The Quest for the Sonnet

The Origins of the Sonnet in Arabic Poetry

in Critical Survey

Headnote

For decades, Arab and Western scholars have wondered about a possible genealogical relationship between the European sonnet and earlier Arabic poetic forms such as the muwashshah form popular in Muslim Spain. Published in 2011, Kamal Abu-Deeb’s Arabic translation of Shakespeare’s Sonnets not only offered a well-received complete translation of the sonnets; it also proposed a bold theory of how exactly this genealogical link might have worked. In the section of his introduction excerpted here, which he has rewritten in English (with a special English epilogue) for this issue at our request, Abu-Deeb lays out an argument that the polyglot Sicilian court of Frederick II (1194–1250) was the forum in which poet Giacomo da Lentini, father of the Italian sonnet, might have heard, adopted and adapted Arabic poetry of muwashshah type. Abu-Deeb also discusses what he calls his ‘fantasy’ of an Arab origin for Shakespeare’s name. We present this valuable document to you as Abu-Deeb wrote it, with minimal editorial alterations. –Eds.

Epilogue (For the English Version of the Study)

Sometime in the mid-1980s, I was working on a paper to be delivered at an international conference on the impact of Arabic literature on Western literatures. As I explored various possibilities, familiar and obscure, a phantom of an idea began to loom in my head: why is the structure of the sonnet so similar to the structure of the muwashshah in Arabic, yet the muwashshah has countless variations of structure and the sonnet has had (until recently) limited variations? And why is English poetry so relaxed about breaking rules of prosody and yet so tightly rigid about the rules that govern the structure of the sonnet, especially in Shakespeare’s time?

These initial questions never had the chance to be investigated in my own work, for reasons that are not relevant to the issue. Then, in the period around 2006, I was playing around with the structure of the sonnet and found myself almost forced to write one. I did.

To my surprise, however, the writing did not stop when I composed the cap, as it has been called by some, or the kharja, as the Arabs had called the last line of the muwashshah: my text went on growing until an additional stanza was completed with the rhyming scheme: HIH IIH or HI HI IH.

Thus, a sonnet, written intentionally in accordance with a pre-established, pre-conceived model, broke through the barrier of the model and invented itself as a sonnet with a different structure. Hence, I had the Shakespearean model:

AB

AB

CD

CD

EF

EF

GG

And my own:

AB

AB

CD

CD

EF

EF

GG

HIH

IIH

Or

HI

HI

IH

What was truly significant is that I knew the model I had intended to emulate so well and that the poem refused to stop where the model became complete. An impulsive force, emanating from the very flow of the inner rhythm, the theme, imagery, vision, etc., of the poem had taken control: I had no will of my own to intervene, to stop, to obey or to disobey.

At the time I had been translating a couple of Shakespeare’s sonnets because a beautiful and rather dakina friend of mine had enjoyed a translation I had made of the famous ‘Shall I compare thee to a summer’s day?’ The intervention of my own composition invoked my old phantom that the sonnet had been developed in Europe with full awareness of Arabic poetry in Spain and, specifically, of the muwashshah, and in particular the muwashshah as song. Now flashes of intriguing details began to shoot into my space: the Arabs had for centuries known a structure they called simt. And the word sonot, sonet, sinot is an Arabic word. Plus, remarkably, the Qurtuba poet Hasan ibn al-Hasan was called al-sonat. Did this sonat compose sonets?

Puzzling flickers of lanterns burning in the dark. But the most significant question became this: why is Petrarch the poet accredited with inventing the sonnet form? He is Italian. How did Italy come into the picture if the sonnet originated in Arab Spain, and why did it take so long, first to develop in Italy then to travel across Europe and reach shake spear (or is it, as he himself signed his name, shakespere, shak esber being an Arab name)? More puzzles. It was like looking for Darwin’s missing link. And I did not even know that a missing link was missing at all. There was no evidence that a link existed. Just an intuition.

Not being a specialist in Hispano-Arabic, I decided to give up because my main work as a scholar had very little to do with Hispano-Arabic and I did not want to waste more of my life than I had already done. Just enough to translate the sonnets, I decided.

The bug, however, never dies if there is something really interesting in it. The itch went on. And I began to scratch again. This time, however, not in Hispano-Arabic but elsewhere. I had for years written on and taught courtly love and its origins in Arabic poetry1 and while doing that I had done some reading about the courts in which courtly love was popular and might have originated; and one of these was not in Spain but in Sicily. Sicily! That is where the reward lay, I felt.

I sat down and charted the structures of all the sonnets written in Sicily – a century or more before Petrarch – in the court of Frederick II (1194–1250), in which a poet named Lentini lived and composed poetry that was unusual and followed different rules and had schemes, or structures, that are variations, yes variations, on structures of the muwashshahat produced by Arab poets all over the place. And a major one of these poets was the Arab Ibn Hamdis al-Siqilli (c.1056–1123), a fine poet in more ways than one.

1

As I had been writing my book with the Arab reader in mind, I had to give a general description of the scheme of the sonnet as Shakespeare composed it. In the process of giving this description, and having used letters, ABCDEFG, to describe this scheme, I went on, for mysterious reasons that I can no longer remember, to use numbers to describe it; this turned out to be truly interesting. Thus I had the following representations:

AB1 2
AB1 2
CD3 4
CD3 4
EF5 6
EF5 6
GG7 7

Wow.

But seven in Arabic culture (as well as some other cultures) is a magical number in much of human activity, from religious thought to poetry: the Qur’an depicts the heavens as seven. The most beautiful poems in Arabic before Islam were seven (though in some versions, ten). A select, and very special, number of suras (chapters?) in the Qur’an are called al-sab‘ al-mathani. The seven?

A cat has seven souls.

And for us all, the days of the week are seven. Etc. etc.

Why seven and where does it come from?

Is this magical, Sufi, mystical property of the seven the mysterious reason for inventing a whole autonomous poetic structure called the muwashshah, or the sonnet, and restricting it to fourteen lines, which are in fact two sevens? Wasn’t Shakespeare = shaik esber thought to have some strange Sufi-like ideas about the soul? Allahu and other experts A‘lam.

I went on transcribing the various ways in which the muwashshah was composed in the selection made by the first Arab author to write on the muwashshah. A truly dazzling variety of structures. No way of exhausting them all except in a specialized study, which mine was not. But enough was revealed to convince me that I was on the right track. Eventually, there it was: the very scheme in a muwashshah that is identical with the structure of the sonnet. And here it is, in a very rough translation just to give the reader of English some idea of the ‘content’ of the muwashshah (the Arabic text will be given later):

Rise early for the wine / and inhale the flowers

Life is in decline / if not drunkenness

Rarely do I forego / sipping glasses

as the one with the magical eyes / helps the drinkers [those sitting round for a drink]

So water me / with the daughter of the finest grapes

Give it to me pure, unadulterated / you, gazelle with eyes so black and white

A wine that resembles a trait / of your moonlit cheeks

A gazelle that is noble & just / to people

and the musk is in the incense / of the wafting breath

So shelter me / from the musk of Darin [a place name].

Its rhyming scheme is:

AB

AB

CD

ED

GG

HI

HI

CD

ED

GG

Is there any better evidence than this that the sonnets’ cap is identical with the kharja (the last phrase) of the muwashshah?

A trinity of items are of special significance:

The very division of the muwashshah and the sonnet into two components or blocks.

The fact that the couplet of lines of the Lentini sonnet could be read as we read Arabic lines of poetry – i.e., as consisting of two hemistichs usually written as one single line across the page with a space (caesura) in the middle.

The dazzling suggestion that the Italian word volta (in the sonnet) means kharja (in the muwashshah). More on this point below.

I propose initially that the Arabic terms qufl/aqfal and bayt/abyat hide crucial clues to the nature of the muwashshah and the possible way in which it influenced the formation of the sonnet in its earliest stages. The immediate meaning of qufl, namely ‘lock’, suggests that the qufl is that element in the muwashshah that establishes (or ‘locks up’) the scheme of rhythm and rhyme that will be maintained in all the following aqfal, which is not the case in the bayt.2 However, I think that more hidden and more significant meanings are at play in the naming of qufl and bayt. I hope to explore this issue further in the future, but I suggest for the time being that qufl (in one of the meanings of the word) refers to the constituent that involves repetition or ‘returning to’ a previous element (such as rhythm or rhyme), whereas bayt denotes the more varied constituent, like a bayt (house, tent) which contains a variety of objects and arrangements. This division appears to be the basis of dividing the sonnet (in its early as well as late form) to the octet and sestet, the first being the element with regular repetition or ‘returning to’, the second being a lot more varied. More on this in a future work, time permitting.

2

I shall now give a version in English of the analysis I carried out to produce the Arabic study. I am only presenting here one section of the Arabic; the other sections contain significant material as well, in addition to examples of the sonnet, a historical brief of its transformations, Shakespeare’s life and work and, most intriguingly, my fantasy that his name may have originated in Arabic in the form shaik esber, which was then slightly modified in writing and pronunciation into Shake spear. Possible evidence resides in the variety of ways in which his name was written in his own lifetime.

3

Here is my Arabic original with very slight changes:

I have said that the invention of the sonnet is attributed to Petrarch in non-specialized (and most specialized) circles – a more grave error indeed. Had it been true, it would have meant that this Italian poet had suddenly invented a complex structure, or even genre, without any previous model even in a budding state of being; this is no more likely than God making Adam out of nothing. And God himself didn’t do that: he created Adam out of mud and water, each in its turn coming out of something existing previously.

For a short time in the past I had thought that Petrarch had known samples of the art of tawshih (and the composition of muwashshah in Arabic poetry) that he managed to emulate and produce a simple form of muwashshah that has the structure for which Petrarch has been famous and which is called the Petrarchan sonnet. And had Petrarch been an inhabitant of Spain, I would have had no hesitation to believe that my conjecture is very likely to have been what actually happened. But Petrarch was Italian and I knew nothing about his relations or the lack of them with Spain (a point that deserves further contemplation, and which might be carried out in the future).

Yet, in reality, there is a hidden unknown, which is a bit of a time bomb that needs to be exploded. The reality is that Petrarch did not invent the sonnet that is famously attributed to him. What is commonly stated in less meticulous literary histories is inaccurate. There is a forerunner to Petrarch who came more than a century before him and who weaved poems which had the structure for which Petrarch became famous later. This time bomb of a poet was a citizen of Sicily and lived very close to the time when the Arabs lived in and ruled Sicily, and close in time to the Arab poet Ibn Hamdis al-Siqilli (possibly 1056 until around 1123). Indeed, the Sicilian poet lived in the very same region in which the Arab poet had lived, namely Syracuse. Ibn Hamdis was famous for the poetry he wrote in nostalgia for Sicily after departing from it. And a detailed study of his poems may reveal interesting links to it and to the circles in which he moved (something I hope to be able to carry out in the future).

The name of the Sicilian poet I have been talking about is Giacomo da Lentini (or Lentino);3 he lived in the court of Frederick II (1194–1250) and served for a period of time as a governor of a fortress called Carsiliato. More importantly, he was the leader of the first group of Italian poets and composed, together with other members of the group, thirty-one sonnets, twenty-five of which are attributed directly to him. There is no reason to believe at this point that anyone else before this group and Lentini had composed sonnets and it seems that he himself was the man who invented the sonnet, as a number of scholars believe. The sonnet was then transported into Italy wherein Petrarch and others learnt the new art and composed their own sonnets.

More important than all these facts, however, is the fact that Lentini lived in the court of Frederick II and worked as a court scribe for him. And the court of Frederick II was the very location in which flourished the genre of poetry known as ‘courtly love’, for which the troubadour poets were famous and which is directly taken from Arabic love poetry, as many scholars believe. In fact, Frederick himself, as we are told, was one of the pioneers of love poetry and the product of a milieu in which cultures like Arabic, Greek and Latin mixed. He was also one of the leaders of the Crusades and knew the Arab East very well. It is as important to know that Lentini himself and his group composed poetry in the local spoken dialect, breaking away from the Provencal language used at the time, and that the language of Sicily was clearly influenced by Arabic. It is likely that Lentini and his group were themselves influenced in their use of the local spoken dialect by what they had known about the use of colloquial Arabic in the genre known as zajal, for which Ibn Quzman was famous, as well as by the important and distinctive role that the kharja (the last line of the muwashshah) played in its construction: the kharja was most attractive when used in the colloquial. It is illuminating in this respect that many types of poetry came into being (in Sicily, Italy, France, Catalonia and Spain) in the local languages of these regions and appeared also in Hebrew in some of these countries, which are all strikingly similar to the azjal of Ibn Quzman in their structures and their rhyming schemes. Yet most important for the thesis I am presenting here is the fact that Jacopone da Todi (1236–1306) composed poems similar to the zajal, which were called laude4 in Italy. Todi might have belonged to the Sicilian school led by Lentini.

If it is true that Lentini borrowed the sonnet from Arabic poetry, then Shakespeare would owe a far greater debt to Arabic poetry and creativity than just the character of Othello in his great play. For Lentini’s impact was immense as the impact of the sonnet spread, and became so powerful first in Italy and France, then in England, wherein the sonnet which burgeoned with Lentini developed and was transformed to a point where it became known as the English or Shakespearean sonnet. And if many generations of European poets owe a debt to Lentini, then many more by now owe a greater debt to ‘Ubada Ibn Ma’ al-Sama’ (d. 1030) and other poets who invented the art of tawshih,5 and developed and refined it by composing dozens of variants of structures of the muwashshah, and making it one of the most exciting genres of poetry and singing for many centuries.

And I would like to entertain a thought, by way of prediction for the pleasure of prediction if not for anything else: to fantasize that the word ‘sonnet’ itself is of an Arabic origin. It is not known in Latin, and is said to be a word in the Provencal language. I suggest that it is a modification of the Arabic word simt, which means the same thing as a necklace, and was used in Arabic to describe a type of poetic structure, and I have seen it used to refer to the muwashshah itself. Simt is a most ancient poetic structure in Arabic and has highly distinctive features, especially as far as the rhyming scheme and the numbers of lines in it are concerned (see Appendix).6 Furthermore, the plural of simt is sumut/somot, and is there anything that sounds more like ‘sonnet’ than this?

Nor would I be shocked to learn at some point in the future that the word ‘sonnet’ is a slight corruption of the Arabic word nasat, deriving from listening as hearing; and listening was the foundation of singing for which muwashshahat were written, and without which the rhythm of the muwashshah often could not be preserved as regular. Insat is organically connected with muwashshah and corrupting nasat into ‘sonet’ is not strange, especially when we remember that the language of Sicily carries in its vocabulary a strong Arabic influence (in addition to that of Catalonian, Spanish, Latin and Greek). Nor is the degree of corruption from nasat to sonet striking when compared with the way in which the Arabic dar al-sina’ah has ended up as ‘arsenal’, al-taraf al-agharr as ‘Trafalgar’, al-Khwarizmi as ‘logarithm’ and dozens of Arabic words ending up in English and other European languages in forms that require a fortune teller to be able to discern the original Arabic in them (Averroes and Avicenna, etc.). So there is nothing unusual about nasat turning up as sonet or sonnet or about shaik esber turning up as shakesper, which is the way William Shakespeare’s name was written in his own lifetime. In fact, there would not be anything strange in seeing the Arabic word saot/sawt, turning up as sonet, and sawt was the very origin of poetry as song in the Arabic tradition as recorded by the great Kitab al-Aghani (The Book of Songs) by al-Isfahani, written in the tenth century and widely celebrated throughout the Empire, including Arab Spain. A sawt is a song rhyming as AAAA or AABA in its first lines.

Yet, the final hint of evidence that we can use to strengthen the possibility that the sonnet originated in Arabic poetry is the significance of the word ‘sonnet’ in its purely linguistic context. It is thought to come from the Provencal ‘sonet’ or Italian ‘sonetto’ meaning the little or small song. This suggests that the sonnet from its very inception in the work of Lentini and others was conceived as singing. And the muwashshah was pure singing; singing was and is its very essence and the means by which its aim and purpose was achieved and the mode by which its rhythm, if not prosodically precise, was turned into a proper rhythmical structure. And in the performance of the muwashshah as song, the goodness or weakness of it was/is measured. Here are some of the passages written by Ibn Sana’ al-Mulk (1155–1211) in his detailed description of the muwashshah:

The muwashshahat have no elements or rules of prosody except through setting to music and singing; that is the ‘arud (science of prosody) by which what is correctly weighed (metred) is known from what is broken, and what is perfectly measured from what has been exposed to deletion of phonetic elements (metrical units). Most muwashshahat are based on the composition of/for the urghun. And singing the muwashshah with another instrument is done by way of borrowing and majaz (transfer/metaphorical).

[There is] one type which cannot be set to music and be carried through it except by relying on a word which has no meaning and which will be a support for the tune and a stick for the singer.

And he (the composer of the poem) does what is not permitted to do and which cannot be carried by the setting to music, thus his scandal will be revealed when sung, for the singer using some instruments needs to change the tightness of the strings when he moves from the qufl (lock) to the bayt (house) of the muwashshah and when exiting from the bayt to the qufl; this is a position which needs to be remembered and taken note of.7

The last statement reveals the great degree of connection in the muwashshah between its composition as words and its singing and the impact this has on its structure.

In this process of composing and performing a song, the role of hearing and close, attentive listening is crucial. Insat thus is at the heart of the process; and if the muwashshah is nothing but song, those who would be listening to it in Spain and Sicily would have considered it merely as song. They might have understood its language, or failed to understand it, but they would be in a state of insat-nasat to its tune. Imagine a group of Arabs and non-Arabs listening and one is asking: what is this? The Arab would probably say ‘a song’ rather than a muwashshah because nobody really knew what muwashshah as a word meant; not even Ibn Sana’ al-Mulk or al-A‘ma al-Tutayli nor you nor I. This is a song, a simt, a small song, or a sonat=sonet=sonetto.

And the sonnet continued to be organically connected with music and singing even in Shakespeare’s time when it got more complex and philosophically and intellectually contemplative. Shakespeare’s Sonnet Two was famous as a song and there are many sonnets nowadays, Shakespearean and otherwise, set to music.

4

The muwashshah’s structure, as Ibn Sana’ al-Mulk has described it in what is thought to be the first definition of it in Arabic, is constituted in aqfal and abyat. And the degree of complexity varies from one muwashshah to another in their structuring of the qufl and the bayt. There are two distinct types of muwashshah: the complete (which begins with a qufl) and the aqra’ (literally, bold) which begins with a bayt. The complete type consists of six aqfal and five abyat. The aqra’ consists of five abyat and five aqfal. The minimum number of parts in a qufl is two and may go up to eight, rarely would it consist of nine or ten; the minimum number of parts in the bayt is three and may be in rare cases two, and can be of three and a half parts, and up to five parts. The part of a qufl is always singular/simple, whereas the part of a bayt can be single or compound. The compound only consists of two or three phrases, or faqarat, although on rare occasions it can have four faqarat.

The rhyming schemes in the muwashshah are of immense variety. One of these countless rhyming schemes is almost identical with the scheme we find in the first of the sonnets composed by Lentini and his group. In fact, if we write Lentini’s sonnets in the way used to write Arabic poetry generally (including the muwashshah), namely by considering each couplet to be one line with two hemistichs, it would become clear right away that the sonnet forms a simple type of muwashshah in which the qufl consists of two parts, each part consisting of three phrases, whereas the bayt consists of four parts with two phrases in each. What is remarkably significant also is that three of Lentini’s sonnets contain internal rhymes, i.e. they divide the line of poetry into what Ibn Sana’ calls tajzi’, or division into parts. This feature occurs frequently in the muwashshah both in the qufl and the bayt.

The present scheme of rhyming in the Lentini sonnet as written in the West is:

A

B

A

B

A

B

A

B

C

D

E

C

D

E

And there may be other variations in the last six lines.

If we rewrite that scheme in the Arab way, it will look like this:

AB

AB

AB

AB

CD

EC

DE

This scheme is no more than one of the simple schemes that the muwashshah may have in the aqfal and the abyat, where each bayt consists of four parts, each part of two phrases, followed by a qufl repeating what is present in the Lentini sonnet, as I have rearranged it. Should it appear that I am performing an unscholarly act by rewriting the sonnet as I did, I would refute that. To my great luck, after I had formulated my arguments, including rewriting the sonnet as I did, I found a scholar who has said the following: ‘The octet rhymes in every case ABABABAB, and was regarded rather as a series of four distiches than as a pair of quatrains’.8 That is exactly the way the bayt is composed in the muwashshah I translated above: ‘Rise early for the wine’.

This leaves me in no doubt that the sonnet written by Lentini was mirroring one or other muwashshah that he had learnt from his environment, which was strongly influenced by Arabic. Indeed, this degree of mirroring becomes sharper the more we explore Lentini’s sonnets and relate them to the muwashshah. In these sonnets the rhyming scheme of the sestet (the part consisting of six lines) differs from the one I described above and can be of any of these types:

CDCDCD

CDCCDC

ACDACD

which is a rhyming scheme that we find in the aqfal of the muwashshah, wherein we also find countless varieties which may well have formed the overall model (the paradigm) which Lentini learnt and of which he and his group borrowed one that they liked more; indeed, they may have borrowed most elements of the paradigm of the muwashshah and modified some slightly for one reason or another. And Allah knows best.

I shall now present samples of muwashshahat in which we see clearly that the structure of the sonnet composed by Lentini is no more than a chipped section of the muwashshah, which forms a bayt of four parts and two qufl each in two parts, each part in more than one phrase. The Arabic text of each muwashshah is quoted, but translating it will ruin the rhyming scheme; I shall transliterate the first muwashshah so that the reader can see the rhymes clearly.

Example 1:
Ma hawa mahasina l dahri illa ghazal
Mu‘riqul jaddayni min fahri ‘ammin wa khal
Nisbatan li l na’ili l ghamri wa li l nizal
Fa ana ahwahu li l fakhri wa li l jamal
Wajhuhu wajhun taliqli l duyuf mushriq wa yadun tastu ‘ala l usdi fa tafraq
ما حوى محاسن الدهر إلا غزالAB
معرق الجدّين من فهر عمّ وخالAB
نسبة للنايل الغمر وللنزالAB
فأنا أهواه للفخر وللجمالAB

وجهه وجه طليق للضيوف مشرق ويد تسطو على الأسد فتفرق

AB

AB

AB

AB

CDED

It can also be read as:
Wajhuhu wajhun taliqli l duyuf mushriq wa yad tastu ‘ala l usd fa tafraq
CDEFFE
This second reading produces a fourteen-line muwashshah with a striking kharja, metrically and grammatically.
Example 2:
بثنايا كالأقاحي فضحتْ سرّ المدامهAB
وقناع كالصباح غلبت ألف غمامهAB
فتنحّوا يالواحي واسألوا الله السلامهAB
فلها على الملاح بجمالها الإمامهAB
ريقها دار الامارة ثغرها عقد9CC
فلذا تصدّ تيها حين لا ترى شبيهاDD
أي حسن ما أجلا ونوال ما أقلاEE

AB

AB

AB

AB

CCDDEE

This is another, and an unambiguously so, fourteen-line muwashshah, most significantly closing with EE, as the Shakespearean sonnet was to do later.
Example 3:

بنت الكرم لها حسيس قد سمعته النفوس

منه نفسي تسمع أمره

بأن أمسي أشرب خمره

أذكي حسي منها بجمره

هذا عرسي شربت سره

على رسمي تجلى عروس لها الثياب كؤوس

ABB

CD

CD

CD

CD

ABB

This too consists of fourteen lines.
Example 4: Another type with three abyat:

سلطان الحسن جم الجمال طاغي التيه

جنات عدن في برده وما تكفيه

يسطو ويجني وبعد هذا درُّ فيه

مظلوم المسواك ثغر هداك بالابتسام إلى الغرام

ABC

ADC

AEC

FFGG

Significantly, this closes with GG too, but consists of thirteen lines.

Example 5: Another type with three-phrase aqfal:

رأيت ألف مليح ولا كهذا الرشا في الدلّ والغنج

دريتم من عنيت لم يدر إلا أنا

عنيت من قد جنيت من غصنها زهر المنى

وطالما قد ثنيت منها قواما ليّنا

ذاك القوام المروح سقوه حتى انتشى صرفا بلا مزج

ABC

DE

DE

DE

ABC

Example 6: And another:

قامة الغصن مالها مالت فيه من غير ريح

وكذا الشمس ما لها حالت عند وجه المليح

فاستمع للسماء إذ قالت فيه قولاً صحيح:

نور شمسي من وجه ذا منسوخ وهي أيضاً تقول

إن بدري لوجه ذا البدر خادم أو رسول

AB

AB

AB

CDE

CFE

The last two lines here are ambiguous; they can also be read as:

CDEF

CDGF

and forming fourteen lines.

5

I have taken these few samples from the book by Ibn Sana’ al-Mulk, but Arab poets have produced hundreds of muwashshahat in addition to the selection he recorded. Some of these may indeed be more closely similar to the structure of the sonnet, something I hope to be able to examine in the future. But it is very significant already that in Example 4 quoted above, we already have a muwashshah that closes with FFGG, which is close enough to the cap of the Shakespearean sonnet. I want to emphasize that normally muwashshahat are longer than the fourteen-line sonnet, but some of them in the way normally written in Arabic are not: I have found one by Ibn Sana’ himself consisting of fifteen lines; more strikingly, Shakespeare himself composed sonnet no. 99 in fifteen lines. Yet even more significant is that – in addition to the examples I gave above – I have found amongst the muwashshahat of Ibn Sahl al-Ishbili (d. 1251), a very famous poet known as the poet and muwashshih of Ishbilia (Seville), who was a contemporary of Giacomo da Lentini, one muwashshah which consists of seven lines in the Arabic system which is in fact fourteen lines if written as a sonnet. The rhyming scheme in this muwashshah is as follows:

AB

CB

DE

DE

DE

AB

CB10

One even more significant truth about the sonnet and the likelihood that it originated in the Arabic muwashshah (and zajal, possibly) is the very structure of the sonnet and its division into (or composition of) two sections both in Lentini’s version and in its Petrarchan version, a division that has remained valid in the composition and analysis of the sonnet in the languages that produced sonnets, including English, despite the many changes affecting the rhyming system. Thus, the sonnet is and was divided into two sections: the first called octet, the second called sestet. This division is no more than a mirror, or rather a copy, of the division of the muwashshah into aqfal and abyat.

Finally, and as a musk of the ending, the sonnet closes rhyming-wise with a couplet sharing the same rhyme, GG, which – as I have indicated – has something distinctive about it, not only rhyming-wise but often with reference to meaning or attitude, making it different in this fashion from the previous twelve lines, and coming like a fatlah or qaflah (like a ‘twist’ or a ‘lock’), which is called the volta. This feature is one of the most distinctive of the muwashshah in the aqfal section, on the one hand, and on the other in the famous kharja (possibly meaning ‘exit point’, deriving from khuruj, or ‘ornamental point’, deriving from takhrij). The kharja has always been described as the focal point and the pillar of the muwashshah, which enjoys qualities that are peculiar to it on the purely linguistic/phonetic level as well as on the semantic level. A crucial piece of evidence seems to me to reside in the connection, conceptually, between the kharja of the muwashshah and the volta of the sonnet; if this conceptual correspondence can be found to be expressed on a linguistic level, then a missing link of a Darwinian dimension can be discovered. And here is what it will look like: the word kharja, as I said, means ‘exit’: the phrase at which the composer will have completed his poem and ‘exited’ from it; the word volta seems to have the same significance in Italian, being also the last phrase at which the composer will ‘exit’11 from his composition, having completed and capped it. If this is true, and it seems to be so, then volta would have been no more than a direct translation by the early composers of sonnets of the Arabic word kharja.

Despite this thrilling ‘discovery’, however, I need to investigate further the semantics and nature of both kharja and volta at some point in the future. For the time being, it is sufficient to say that by the ‘nature’ of the volta and kharja, I mean the linguistic and artistic qualities each of them possesses as a closing element, or a closure, in their respective genres. In this respect, Ibn Sana’ al-Mulk’s description of the kharja plays a most illuminating role. Here it is in a relatively unpolished translation:

The kharja is the last qufl of the muwashshah; it has to be Hajjajiyya12 in triviality, Quzmaniyya in its linguistic deviation [from Fusha Arabic], hot, burning, sharp, ripening, [taken] from the speech of the masses or the special class (select group). If it is grammatically proper and woven on the model of the abyat and aqfal preceding it, the muwashshah ceases to be a muwashshah, except however if it is one in praise of some person and the name of this person is mentioned in the kharja … The kharja may be grammatically proper even without the name of the praised person but on condition that its words be highly erotic, bewitching, enchanting, magical and akin to love and yearning, which is an unachievable and highly demanding condition…

What is required, indeed imperative, in the kharja is that exiting towards it [or the approach path to it] be in leaps and digression and speech borrowed and attributed to [uttered by] some tongues (other than the composer?) either of those who can speak or those who cannot, or be about topics belonging to different species. It is mostly made to be uttered by boys and women and drunken males and females. The bayt which precedes the kharja must contain the words ‘he said’, or ‘I said’ or ‘he sang’ or ‘I sang’ or ‘she sang’.

The kharja is the mixed nuts and seeds13 of the muwashshah and its salt and sugar and musk and ambries; it is the final outcome and should be praiseworthy, and the seal [khatima] yet in fact it is the first, even though it is the last. I say the first because it is the item that must be born in the mind first. Thus, however the words and the metre come unto him [the composer] light unto the heart, elegant into the ear, naturally flowing into the Self, sweet unto the taste, he would handle it, craft it, treat it artistically, then upon it construct the muwashshah, because in it he has found the foundation and captured the tail and upon it erected the head.14

The significance of this description can be so crucial in the light of the fact that the Italian poet Jacopone da Todi (1226–1306), as I mentioned earlier, composed poems which resemble the azjal of Ibn Quzman (who is mentioned as a model for the kharja in the text quoted above), which were called laude. Examining Todi’s poetry and the properties of his language, especially this laude epithet, in the light of Ibn Sana’ al-Mulk’s description of the kharja may open new horizons for contemplation. Will Fate permit?

6

I shall now present the sonnet thought by one scholar to be the first to be composed by Lentini, in its original language first then in an English translation,15 hoping that the reader will be able to trace the rhyming scheme in it with ease.

Molti amadori la lor malatia
portano in core, che ‘n vista non pare;
ed io non posso sì celar la mia
ch’ ella non paia per lo mio penare:
però che son sotto altrui segnoria,
né di meve nonn-ò neiente a∙ffare,
se non quanto madonna mia voria,
ch’ ella mi pote morte e vita dare.
Su’ è lo core e suo sono tutto quanto,
e chi non à consiglio da suo core,
non vive infra la gente como deve.
Cad io non sono mio né più né tanto,
se non quanto madonna è de mi fore,
ed un poco di spirito è ‘n meve.16
In English:
Many a lover beareth his distress
Within his heart, away from others’ sight,
Yet can I not conceal my bitterness
So that my look shall not reveal my plight.
Another holdeth me in her duress,
And over mine own self I have no might
Save as my lady deigns to acquiesce,
Who giveth life and death as of her right.
Hers is my heart, hers am I all in all;
And he that hath no counsel of his heart,
Liveth in gentle company but ill.
Nor am I verily in life at all
Save through my lady, from myself apart,
And the mere breath that bideth in me still.
The rhyming scheme here is clearly ABABABAB CDECDE. One does not need to gaze long to realize that this structure is almost identical with the structures of muwashshahat I have described above.

7

To close with a more aromatic misk khitam, I would like to formulate my hypothesis in the following way:

The sonnet that was ‘invented’ by the Sicilian Giacomo da Lentini, is no more than a ‘chip’ or a ‘block’ severed from the structure of the type of muwashshah called aqra‘ (‘bold’) by Ibn Sana’, which begins with a bayt (rather than a qufl) consisting of four parts, each of two phrases, followed by a qufl consisting of two parts, each of three phrases. In such a way, the bayt will be one in metre but not in rhyming with the other abyat of the muwashshah whereas the qufl will be normally different in rhymes from the other aqfal. In this way, a varying system of rhyming appears in the section of six lines in the muwashshah, i.e. in the qufl and in the sestet of the sonnet. And because the variations in the qufl are many as far as rhyming is concerned, it is the case that the variations in the sestet of the sonnet are many. Here is a description of what I am saying:

Octet:

AB

AB

AB

AB

Sestet: CDE, as well as any of the following mathematical possibilities:

CDE

CED

DCE

DEC

ECD

EDC

CDC

DEE

EFD

FDE

EGC

What is most exciting is the fact that when I wrote the Arabic text of this study back in 2007, I had only identified one muwashshah which closes with FFGG, but having since read and re-read many muwashshahat, I have found many in which the qufl actually closes with the rhyming couplet GG. This is a hugely significant type17 although it is rare and is called by Ibn Sana’ shadh jiddan (‘very rare, unusual’, ‘irregular’ or ‘out of the way’) because its bayt consists of two compound parts, each consisting of two phrases. The cap of this type of muwashshah is the cap that appeared as a fundamental and distinctive feature in the development of the structure of the English or Shakespearean sonnet centuries later.

8

Now the idea of severance (or partitioning or detaching) may appear odd but it is important for my present thesis: the muwashshah is a larger structure than the sonnet and this may be the secret behind calling the latter a sonnet – a small song (sonet; sonetto) – when it was ‘discovered’, rather than calling it simply a song or a large song. Indeed, it is a small song in relation to the muwashshah, because it is no more than a part chopped out of its five possible parts. This may well be a decisive piece of evidence in supporting my argument. Yet only God and Giacomo da Lentini know the truth of what happened in history, while I remain steeped in my ignorance.

Prologue

The thesis I have presented here may strike some as being far-fetched and the idea that Lentini or anyone else in his time could have seen or heard the Arabic muwashshahat and azjal written and sung, then out of one or more examples borrowed the basic structure and maintained the basic rhythmic and rhyming scheme, but introduced variations here and there, especially in the sestet, may appear to some to be even more far-fetched. For me personally, neither of these ideas has the slightest degree of strangeness or unfamiliarity if we look at the history of literatures throughout the world. Stranger things happened when the Arabs translated Greek science, philosophy and literary texts, such as Aristotle’s Rhetorics and Poetics. Closer to home, Petrarch himself borrowed Lentini’s structure of the sonnets and introduced his own variations. In England, the sonnet inherited from Petrarch underwent radical transformations (which included Edmund Spencer’s version, that goes ABABBCBCCDCDEE, which is more strikingly like a real Arabic muwashshah) then settled down into its Shakespearean format.

Yet, the story did not end there: despite Shakespeare’s dominant position in English literature, right after him and up to our own days, new variations and various ways of structuring the sonnet have taken place. Some of these have been mentioned in the Arabic version of my book, but the English reader will probably not need to be fed this piece of information. If my own personal experience could be of any interest, it is indicative of the ease with which a poem, being written in full awareness of a model, and consciously trying to emulate that model, can overflow the model spontaneously, and end up making significant changes to the structure of the model it had aspired to emulate. (In my case, producing the structure ABABCDCDEFEFGGHIHIIH.) Examples of such an act can be easily multiplied and shown to occur not only in the genre of the sonnet but also in all aspects of creative activity and artistic production. Edward Said’s brilliant article on the migration of theory demonstrates this so eloquently and convincingly.18

Appendix

The reader of Arabic will see the significance of these examples of the musammat structure and its definitions immediately, I hope. For the reader who cannot read Arabic, translating these examples will not be very useful, but seeing their scheme of rhyming will carry considerable significance. Even the word musammat itself carries such significance, as it is the past participle of sammata (to make a simt), thus suggesting that the later structure was called muwashshah with full knowledge of musammat and taking the musammat and the art of tasmit as a model to produce muwashshah and tawshih. The actual difference between simt and wishah (‘shawl, sash thrown over the shoulder and chest’) clearly underlies and underpins the coinage of tawshih and traces the process of development from tasmit to tawshih. The physical appearance of simt is similar in fact to a necklace, whereas the physical appearance of wishah may have inspired the layered or striped structure of the muwashshah into aqfal and abyat, each with their own ‘colours’ of rhythm and rhyme. Who knows?

The musammats of ‘Umru’ al-Qays:

  1. 1)AABBBBA
  2. 2)AAAAB
  3. 3)simpler examples with this scheme:AAAB

Kamal Abu-Deeb

Oxford

23 May 2016

Notes

William Shakespeare, Al-Sūnītāt al-Kāmila bi-l-‘Arabiyya wa-l-inklīzzīyya [The Complete Sonnets in Arabic and English], trans. Kamāl Abū-Dīb (London: Dār al-Sāqī, 2011).

1

See on this topic Sigrid Hunke, Allahs Sonne uber dem Abendland: unser arabisches Erbe (Frankfurt am Main: Fischer Bucherei, 1965), translated into Arabic as Shams al-‘Arab Tasta’ ‘Ala al-Gharb by Faruq Baydun and Kamal Dasuqi, 8th edn (Beirut: Dar al-Jil & Dar al-Afaq al-Jadida, 1993).

2

In other words, qufl suggests locking the poem into a rhythmic or rhyming scheme to be followed by other sections in the poem occupying a corresponding position to the first qufl.

3

In fact, this text originally referred to the poet as da Lentino; Abu-Deeb has, however, given editor Katherine Hennessey permission to substitute ‘da Lentini’, the form most commonly found in Italian language texts.

4

See references to these aspects in Henk Heijkoop and Otto Zwartjes, Muwashshah, Zajal, Kharja: Bibliography of Strophic Poetry and Music from al-Andalus and Their Influence in East and West (Leiden: Brill, 2004), ix–xvi.

5

The birth of the muwashshah remains obscure; some date it to the ninth century. See on this Jawdat al-Rikabi’s introduction to his edition of Ibn Sana’ al-Mulk’s Dar al-Tiraz fi ‘Amal al-Muwashshahat, 2nd edn (Damascus: Dar al-Fikr, 1977), 12.

6

The Appendix gives a number of examples of musammat poems composed by Arab poets, including even the great pre-Islamic ‘Umru’ al-Qays.

7

The three passages can be found in Ibn Sana’ al-Mulk, Dar al-Tiraz, ed. al-Rikabi, 47–50.

8

Ernest Wilkins, ‘The Earliest Sonnet’, in Studies in Petrarch and Boccaccio, ed. Aldo S. Bernardo (Padova: Antenore, 1973), 22.

9

A word is missing here but it ends with the sound sequence ‘arah as the rest of the muwashshah shows, i.e., CC.

10

See Yusri ‘Abd Alla, ed., Diwan Ibn Sahl al-Ishbili (Beirut: Dar al-Kutub al-‘Ilmiyya, 1988), 37.

11

I owe the revelation of the meaning of volta to my friend and colleague Tullio Lobetti with whom I have discussed a number of points on the sonnet related to Italian literature; his valuable ideas have helped me formulate a couple of points while preparing this English version of my study. One of his most crucial items of information is that when Italian students are taught the sonnet at school, they are taught that it consists of a number of elements that end with an ‘exit’. How exciting! However, even in its sense as ‘turn’, volta appears to me to be a reflection of the kharja.

12

Ibn al-Hajjaj is one of the most formidable poets of the ‘Abbasid age in Baghdad, but his vulgar language, shocking vocabulary of homosexuality and such items gave him a bad name. His Diwan is totally un-publishable even today, with all the libertarianism of the postmodern, postmoral, postALL world.

13

Nuts and seeds (abzar) of various plants such as melon and pumpkin are roasted and served with wine; without them, drinking is not much of a pleasure to the connoisseur. However, abzar also means ‘spices’.

14

See the full description in Ibn Sana’ al-Mulk, Dar al-Tiraz, ed. al-Rikabi, 40–44.

15

Both taken from Wilkins, ‘The Earliest Sonnet’, 22.

16

Katherine Hennessey has, with Kamal Abu-Deeb’s approval, edited the Italian text of this sonnet, following the text provided in I poeti della scuola siciliana, vol. I°: Giacomo da Lentini, ed. Roberto Antonelli (Rome: Bulzoni, 1979, since reprinted by Mondadori/Giulio Einaudi, 2008 and 2009).

17
Another very important type is evident in muwashshah no. 32 in Dar al-Tiraz, which can be described as follows:

AA

BBB

AA

CCC

AA

DDD

AA

The degree of variation in this type is significant and its closure with AA is particularly interesting.
18

In ‘Travelling Theory’, Said demonstrates that theories which are formulated in one culture and adopted or used by an individual in another culture undergo considerable transformations and acquire qualities which are determined by various conditions that prevail in the receiving culture. He traces some of the most influential theories in the humanities in their travels to illuminate his argument. See The World, the Text, and the Critic (Boston: Harvard University Press, 1983).

If the inline PDF is not rendering correctly, you can download the PDF file here.

Contributor Notes

Kamal Abu-Deeb is Emeritus Professor of Arabic and Comparative Literature at SOAS, University of London. He has taught at Oxford, Pennsylvania, Berkeley, Columbia and Dartmouth as well as at Arab universities. His recent publications include a philosophical/political text on freedom, The Book of Freedom (in Arabic, 2012), and a complete translation into Arabic of Shakespeare’s Sonnets (2011). In 2016 he received the Sultan Bin Ali Al Owais Cultural Award. He lives and works in Oxford.