guidelines are the two basic inferences drawn from reading the Facsimile of Ibn Quzman's Dlwdn, and illustrated in some detail in the preceding pages.They can be summed up in what I referred to as the two antipodes, a poet fully abreast of classical Arabic diction, idiom, metaphor and figures of speech, and a copyist who seems to be at sea with literary Arabic be it classical or popular.Apart from being a good hand at calligraphy, the copyist, it would seem, had little else to commend him for his task.
In the light of these conclusions we can be constandy on the lookout for the original and expressive literary language of the poet and the ersatz to which the clumsiness of the copyist can often reduce it.
What follows here are only a few examples to emphasize still further the remarks made in the earlier sections.
In zajal 14,9 and after some fairly fulsome praise of a Maecenas of his, Ibn Quzman resorts to the traditional metaphor of the moist or dewy palm to describe his benefactor's generosity.He then goes on to compare him to the rain-bearing clouds.He says:

LAJ) Ô-^ L?*-^' *--°'^^
The confusion, or perhaps an element of absurdity in this case, appears in the last verse quoted.'His palm', says the poet, 'is more moist than water, so that I am (we are) bewildered, whether he is a raincloud or a palm (of a hand).This is how both Garcia Gómez and Corriente have read and translated the verse.Garcia Gómez translates the last two verses: Corriente, likewise, translates: Perplejo entre dos estoy, si es nube o es mano (1996: 86).Remembering, however, that Ibn Quzmán knew his Arabic well, he would have been sure to associate rainclouds with wakf or 'shedding', 'dropping' their rain.He would have been more likely to wonder whether the Maecenas he is praising is a raincloud, or more copious with rain, so to speak, than rainclouds, in which case we should read: In fact the Facsimile text has a seemingly otiose a//f written after t_jLx-uj, which should have guided editors to the correct reading.It is the alif of j\ of which the copyist dropped the j.He could not visualize, perhaps, how jl could appear on its own, and then as part of the following comparative ^-Û^ jl.
Ibn Quzmán, knowing his Arabic all too well, would be aware that the verb hcijja, apart from referring to making the pilgrimage, is extensively used as a transitive verb, and that among its primary meanings, is 'to visit repeatedly or frequently'.These seem to be the twin implications of the imperative ^^ in zajal 101,1: Repair to or make your pilgrimage frequently to the cup, even if you should be (bent double and) walking with a stick.
The text is patently clear, and I see no reason or justification for Corriente's surmise that «. 'is of Romance origin meaning 'suck' or 'suck at the breast ' (1995: n. 7, 321).^^lUl ^^ conveys the poet's characteristic irreverence for the injunction against drinking.Besides, the striking paradox of repeated visits being enjoined when one is no more able to walk is not conveyed by the suggested Romance reading.Garcia Gómez reads the term as Arabic, although ^ The poet, of course, would be aware of the il0 ' involved in rhyming JHU with o-il, but as a zajal or hazl writer, he was not one to observe the fine points of prosody or grammar.wa-hag¿ (sic) bi-l-kds law annak 'ala 'ukkds (1972, II: 524).
In the same zajal and the same strophe, 101, 1 the poet seemingly aware that ^ like ^i/<l, ^^±i-il, ^JLii and ^1 refer to dawn, daybreak or the sun shining brightly, speaks of cheeks from which light emanates or shines brightly: Corriente reads ^Ju, which is sufficiently clear in the text, as ^.L and comments in a footnote that this term too «is of Romance origin, meaning *to light' or *to shed light '» (1995: n. 1, 322).Garcia Gómez reads ja//^ instead of ^Jb, but, apparently surmising the sense of the text, gives a translation which tallies with ^Ju: Gente con luz que les sale de la faz (1972, II: 524-5).
^_=».appears in zajal 20, 29, where it stands to reason that it is, in this case, the copyist's mistake for 4.^1^ 'need': If a need (of mine) should arise, I would go after in on foot, It is only right if you have a need that you should be relentless (in seeking it).
Both the context, (seeking help from a benefactor), and the balanced phrases in the two verses leave little doubt that the copyist wrote ^^, suggesting 'pilgrimage', for i^La.'need', in the first verse.Corriente reads ^_a., and reads (^ jLt^ for ^Ju^, which is clearly written in the text, and which is more compatible with the 2nd person form of address in the second verse.He gives the reader no grounds for his emendation, or for reading as interrogative what seems to be a declamatory statement by the poet.Corriente translates: Aunque peregrinar a pie me pidiera, ¿estaría bien discutirle lo que fuera?(1996: 102).Garcia Gómez reads Mga, but he reads the imperfect subjuntive ^ jLx3, following appositely after ¿,1, as a Romance term tomar.In order to sustain such a reading, however, he changes the text of the second verse out of recognition:

V. THE ROMANCE ELEMENT IN THE DIWAN OF IBN QUZMAN, AND CARITATIVE DIMINUTIVES
It is no secret, of course, that Ibn Quzman peppers some of his zajal strophes with Romance terms and expressions, or what Garcia Gómez aptly calls Romancismos.It is not the intention here nor the place to study these Romancismos.But, as in the preceding sections, the intention is to trace any giveaway secrets or guidelines for a better reading of the Dlwdn of the poet.The Romancismos are too limited, or simply an insufficient indication as to whether Ibn Quzman was at all bilingual.In fact, the Romancismos by themselves tell us little about Ibn Quzmán's knowledge of Romance, or what one may call his linguistic acculturation.Some features of the poet's Arabic diction, however, are more revealing about his familiarity with Romance than the isolated use of Romance vocabulary.
One such feature which has received little attention is the poet's use of the Arabic verb arMa (Form IV), as the equivalent of the Romance querer.While he often uses the verb to convey its familiar Arabic denotations 'to want', 'to want to have', 'to desire', etc., which are equally shared by querer, he quite frequently uses the Arabic narîd, tarîd, yarîd to convey the other primary meaning of querer, i.e. 'to love', 'to like'.In fact narïd is used in the very first verse of the Diwan, zajal 1,1 in a sense peculiar to querer: I love and for fear of infatuation (lit.getting hooked), I cry.
This translation is more to the point than Corriente's who translates narîd as 'want' or 'lo que quiero'.Ra'ayt, likewise, does not suggest conseguir as Corriente translates it: Si consigo lo que quiero (1996: 279).
Another feature of Ibn Quzman's poetic diction which betrays Romance influence is his prolific use of the diminutive.It must be pointed out, however, at the outset, that the distinctive feature referred to here is the poet's use of the diminutive as a caritative form, or as a form of endearment.There is no suggestion made here that the diminutive form is not used by other Arab poets in Islamic Spain or outside it, or that the famous al-Mutanabbi, for example, did not use it perhaps as profusely as Ibn Quzman did.But whereas al-Mutanabbî used the diminutive to scoff at his opponents, or as an expression of contempt, Ibn Quzman invariably uses it to convey notions of affection, fondness and endearment, in the same way that it was used in contemporary Romance, or in the way it is attested in njany of the surviving Romance kharjas.In this respect, Ibn Quzman stands out as the sole representative among all the poets and writers in Arabic in Islamic Spain of the tradicionalismo which Ramón Menéndez Pidal chose to identify as a constant in cultural developments throughout the history of Spain.Ibn Quzman's use of the caritative diminutive, in other words, is one of those perennial literary traits which seem to defy time in literary productions in Spain, or which survive, so to speak, with every ecological succession.
I have examined the diminutives in the DTwdn of Ibn Quzman at some length in a study published back in 1989.^^There is no need to repeat here the arguments produced in the 1989 study in support of likely Romance influences on the poet's predilection for the diminutive, and for coining, all too often, diminutive patterns of his own.One example should suffice to illustrate what one can call the Hispanic tone of many of his diminutives.AQ.XX, 1998 The poet uses 4 n^n^ (feminine) as the diminutive of ^, at a time when both the Arabic term, and its classical Arabic diminutive à^jÀ are masculine in gender.The feminine form 4 n^n^ is patently clear in the poet's text.Equally clear also, in the example given below, 62, 6 is the feminine epithet faww^a qualifying fumayma and the feminine suffix pronoun referring to it, bi-hd: My beloved came to me with a fragrant mouth as if he had just eaten an apple.
When the poet goes on to use 4.^^^^ consistently as a feminine diminutive, as in 10, 6 in 67, 2 and 141, 3 it stands to reason that fumaymah constituted a translation in the poet's mind, or perhaps a 'caique' of boquita, or bokella as attested in the Romance kharjas (Garcia Gómez, 1965: e.g. 124, 144, 186, etc.).
But our primary purpose here is to see how far the poet's flair for the diminutive, and particularly for caritative diminutives, can be put to use as a guideline in reading his problematical or obscure texts.
Strophe 7 in zajal 54 has a caritative diminutive in each of its aghsdn, and a fourth in its qufl: All the beauty of the world (lit.time) is gathered in one youth He will be nine soon, or a little older One who is noble and honourable, shapely of stature and charming I have seen how noble he is, and his comely stature, prevents me from bending it.
Subay, shuway, hulay are all caritative diminutives which give a clue or a fair expectation that qudaydu with an u vowel on its first radical is yet another diminutive.It is a caritative diminutive of qadd 'stature'.Garcia Gómez misses the diminutive and reads qadruhu instead of qudaydu.He transliterates: Fa-qadrU'hu yamna '-ni an numaiyalu (1972, I: 276).Corriente reads qudaydu and translates accordingly 'tallecito ' (1996: 168).But both Garcia Gómez and Corriente miss the signification of munsabak in the third ghusn which anticipates the caritative qudaydu in the qufl.Both munsabak and the cognate masbilk imply 'shapely', 'well-shaped' or, literally 'well-cast' in a variety of Arabic dialects spoken today.Masbilk sabk likewise refers to a perfect form or stature, the implication being that it is without flaw or blemish as if it were cast masbilk by hand.
Keeping the caritative diminutive in sight can help us to to clarify an analogous difficulty in zajal 116,5 where the poet says: jjjjâ J 6.A J. 1 7k.!What has confused the editors of the DTwdn in this case is hudaydah which is a diminutive of hadd or haddu meaning 'near' or 'near him' respectively.It is familiar in a wide range of present-day Arabic dialects in such statements like bay tu hadd baytl «his house is close to my house» (Nakhlah, 1960: 121).The poet says: AQ.XX, 1998 Corriente, likewise, reads the diminutive as (_jdaJL=J and translates: Corre, pronto y sin volverte, a mi hierro, pues besarlo causa fatiga y desengaño (1996: 296).
One other caritative diminutive which has confused all readers of Ibn Quzman appears in zajal 10,4.It occurs in a verse in which the poet wishes that God had created his love as a little stem, or, one might say 'spriglet' of lavender: ¿Uia^iw <ÜJ1 AÏJU^ JÁ O. J. Tuulio reads this diminutive as khudayma and translates: «Puisse Allah te transformer en une (pauvre) petite servante!»(1941: 4). A. R. Nykl transliterates this diminutive asjudhayma and translates: «Si Dios te hubiera hecho manquillo».He adds in brackets after «manquillo» by way of explanation of his conjecture, «para no gastar!»(1922: 24 and 375).C. Brockelmann, quoted by C. Appel, translates the verse in question: «Moge Gott zu einem... (?) machen!».Brockelmann, it seems, refused to hazard a guess similar to the previous two, and supplied blanks for the diminutive in question.^^ Garcia Gómez transliterates the verse in question: ¡lau ga'alak Allah gudaima!He translates, omitting the reference to God: «de volverte leprosillal» «that you should turn into a little leper!» Garcia Gómez adds in a footnote that the strophe, in which this verse appears is difficult, and wonders whether the poet's use of 'little leper' was not due to the fact that "small pieces of money were thrown to lepers" (1972, I: 58-9).Corriente reads <LA-Í1:=>., and adopts a similar explanation to those of Nykl and Garcia Gómez.It is no doubt baffling that when the poet is clearly heaping terms of endearment on his love in zajal 10 (four caritative diminutives in as many verses in strophe 6), that so many scholars should give credence to the conjecture that he is wishing him or her to be a leper or a cripple lacking one or both hands or arms.In zajal 14, 6 the copyist has written à^'^i^ j 4-p which makes no sense, and which has consistently been read à^'^i^ j 4.^ by the poet's editors.In the same way in zajal 10,4 we should read ^^^^j^ and not 4.-a^Ii^ which, to say the least, seems far-fetched.
Having said this, however, one should not fail to point out here that scholars have puzzled over another diminutive in strophe 4 of zajal 10, hujayrdt, where the script presents no problems.Hujayrdt is used with mathdqil, gold dinars'^ or gold coins, in the same verse, and both terms seem to be used metaphorically in the context of a mounting crescendo of adulation and endearment of the beloved.In his 1980 edition (81-82) Corriente equates hujayrdt with ^^^-^ 'small stones', 'pebbles', indicating a diminutive of hajar 'stone'.In his first translation, however, (1984: 66) he clearly seems to view hujayrdt as the plural of hujayrd, diminutive of hujra 'room' or 'cámara', and he translates the term as 'Algorfas'.In his 1995 edition he points out that he now subscribes to the view that hujayrdt is the plural of the diminutive of hijr, meaning 'bosom' or more popularly 'lap', and he translates hujayrdt accordingly as ' 'Haldas'' in place of "Algorfas" (1995: n. 7, 55 and1996: 77).
Since the context is one of lavish praise of the beloved, hujayrdt is more likely to be the plural of the diminutive of hajar 'stone', and since it is coupled with mathdqil, the poet is in all likelihood referring to lj\ -%-% 4^-^^ or A'l^n^ SjK-v, meaning 'precious stones' or 'gems', and not 'pebbles', j ^ 11 in classical Arabic referred also to 'gold' and 'silver', and together the two metals were called ¿}lj->.-.11 (Lane: art.hajar).It no doubt stands to reason that the poet in zajal 10,4 was not wishing his loved one to be a leper or a cripple, as has been claimed so far, so that those wishing to have a close look at her or his beauty would confer 'roomfuls' or 'lapfuls' of gifts on her or him.What the poet is saying is simply that his love is as precious as gems or gold.With the help of scholars like Emilio García Gómez and Federico Corriente we are nearer to having a definitive edition of Ibn Quzman's DTwdn.But the edition which is to transform the 1896 Facsimile into print is still a long way off.The definitive translation of the text it follows is still a long way off too, and we might perhaps have to accept the fact that it will always have a tentative aspect about it.
This article, as its title indicates, was meant to be restricted to the study of textual uncertainties in the DTwdn of the poet.Literary translations quite often, of course, reveal differences in interpretation.In the case of Ibn Quzman's work such differences seem to be numerous and wide apart even where there are no textual uncertainties.One or two examples should suffice here to show that translating Ibn Quzman needs as much deliberation and caution as reading his erratic or bungled texts.
An Arabic term which has the connotation in popular usage of «to be left without protection or support», or «to become exposed and vulnerable» is inkashaf.It is obviously in this sense that Ibn Quzman uses the term.In strophe 14 of zajal 83, an elegy he wrote upon the death of his benefactor the ^dí/í of Cordova, Abù'1-Qâsim Ibn Hamdm, he addresses himself to the dead qddT, and says: Many people have become unshielded (unprotected) after your death, Like fleas on a bald man's brow.
I am not commenting here on Ibn Quzman's libertinism in seeming sardonic or even scurrilous in the context of a sombre elegy.But Garcia Gómez, apparently missing the popular nuances of inkashaf, translates the term as 'sale\ 'to come out': Ido tú, sale tropel de gentuza, como en la frente del calvo las pulgas (1972,1: 417).
Readers might note that the copyist misses writing Sâ.t.^ with a td' marbiita.
Corriente, however, seems to confuse the subject and object in the first verse, and translates as a result: Si el amor cuece a algunos yo me lo como crudo (1996: 333).
Side by side with the editions by Garcia Gómez and Corriente, the Dïwân should continue for a long time to come to be published in Facsimile, until a consensus develops concerning the textual problems it presents whether in the Arabic which is difficult to surmise, or the equally enigmatic Romance which should fit into a convincing harmony both with the sense and the overall rhythm of the metre.Despite the numerous readings which Corriente has convincingly put right in his recent edition, the 1972 edition of Garcia Gómez still has the advantage of having offered tentative readings and a detailed discussion of what is clearly Romance or supposed to be in Romance in the texts (1972, III: 323-513).
The rest of this conclusion is a further brief attempt to show how what I have described as 'giveaway secrets' in the text of the DTwdn can be put to use in helping us to arrive at a better reading and better interpretation of the Facsimile.
In zajal 83,5 we read what amounts to an injunction from the poet: Both Garcia Gómez and Corriente have read the offending word as assdrra which makes no sense either.Garcia Gómez translates the first verse: Come lo justo, prescinde de todos (1972,1: 415).
Neither of the two translations tells us what it is the poet is enjoining us to eat, although the drift or tenor of what he is saying is fairly clear in both of them.Being aware, however, of the poet's resourcefulness, and the incompetence of the copyist, we can be as good as certain that the poet is referring to al-sdb tree, the equivalent of the bitter aloes j».^il which the copyist writes down as al-sdr.«Sustain yourself on the bitter sdb», the poet tells us, «rather than seek the help of other people.»The poet tells us, after all, in an equally proud and dignified tone in zajal 9, 29: J!; AIU JIJLI iuS jjb liiU I am one who abides by the popular saying: Keep your personal dignity, even though you should be carting charcoal.
Zajal 9, addressed to Abü'l-'Alá' Ibn Zuhr, a benefactor of the poet, is in 42 strophes.While many references have already been made to this zajal, it would be useful to quote one of its strophes, n.*^ 24, in its entirety here to illustrate the complex problems of reading and interpretation which the Facsimile still presents: What seems to be ankd in the second verse has proved to be problematic.The poet is praising the prowess of his Maecenas when he is accoutred for battle on horseback.He describes him when he charges from a distance, goading or spurring his horse for the charge -hamaz hamza.Prodding the horse with the spur seems to suggest to the versatile poet what might be called the 'logical sequent', thrusting at the enemy with the spear ^^^^j juijLaJL or jLij-=JLi lliij, terms which have besides the wider implication of inflicting slaughter on the enemy with both sword and spear, or, collectively, ju,^-.JL i.e. with steel (weaponry).The copyist, as far as one can see, unfamiliar with either ^^ or 1£LJ could well have introduced an element of metathesis and written ^^^i.Garcia Gómez reads wa-'ankà (1972, I: 48).The copyist on the other hand could have written (^^1 j wanakâ, for ^^ j giving the conjunction a long vowel instead of its short one. Corriente reads jLa.\' ^ ib ll3lj justifying his emendation by metrical considerations (1995: n. 1, 50).llnl , of course, requires a different preposition, which would create further metrical difficulties.While it is difficult to make out the sense of juaju^JL ll^l in Arabic, the Spanish translation by Corriente «embraza el hierro» is not free from obscurity either.
The poet says in strophe 24: He who, whenever he charges from a distance Prods (his horse) and perpetrates slaughter on the enemy with his weapons (al-hadid).
(Then) those who fall upon us (in battle) are (turned into) slaves and those appointed to rule over us become (our) servants.
Corriente reads i'.^ir Iji.*^! in place of LiJl IjJuJal although it is the preposition ^1 which denotes direction.He also reads \\¿\c ^j-lj-lj ''los que albórbolas en nuestra contra hadan» in place of the clearly vocalized passive I'.^if IjJ^ (1995: 50, and1996: 73).His reading seems far-fetched.For one thing it is not the women encouraging the combatants that Ibn Zuhr is praised for enslaving, and for another, Corriente's reading misses the sharp case of antithesis, tibâq, in the poet's vaunt.«Those appointed to rule, or lord it over us, become (our) servants.»Every caution, as can be seen, should be exercised before any emendations are introduced, especially where the text seems clear, and the copyist's bungling or incompetence is not suspected.
In zajal 20, 17 where the poet is describing one of his escapades with women, his friend, whom he has enticed to go home with him, seems to ask for something wherewithal to cover her nakedness.His ingenuity seems to rise to every occasion.He promises to write a note with his own hand through which'a dress of a particular texture would summarily be provided: Jcxsi^ (3^ ^.* ^ '"-^'^ t-j|j CiJJâ I said I will direcdy write a note with my own hand and a see-through dress will be dispatched (lit.will come) to us.
It is no doubt legitimate, if only for purposes of consistency and standardization to read ^LaS in place of the copyist's ¿jILn^.But both Garcia Gómez and Corriente unnecessarily emend (^jl^ i.e. 'transparent', 'see-through', and read instead badârï, which they translate as 'new', (^jlj^, meaning 'see-through', is still attested in a wide range of Arabic dialects in our own day, and it is, with little doubt, the appropriate term in the context.
The Romance terms and expressions in the Dîwdn of Ibn Quzman could not have fared better than the Arabic has done, at times, at the hand of the copyist.To say the least, they need to be read and ascertained with equal care and caution.
An attempt was made in the course of this study to show that Romance influence on Ibn Quzman goes far deeper than the mere use of isolated items of vocabulary in his poetry.This is particularly apparent in what might well be described as his native use of the caritative diminutive.But it could well be said that the study of the poet's use of Romance terms and expressions has not gone much further than deciphering these terms or expressions, where the script is vague, or deciding their true identity with an element of certainty.Are the Romance terms used by the poet used solely for macaronic purposes, or is it possible to speak of any of these terms as loan words in popular Andalusian Arabic?These are the type of questions which the treatment or study of the 'Romancismos' has not so far tried to raise or answer.
Strophe 10 in zajal 20 has an exquisite 'Romancismo' used in it at the end of the qufl, or in the rhyme position.The poet says, expressing a seemingly universal failing: Whatever (lit.whoever) my eye sees, my palate (Sp.Taladar") would covet.
j\ U\s appears also in zajal 88, 3 but again in the rhyme position.The two terms appear in the rhyme position in the aqfdl.Zajal 20 has 29 strophes, and zajal 88 is not far behind with 27 strophes.Supplying 29 or 27 words rhyming in âr can tax the ingenuity even of a poet like Ibn Quzman^^, and the use of one or two Romance terms prove unavoidable, if not simply handy.In the light of such considerations, paladar cannot be viewed as a likely loan term in Andalusian Arabic.
Only what seems to be the last two strophes of zajal 3 in the Dfwan, a zajal of praise, have survived, and they are clearly addressed to a member of the Banú Hamdîn family.Strophe 2 ends with the following verses: dX.>£l.t...,o ¿)L^ Lajl ^ I heap curses upon whoever is an enemy of yours May God smite his head with a blow of a ''segur".
It is fairly clear from the context that shuqilr, ot the end of what seems to be the final qufl is the Romance term 'segur' 'áxe', and both Garcia Gómez and Corriente have read it as such.It accords in rhyme, or is perhaps made to have accordance of rhyme with al-budür, the rhyme word in the qufl of the preceding strophe.The appearance of the term in the rhyme position, as already indicated, is no adequate basis for designating or thinking of it as a possible loan word in Andalusian Arabic.The poet, however, uses the term again in zajal 86, 5 where it is clearly the context, and not any exigencies of rhyme or metre, which prompts its use.The context this time is purely martial, and descriptive of battles and weaponry throughout the ten strophes of zajal 86.The poet says: With JMJÙ used in such a context, it becomes legitimate to ask whether the term was not a loan word from Romance in popular Andalusian Arabic.It becomes equally legitimate to ask whether the term was JJQ.O as used in zajal 3, 2 where it has to accord with a particular rhyme scheme or more likely JMJÙ as in zajal 86, 5.It is rather surprising, however, that neither Garcia Gómez nor Corriente have identified the term in zajal 86.They both emend it to read ÔJJLOÎI, and both translate 'acero' or 'aceros\ 'steel' or perhaps 'swords' (Garcia Gómez, 1972,1: 433, andCorriente, 1996: 230).
All this shows, however, that both the Arabic and the Romance elements in the D Iwdn of Ibn Quzman are still in need of much scrutiny.
that this saying was salvaged or restored {rescatado) by an observation of al-Ahwânî.Without rectifying the text of the saying, al-Ahwânî could only have made a wild conjecture about it.Corriente does not point out where al-Ahwânî makes the comments he refers to.In his work Al-Zajal fTal-Andalus (1957: 180) al-Ahwânî reads what the copyist has written: the baffling term iltimdq into iftirdq (separation) and reads: AQ. XX, 1998 THE SCRIPT AND TEXT OF IBN QUZMAN'S Z)/W4iV 283

( c )
Consejo Superior de Investigaciones Científicas Licencia Creative Commons Reconocimiento 4.0 Internacional (CC BY 4.0) http://al-qantara.revistas.csic.es he introduces emendations which mar the forceful use of the imperative by the poet.He reads:

( c )
Consejo Superior de Investigaciones Científicas Licencia Creative Commons Reconocimiento 4.0 Internacional (CC BY 4.0) http://al-qantara.revistas.csic.esAQ.XX, 1998 Such a usage of narîd, tarïd, etc., unfamiliar in Arabic, but which could well be taken for granted by the Spanish reader, appears in numerous other contexts in the DTwdn.Thus in 16, 4: You love whoever comes seeking your help, and become fond of him.And in 28, 8: Whoever sees you falls in love with you.Whoever sees you develops a desire for you.This verse may be contrasted with the verse which immediately follows it, where tarîd simply implies you want: All the people you see are your slaves, enthrone and dethrone whomever you please (lit.*want')-ju^, used with a clear Romance reference, appears twice in the same verse in 43, 2: I love him, gentlemen, I do.The Romance signification is patently clear also in 48, 7: Do not you doubt my affection, and be sure that I love you.And in 106, 1: I never complain of any worries if I see the one I love.I have retained the Facsimile reading where là! and CJJÍJ are written together.(c) Consejo Superior de Investigaciones Científicas Licencia Creative Commons Reconocimiento 4.0 Internacional (CC BY 4.0) http://al-qantara.revistas.csic.es

( c )
Consejo Superior de Investigaciones Científicas Licencia Creative Commons Reconocimiento 4.0 Internacional (CC BY 4.0) http://al-qantara.revistas.csic.esAQ.XK,l998 THE SCRIPT AND TEXT OF IBN QUZMAN'S DIWAN 303 one day come to realize that Ibn Quzman is by far the most resourceful, talented and sensitive of all the poets who wrote in Arabic in al-Andalus.His Diwdn should be treated with great care, and no changes or emendations should be introduced on the single manuscript we have of it, unless *^ See zajal 19, 2 and Corriente, 1995: n. 3, 86.(c) Consejo Superior de Investigaciones Científicas Licencia Creative Commons Reconocimiento 4.0 Internacional (CC BY 4.0) http://al-qantara.revistas.csic.esthey are the kind of changes which, as Emerson says, constitute their own evidence.

I
^r^ C ^] j".S g I ¿\ jL^Jül dJlJ (^J^ {J^^^ Ibn  Quzman makes practically the same vaunt about his unrivalled knowledge of Arabic, and quite rightly so, as al-Mutanabbî does.In zajal 9, 35 he says: ^j^ ,^LuJ ^i v^ijiJij i.e., «the most far-fetched terms are at the tips of my fingers».(c) Consejo Superior de Investigaciones Científicas Licencia Creative Commons Reconocimiento 4.0 Internacional (CC BY 4.0) http://al-qantara.revistas.csic.esAQ.XX, 1998You could hardly see that day anything other than axes and spears*^ (lit.sticks) Shaft following upon shaft, and arrows as numerous as stones.