The Arabic Qâfiyah & English Rhyme – The Use of “Microrhyme” for Adapting Arabic Poetic Forms into English
by David JalajelThe word “qâfiyah” is etymologically of Arabic origin. It literally means “that which follows after.”
The term qâfiyah in Arabic poetics is defined in different ways. Some definitions are highly theoretical and discuss other quantitative aspects of the end of the verse and not just the rhyming sound. Other definitions are more straightforward, like Ibn Kaysân’s: “Everything that must be repeated at the end of each verse.”
The Persians adopted the word qâfiyah into their own literary discourse, just like they adopted a large number of other Arabic poetic terms. However, the Persians had their own understanding of how these terms applied to Persian poetry and fitted in with Persian sensibilities – a process that English-speaking poets who write in foreign forms, like the Persian ghazal or the Japanese haiku, should be quite familiar with.
Rhymes in Arabic can be short and simple or long and complicated. In all cases, however, they centre on a single letter, called the rawî. Some short rhymes based on the rawî “b” would include (bi) (ba) and (b) on its own.
Al-Mutanabbî (915-965 AD) – the so-called “Shakespeare” of the Arabs – writes:
Fahimtul-kitâba abarral-kutub * fa-sam`an li-amri amîril-`arabI divided the long line into its two component hemistiches by using a gap and an asterix to distinguish them. I did not bother with couplets, since the Arabic poet only writes the bayt (verse) in two lines if the page is not wide enough to fit the whole bayt on one. And then the poet – or more often the typesetter – sometimes has to split a word between the two lines.
As we can see in the example above, the only common denominator with respect to sound is the “b” sound itself, which is the rawî of the poem’s qâfiyah. This sounds like a full, perfect rhyme to the Arab ear. Actually, the metrical uniformity of the final words helps to make the rhyme work – the terminal “b” in this qâfiyah is always preceded by a succession of three consonants, the first un-voweled, followed by two consonants each having a short vowel. Of course, this consistency is part of the quantitative Arabic meter, though Arabic poetic theorists include these considerations in the definition of the qâfiyah.
In the example above, we also see the optional convention of tasrî`, which is to have the first verse of the poem employ the rhyme in its first hemistich – something it seems to me Persian poets adopted into their initial couplets. However, in the Persian couplet (sher), it is an end rhyme, whereas in an Arabic bayt, it might be better understood as a leonine or medial rhyme.
Another nice short qâfiyah for us to look at is (di) with the condition that it must be preceded by an un-voweled consonant. The fact that the preceding consonant must be un-voweled is part of the qâfiyah, but that consonant can be any letter, as we can see in the following verses by al-Mutanabbî:
Wa man yashabi-sma-bnil-`amîdi muhammadin * yasir bayna anyâbil-aswâdi wal-usdiThese are also perfectly good rhymes to an Arab ear.
The following verses by Abu al-Atâhiyah (d. 825) exhibit the long (ee) sound as the rawî of its qâfiyah, though this is quite rare:
Narûhu wa naghdû li-hâjâtinâ * wa hâjatu man `âsha lâ tanqadîThere are dozens of other qâfiyahs exhibiting short rhymes like these. What I have cited is just for illustration. They are not in the least bit unique.
This is why I regard the English slant rhymes that I use in some of my poems – like (er) (en) and (ee) – as being consistent with the Arabic qâfiyah. Of course, these are not necessarily the same rhymes that are used in Arabic, since the dictates of each language are different. However, these rhymes are similar to the Arabic rhymes in the function that they play in the poem. The condition I place on these rhymes, to conform more closely with Arabic rhymes, is that the repeated sound should always be on the end of the word and it should be either a vowel-consonant combination like (er) or (en) or a stressed terminal vowel like (ee). I call these “microrhymes”.
Also, for English poets who wish to follow the Arabic form as closely as possible, I find such terminal slant rhymes (or “microrhymes”) necessary for practical as well as aesthetic reasons.
Practically, it is the only way I can see in English – a notoriously difficult language for rhyming in – to keep a monorhyme going for dozens of lines, which is the case with Arabic poems. They can be quite long. (And we must remember, the Arabs HATE repeating the same rhyming word in successive lines of poetry. The same word should not reappear in the qâfiyah of a single poem unless a good number of verses have passed between it and its previous appearance.)
There are few English full rhymes that are ubiquitous enough to be used as a monorhyme for a long poem – (air) being one of them, which I use in my poem “A Frog”. I also got some mileage out of alternating the sounds (ease) and (eaves), which though technically a slant rhyme, is a very subtle one to the ear and has relatively the same aural strength as a perfect rhyme. This can be seen in my poem “qasîda #1”.
Therefore, I feel it is better in general to resort to those terminal sounds that, for grammatical reasons, tend to repeat themselves in English quite often, like (er) for the active participle, (en) for the passive participle, and (ee) for the adverb. These can then be mixed with a few other words that are not within those grammatical categories but have the same end sound – like (cylinder), (chicken) and (tea) – to provide a robust and varied set of rhymes for a poem.
On a purely aesthetic level, microrhymes seem to work better for longer poems. Using perfect rhymes as monorhymes can become obnoxious to the English ear after a while, because rhyme is such an artificial device in English verse. In Arabic, the rhymes are so close to natural speech patterns that they are no more obtrusive as a formal poetic device than iambic pentameter is in English.
The microrhyme, being a type of slant rhyme, is subtle enough to remain pleasant to the ear over the long haul, while being consistent enough to act as a unifying formal device. Even then, it can sometimes become too much. The rhyme on (ee) can sound obsessive after a while – a quality that I tried to exploit to my advantage in “The Tea Party”.
Defects of the Qâfiyah (`Uyûb al-Qâfiyah) – Arabic Slant Rhyme?
All works of Arabic poetics must include a chapter on the defects of the qâfiyah. Leading the list of defects in this chapter we usually find the truly despised repetition of the rhyming word. This is followed by a discussion as to whether rime riche is to be considered a defect as well, and an enumeration of the contexts in which word repetition it is acceptable for stylistic reasons.
The second defect is enjambment. Though theorists list it as a defect, they confess that it is “extremely common in poetry.” It is, moreover a “defect” that poets in every age seem to have taken a singular pride in committing and that anthologists have taken great care in showcasing.
These two defects are followed by a bewildering categorization of various kinds of “defective rhyme” or slant rhyme. We might be surprised at this, after seeing how brief the Arabic rhymes can be, and how they are already comparable in sound to English slant rhymes. Nevertheless, Arabic poets manage to slant even these.
We must remember that the qâfiyah is always centred on a single letter called the rawî, which is then often associated with other letters and sounds, as well as with certain metrical conditions. Some slant rhymes entail retaining the rawî while changing some of the other sounds. A few categories of slant rhyme entail changing the rawî itself while retaining its surrounding structures. These categories of defects have various degrees of acceptability in Arabic poetics, though most of them can be supported with a liberal number of examples from the earliest Arabic masters.
One poet who was notorious for his use of slant rhyme is al-Nâbighah al-Dhubyânî (535-604? AD), whose following verses are often cited as a prime example of one kind of slant rhyme known as iqwâ’, where the rawî of the qâfiyah takes variously the vowels (u) or (i) after it:
Amin âli mayyata râ’ihun am mughtadî * `ajlâna dhâ zâdin wa ghayra muzawwidiAnd elsewhere in the same poem:
Saqatan-nasîfu wa lam turid isqâtahu * fa-tanâwalat-hu wattaqatnâ bil-yadiThis is an extremely common form of slant rhyme in Arabic poetry. The language scholar al-Akhfash says: “Scarcely has a qasîdah been free of it.”
Even the indomitable Imri’ al-Qays (d. 560 AD) employed it. In one poem of his, written on a qâfiyah with the rawî of (m), three of its 23 verses end on (mu) while the rest end on (mi). Two of the three instances of (mu) occur in lines 14 and 16:
Jâlat li-tasra`anî fa-qultu laha-qsirî * innî imru’un sar`î `alayki harâmuSlant rhyme can be achieved by replacing the rawî. This is considered more acceptable when the substitute letters are articulated in a similar manner to one another. It is less acceptable when the letters are distant from each other in sound. Some letters that the Arabs consider close to each other in their articulation are (l) (m) (r) and (b). The following example by al-`Ujayr al-Salûlî (d. 709? AD) exhibits all four of these letters as the rawî in four successive lines:
A lâ qad arâ in lam takun ummu mâlikin * bi-milki yadayya innal-baqâ’a qalîlu(Note: The long vowel that appears in these verses before the rawî is part of the qâfiyah. When a long vowel appears before the rawî in a qâfiyah is called the ridf. The long (î) and (û) are interchangeable as the ridf and their alternation does not constitute part of the rhyme’s slant. By contrast, replacing the ridf with an un-voweled consonant would be another type of slant rhyme.)
What implications might Arabic slant rhymes have for English poets who wish to emulate the Arabic forms? Indeed, the main purpose of this article is to explore such questions.
To me it seems obvious that there would be little point in adopting into English anything quite like the Arabic slant rhymes as they appear in Arabic poetry, since these would hardly sound to us like rhymes at all. Nevertheless, an equivalent practice would be to allow the poet greater flexibility with the microrhyme, by darkening and lightening the vowel colour of the rhyme for whatever poetic effect the poet wishes to produce.
For instance, if the poet is writing a poem with the rhyme on (er), it would be possible to introduce the sound (ir) or the sound (or). If the poem rhymes on (en), the poet might want to bring in the sound (an).
Also, these Arabic slant rhymes provide us with a basis for using other forms of English slant rhyme in Arabic-based poetic forms, like the (eaves/ease) alternation in my “qasîda # 1”.
These variations could be used to add variety or soften the monotony of the monorhyme in longer poems. They could also be employed as a foregrounding technique or to change the tone of certain verses for poetic effect.
* All of the English poems that are referred to in this article can be found in “English Ghazals Based on Arabic Forms” published in the July 2007 Special Issue of The Ghazal Page.
al-Akhfash, Abû al-Hasan Sa`îd b. Mas`adah. al-Qawâfî. Dr. `Izzat Hasan (ed.). Damascus. (1972).
© 2007 David Jalajel
All of the English poems referred to in this article can be found in "English Ghazals Based on Arabic Forms," published in the July 2007 Special Issue of The Ghazal Page
Back to the Prose index
Back to the 2007 special issue 1