David Jalajel

This article is an attempt to address a gap that I have observed in how the ghazal form is being discussed by poets writing in English. Although most of the poets writing in the field acknowledge the ghazal to be as much an Arabic art form as it is as a Persian or Urdu one, they focus almost exclusively on Persian and Urdu poetic structures. 

The Arabic poetic form differs in a number of ways. For one thing, the Arabic ghazal does not have a repeated word at the end of the line. It merely requires monorhyme (qâfiyah). Arabic is one of the easiest languages in the world to rhyme in, and a single poem can keep up the same rhyme scheme for hundreds of lines without sounding tiresome. 

Another significant way that Arabic poetry differs from its Persian and Urdu counterparts is in the role that Arabic poetry recognizes for the long line (the bayt, which is equivalent to the Persian sher or couplet). Arabic poetry does not see each long line as needing to possess absolute autonomy. Most Arabic poems have carefully constructed arguments and elaborate descriptions that develop over the course of many lines, though each line is more often than not an independent grammatical unit. 

Even this grammatical autonomy does not always hold true. Enjambment exists in Arabic poetry between the long lines, and mild forms of enjambment are not all that infrequent. The lack of autonomy of the individual long line is at its most evident when enjambment is present, and this is what the article seeks to explore. 


The Structure of the Arabic Poetic Line

The Arabic poetic line is a long, complex, metrical unit called a bayt. This unit is divided into two parts, which are metrically either identical or near-identical. The half-line is variously referred to as a shatr (half-line), a misrâ` (hemistich), or a qasîm (division). These are synonymous terms. 

The first half-line is referred to as the sadr (precedent) while the second half-line is called the `ajuz (posterior). The terminal foot of the first half-line (sadr) is called the darb. The terminal foot of the second half-line (`ajuz) is called the qâfiyah, which is incidentally the very term used for end-rhyme. 

A line of Arabic poetry can be written in one of three ways: as a single line, as a line with a gap between the hemistiches (which is most common), or as a couplet (which is resorted to only when the page is too small to accommodate a long line). 

It will be easier to illustrate these three approaches with examples in English rather than with translations or transliterations of Arabic verse, since we are concerned here with the visual presentation and typesetting of the line. Sir Philip Sydney will help us by providing some alexandrines from Astrophel:

And all with teares bedeawd, and eyes cast vp on hie,
O help, O help ye Gods, they ghastly gan to crie.

If this were an Arabic poem, the lines would usually be written with a gap at the medial caesura:

And all with teares bedeawd,                    and eyes cast vp on hie,
O help, O help ye Gods,                  they ghastly gan to crie.

. . . or, if need be, in the couplet form that we are used to seeing English ghazals in:

And all with teares bedeawd,
and eyes cast vp on hie, 

O help, O help ye Gods,
they ghastly gan to crie.


Enjambment Between the Hemistiches

The metrical structure of the bayt is such that the two hemistiches produce a natural place for a caesura right in the middle of the line. The Arab poet has many options as to how to exploit this caesura artistically. The poet may choose to emphasise the ever-present metrical division with a syntactical one. Often this is the case, and this leads to the line resembling something like a couplet. This approach is developed much further in Persian and Urdu ghazals, as well as in the English ghazals based on those forms. 

The poet may decide to not only bring a syntactical pause at the metrical caesura, but to employ internal rhyme to further emphasise the caesura. Andalusian poets were particularly fond of the technique of rhyming the darb of the first hemstitch of each successive line on a different sound than that of the qâfiyah, creating an effect we would recognize as a set of couplets with an A-B A-B rhyme scheme. 

The poet may take the opposite approach and write the line as a single, flowing syntactical unit with no syntactical pause, or go even further and place syntactical pauses at other places in the line and thereby use the syntax to totally obscure the location of the metrical caesura. 

This is all a routine question of rhythm and pacing for the Arab poet. How the Arab poet chooses to approach the syntax of the line is purely a question of style. 

When the Arab poet chooses not to bring a syntactical pause at the metrical half-line, we would interpret this as enjambment if we conceive of the Arabic bayt as a couplet, and that is how we will analyse it if we write Arabic-style ghazals in English and choose to arrange the long lines as couplets. However, it is more accurate to say that the Arabic bayt is a single long line whose metrical structure is highly conducive to a medial caesura – but a caesura that is entirely optional for the poet. 

There is one practice in Arabic poetry, however, that deserves our special attention. This is tadwîr (pivoting), the very common situation where the hemistich comes to an end in the middle of a word. Lines that exhibit this feature, nevertheless, are visually represented in the same three ways as shown above. Sir Philip Sydney will again help us to illustrate this: 

Tho with benign aspect sometime didst vs behold,
Thou hast in Britons valour tane delight of old,

These verses could be written with a gap between their metrically equal hemistiches, necessitating that we break a word in half:

Tho with benign aspect        sometime didst vs behold,
Thou hast in Britons va-        -lour tane delight of old,

Or in the couplet form:

Tho with benign aspect
sometime didst vs behold, 

Thou hast in Britons va-
lour tane delight of old,

The second and third arrangements may seem odd to us, but this is exactly how Arabic poems are type-set and presented. 

Tadwîr is quite normal in Arabic poems. In some poems, it is the exception. In others it is the norm. There are poems where every single line exhibits tadwîr. There are others where it is entirely absent. It is purely a matter of style. 


Enjambment Between the Long Lines

The formal term for this kind of enjambment is tadmîn. 

It is quite common to find examples of mild enjambment in Arabic poetry, for instance, where the first line is a conditional clause that is resolved in the second, making a complete sentence. 

We see this in lines three and four of al-Shâf`î’s famous Da`il-Ayyâm:

If your faults become widely publicised
And it would please you to have them covered up, 

Then conceal them by being generous, since every fault
Is obscured – as they say – by generosity.

This admittedly non-poetic translation faithfully shows how the sentence’s grammatical structure relates to the two lines of the poem. I have opted to present the long lines as couplets, breaking them at the hemistich, to aid in the comparison with the Persian and Urdu forms that English readers are generally more familiar with. It is nevertheless important to keep in mind that each couplet actually represents a single line (bayt) of Arabic poetry. 

Mild enjambment of the kind exhibited in al-Shâf`î’s poem is extremely common. Al-Zinjânî says: “It is rare to find a poem that does not exhibit this type of enjambment.”[1] 

Radical enjambment is far less common. This is due more to poetic expectations about what sounds right as opposed to a formal rule. When radical enjambment does take place, it is clearly a foregrounding technique because of its rarity. It is not, however, a violation of any poetic rule of form. It is rather regarded as a defect of rhyme, and is indeed discussed in Arabic books of poetics along with slant rhymes. 

Most poetry theorists do not distinguish between mild enjambment and radical enjambment in the their discussions of tadmîn. They simply distinguish between those cases they find distasteful and those they do not. 

There is, however a lot of variation between theorists in how they define tadmîn. 

Al-Jurjânî defines tadmîn in poetry as: “where the meaning of the line is inter-related with that of the line before it so that it does not stand correctly on its own.”[2] 

Al-Qâdî Abû Ya`lâ al-Tannûkhî defines it as: “the completion of a line’s meter before its meaning is complete.”[3] 

These definitions are very broad and conform with the idea that most cases of tadmîn fall into the category of enjambment that we would call simple or light enjambment. 

Ibn `Abd Rabbihi, the author of al-`Aqd al-Farîd, provides us with two very different definitions of enjambment in the same work. 

In one place, he defines tadmîn as: “where the first line is connected to the second so that its meaning is incomplete without it.”[4] 

Elsewhere, he defines the term as: “where the terminal rhyming word (qâfiyah) is not independent of the line that follows it.”[5] 

His second definition of tadmîn is a far narrower definition, since it restricts the scope of tadmîn to cases where the final word or phrase of the line is directly dependent on what comes after it. This is close to the concept in English of radical enjambment. 

He immediately follows this definition with an example of this kind of enjambment from the poetry of al-Nabighah al-Dhubyânî:

And they watered their herds at the wells of Jifârin in spite of Tamîm
As they were the victors on the day of `Ukâz, and I 

Extolled to them many of their good victories
That let them know of the love in my heart.

The translation of these two long lines is faithful to the syntax of the original. Here the rhyme in the first bayt is on the words “and I” which is the subject of the verb that starts the second bayt. 

After citing this example, he reverts back to the former, more general definition when he comments on it, saying: “This is distasteful, because the first line is tied in with the second so that it does not stand on its own, and this is something quite prevalent in poetry.” 

The narrower meaning also features in al-Khafâjî’s definition of tadmîn: “where the rhyming word does not stand on its own in its meaning until it is connected to the first words in the following line.”[6] 

We can see that the term tadmîn is discussed in Arabic poetics alternatively as something referring to the overall meaning of the lines involved, which is a broad definition encompassing all forms of enjambment, and as a term involving the terminal rhyming word of the first of the two lines, which gives a far narrower and more restricted meaning to the term. 

It is important to remember that the issue of tadmîn is addressed by poetry theorists in discussions regarding the defects of the terminal rhyme. Definitions of tadmîn, however, vacillate between addressing the overall meaning of the lines and those that focus on the terminal rhyme word itself. 

I believe this confusion regarding the definition of tadmîn arises from the fact that the Arabic academic tradition had long ago pigeonholed the topic of enjambment as an aspect of the study of rhyme and not that of versification. Therefore, when Arab scholars wished to speak about enjambment in broader terms, they had the difficulty of doing so within the narrower framework of rhyme elements, leading to the confusion that we see. 

The Attitude of Arab Theorists towards Tadmîn

The classification of tadmîn as a “defect of the terminal rhyme” shows that, traditionally, Arab poetry theorists frowned upon tadmîn. However, they were generally in agreement that as long as each line is syntactically complete, it is not at all objectionable if the following line elaborates or clarifies the meaning of the former one. This is stated by a number of theorists, including al-Shantrînî[7] and al-Tabrîzî[8]. 

The theorists only really express disapproval where the individual long lines are syntactically incomplete. In truth, even then, this disapproval is more a theoretical ideal than an actual reflection of poetic practice. Ibn `Abd Rabbihi betrays this fact when he says after citing al-Nabighah al-Dhubyanî’s example: “. . . and this is something quite prevalent in poetry.” 

However, Ibn `Abd Rabbihi also says: “A verse is only praiseworthy if it can stand on its own.”[9] 

Here we witness a dichotomy between theory and practice that goes back to the very first theorist on Arabic poetics, al-Khalîl (d.786 AD). His attitude, though, was already at variance with practice. 

Amidu Sanni points out that the reason why the theorists always cite al-Nâbighah al-Dhubyânî’s lines is not because of a paucity of other equally good examples from the pre-Islamic era, but because of his high ranking among the poets of his day. He writes: “. . . what makes this particular example a popular illustration is, perhaps, al-Nâbigha’s position as a literary archon at the `Ukâz fair, to whom other poets submitted their works for assessment.”[10] 

Due to the prevalence of tadmîn among even the best of the early masters, theorists following in the footsteps of al-Khalîl had to find ways to explain its presence, usually by grading instances of tadmîn on a scale of acceptability. In brief, the general attitude of these theorists is that the smaller the syntactical bond is between the two lines, and the more distant the syntactical dependency is from the first line’s rhyming word (qâfiyah), the more acceptable the enjambment is. 

Regardless of the discomfort it caused the theorists, the frequency of tadmîn only continued to increase in the developing urban Arab society. Within a century of al-Khalîl, theorists began changing their minds about enjambment. Ibn Kaysân (d. 911 AD) declared the fault is but a trifling one, albeit one to be avoided, especially by outstanding poets in whom it would betray remissness and poor artistry.[11] 

Further strain was soon put on the critics as the urban and literate Arab poets started experimenting more and more with foregrounding techniques and began going out of their way to produce clever enjambments. 

Amidu Sanni summarizes the ensuing critical crisis quite nicely:[12]

One interesting aspect of the debate concerns the deliberate (rather than inadvertent) use of tadmîn: that is, when it is employed as a conscious device, becoming in extreme cases, an all-pervading feature of a given piece. The attitudes of the theorists to this vary widely. An erotic piece belonging to the love-poet, `Umar b. Abî Rabî`ah, but variously attributed to many others, is an example of enjambment par excellence. The opening line ingeniously dovetails into the second, the second line into the third and so on till the end of the piece. In the view of Ibn Kaysân, such a deliberate commission of tadmîn should by no means be regarded as a fault, but rather, a demonstration of skill and creative effort. Similarly, Muhammad b. Dâwûd al-Isfahânî (d. 910 AD) is reported to have considered the enjambed piece so charming because “it deviates from the well-known”, that he incorporated it into al-Zahra, his magnum opus.

Contrarily, al-Sûlî (d. 946 AD), and following him, al-Marzubânî, consider the piece crassly inartistic. He ultimately concludes:[13]

. . . what was adjudged to be one of the most important prosodic defects – with not a few illustrations from the pre-Islamic corpus – gradually evolved into an acceptable device in poetic praxis . . . tadmîn gradually developed from an objectionable blunder into a belle laide admired by poets and critics alike.
Be that as it may, in a tradition-oriented society like that of the Arabs, many theorists in their works on poetics persisted in enumerating tadmîn as a defect of rhyme, giving a more and more half-hearted lip service to the idea of its being a defect. This has at times led to some very odd conclusions, which show just how much theory and practice could become divergent on the issue. 

For instance, Mahmûd Mustafâ, an early twentieth century Egyptian scholar who in 1931 founded al-Azhar’s Arabic language faculty, discusses in his textbook on scansion what he considers to be “good” and “bad” enjambment. He disapproves of enjambment wherever the first line is not a complete sentence, which would include many of the most mild and ubiquitous cases of enjambment in Arabic poetry, like al-Shâfi`î’s example where the second line resolves a conditional clause set forth in the first. 

At the same time, he sees no objection to any instance of tadmîn where the first line is a complete sentence, even in cases where the second line begins with a prepositional phrase that is utterly dependent on the first. We would rather regard this as a very strong form of enjambment. 

The example he gives for this “good” enjambment is: “. . . she did not think / about anyone more amorously than she did of me . . .”[14] 

Essentially, since the qâfiyah ends on “she did not think” which is a complete sentence, the enjambment between the verses is not problematic. 

The twentieth century also witnessed other theorists who chose, in hindsight, to view enjambment as an unqualified strength of Arabic poetry. Another Egyptian scholar, Dr. Ahmad Kishk, writes that enjambment: “. . . affirms that our Arabic poem is not something fragmentary and disjointed, but possesses the organic cohesion to make it both an artistic and expressive whole.”[15] 


  • [1] al-Zinjânî, `Abd al-Wahhâb b. Ibrâhîm b. `Abd al-Wahhâb al-Khazrajî. Mi`yâr al-Nazzâr fî `Ulûm al-Ash`âr. Dr.Muhammad `Ali Rizq al-Khafâjî (ed.). Dâr al-Ma`ârif, p. 104. 
  • [2] al-Jurjânî, `Alî b.Muhammad al-Sharîf. al-Ta`rîfât. Dr. Muhammad b. `Abd al-Rahmân al-Mar`ashlî (ed.). Beirut: Dâr al-Nafâ’is (2003), p. 123. 
  • [3] al-Tanûkhî, Abû Ya`lâ `Abd al-Bâqî `Abd Allah. al-Qawâfî. Dr. Muhammad `Awnî `Abd al-Ra`ûf (ed.). Cairo: Dar al-Kutub wal-Wathâ’iq al-Qawmiyyah (2003), p. 202. 
  • [4] al-Andalusî, Ibn `Abd Rabbihi. al-`Aqd al-Farîd. al-Warrâq, vol. 2 p. 343. 
  • [5] ibid. vol. 2 p. 385. 
  • [6] al-Khafâjî, Muhammad `Abd Allah. Sirr al-Fasâhah. al-Muta`âl al-Sa`îdî (ed.) Cairo (1952), p. 219. 
  • [7] al-Shantrînî, Abû Bakr Muhammad b. `Abd al-Malik. al-Kâfî fî `Ilm al-Qawâfî. Dr.Muhammad Ridwân al-Dâyah (ed.), p. 104. 
  • [8] al-Tabrîzî, Abû Zakariyyâ al-Khatîb. al-Kâfî fî al-`Arûd wa al-Qawâfî. al-Hassânî Hasan `Abd Allah (ed.). Cairo: Maktabah al-Khanjî (1994). Beirut: Dâr al-Anwâr (1968), p. 167. 
  • [9] al-Andalusî, vol. 2 p. 343. 
  • [10] Sanni, Amidu Ahmad. “On Tadmîn (Enjambment) and Structural Coherence in Classical Arabic Poetry” Bulletin of the School of Oriental and African Studies, University of London, Vol. 52, No. 3. (1989), p. 464. 
  • [11] ibid. p. 464 
  • [12] ibid. p. 465 
  • [13] ibid. p. 466 
  • [14] Mustafa, Mahmûd. Ahdâ al-Sabîl ilî `Ilmay al-Khalîl. Dr. Muhammad Ahmad Qâsim (ed.). Beirut: al-Maktabah al-`Asriyyah (2005), p. 123. 
  • [15] Kishk, Ahmad Muhammad `Abd al-`Azîz. al-Qâfiyah Tâj al-Îqâ` al-Shi`rî. Cairo (1983), p. 87.