The Quran in relation to Pre-Islamic poetry

First of all: Fantastic news! Last week Wednesday I got word that I got the VENI research grant to research Arabic of the Early Islamic period. I'm incredibly happy that I was lucky enough to get this grants, and cannot wait to start my research.

Tangentially related to my research is of course my work on the Quranic Consonantal text, and today I'd like to talk about a topic of interest, for which we have very little information, and therefore it will not turn into an academic article anytime soon, this is the poetic style of the Quran, and some of the Poetic-texts that we have so far discovered in the epigraphic record of the Pre-Islamic period.

The Quran is written in a literary style that might be called 'Rhymed prose'. All verses of the Quran rhyme with following or preceding verses, and many of the Surah's of the Quran are completely in Monorhyme. Rhyme in the Quran is a bit more free than we would normally consider rhyme in a Western poetic context, although surprisingly close to the more free rhyme that we often find in the lyrics of popular music. These rhymes are consistently "pausal", that is, they are pronounced in the way that words are pronounced at the end of a sentence in Classical Arabic. This means that final un, in, -u, -i and -a are lost, and that the sequence -an of the indefinite accusative is pronounced as -ā. Moreover, if the feminine marker -at ends up in word-final position due to the loss of these elements, it is pronounced -ah.

This Pausal rhyme is uncommon outside of the Quran. In Islamic Poetry, as well as the so called Muʿallaqāt, the Pre-Islamic Poetry transmitted within the Islamic tradition, considered the utmost example of Classical Arabic, consistently use these final short vowels (which make up the case and mood markers in Classical Arabic) to build the rhyme. This stark contrast has caused me to wonder whether the language of the Quranic Consonantal Text had these final short vowels in the first place, as it is difficult to imagine such a difference between the proper Classical Arabic poetry, and the text of the Quran having such different ways of treating pausal pronunciations (any reader that knows examples of optional pausal pronunciations in other languages; or examples of a large variety of choice of pausal pronunciations is kindly asked to please let me know).

Besides the difference of pausal versus context pronunciations being used in the Quran and Arabic poetry, there is yet another difference between the two, namely, Classical Arabic poetry adheres strictly to a quantitative meter, while such a system cannot be detected in the Quran. This does not mean the Quran is completely without meter, a fantastic paper by Devin J. Stewart"Sajʿ in the Qurʾān", shows an in-depth overview of the "Saǧʕ" style of rhymed prose of which there are many examples within the Quran. While it appears to be somewhat controversial to consider the Quran itself an example of Saǧʕ (for reasons you may read yourself) there can be no doubt that there are clear examples of this style all throughout the Quran. The Saǧʕ style is exemplified, by two ore more lines with a similar rhyme, and a roughly equal amount of stressed phrases. Several words, especially proclitic particles are stressless, and are therefore part of a larger "stress word". The first line of a Saǧʕ unit in the Quran may be preceded by a short phrase that falls out of the larger stress pattern.

Q94 is a nice example of a Surah which consists of three separate "Saǧʕ units", where in between these units the rhyme also changes (which is not always the case):

The "meter" then is not at all syllable based, but quite strongly stress-based (although even in this small example you can also see some clear syllable-number parallels). Notice also the subtle A-B-A-B near-rhyme scheme of the first Saǧʕ íCrak/áCrak which certainly is intentional. The next Saǧʕ Unit has once again a metrically similar rhyme úCrā, as well as the last rhyme: áRCab. While this style of composition has been discussed to a great extent already by Classical Arab scholars, a full analysis of the style and development of Saǧʕ in the Quran does not seem to have been undertaken but should be highly interesting, the extent of it has, I believe been underappreciated by Western scholars.

An interesting part of the rhyming scheme in the Quran (and one not clearly remarked upon by Stewart) is that the post-stress pronominal clitics -hā, -hum, -kum and the indefinite accusative marker -ā are never enough to constitute a rhyme, the syllable (or syllables) preceding it must also rhyme. This must be understood from the strong stress-based meter. These clitic syllables are always unstressed, simply by the way the stress system of Arabic works. As a result, they cannot constitute the core of a stress-based rhyme. This is not unusual in languages that have stress based rhyme. in Dutch, for example one could not choose to rhyme two consecutive words that end in the unstressed [ə] without also rhyming the preceding syllable, e.g. trappen [ˈtʁɑ.pə] and lopen [ˈlou̯.pə] can't rhyme despite their final syllable being identical because it is unstressed. Nor can, otherwise identical syllables rhyme if one is stressed and the other one isn't: meegaan [ˈmei̯.χan] does not rhyme with gegaan [χə.ˈχan].

What is, however interesting about these unstressed elements, is that they can also occasionally rhyme somewhat with each other, or be invisible to the rhyme of the word.

For example in Q47:5-10 we find a rather complex rhyme scheme āRa(hā/hum/kum):

Saǧʕ unit 1:
sayahdī́-him / wa-yúṣliḥ(u) / bā́la-hum //
wa-yudḫílu-hum(u) / ulǧánnat(a) / ʕárrafahā la-hum (or rather ʕarrafa-hā́=la-hum as one stress unit?) //

Saǧʕ unit 2:
(yā-ʔayyu-hā llaḏīna) ʔā́manū / ʔin tánṣurū / allāh(a) / yanṣúr-kum / wa-yuṯábbit / ʔaqdáma-kum //
wa-llaḏī́n(a) / káfarū / fa-táʕsā(-an) / láhum / wa-ʔaḍáll(a) / ʔaʕmā́la-hum //
ḏā́lik(a) / bi-ʔánna-hum / kárihū / mā ʔanzala allā́h(u) / fa-ʔáḥbaṭ(a) / ʔaʕmā́la-hum //
ʔa-fa-lam yasī́rū / fi lʔárḍ (i)/ fa-yánẓurū / káyf(a) / kā́n(a) / ʕā́qibah(-tu) / allaḏī́n(a) / min qáblihim / dammara allā́h(u) / ʕálayhim / wa-li-lkāfirī́n(a) / ʔamṯā́lu-hā //

This second Saǧʕ-pattern is an example of of two approximately equal length lines followed by a third line of about twice the length commonly called rubaʕī"Quatrain" If we take the verb-subject sequences mā ʔanzala llā́h(u) and dammara llā́h(u) as a single stress units, it can easily be analysed as a perfect 6-6-12 distribution. As you can see, even in between the same Saǧʕ unit 3 different clitics are used.

Occasionally we even find such unstressed patterns being completely "invisible" for the Rhyme, as found, e.g. in Q90:1-7

Saǧʕ Unit 1:
lā ʔúqsim(u) / bi-hāḏā lbálad(i) //
wa-ʔanta ḥíll(un) / bi-hāḏā lbálad(i) //
wa-wā́lid(in) / wa-mā wálad(a) //
la-qad ḫaláqnā / alʔínsān(a) / fī kábad(in) //

Saǧʕ Unit 2:
ʔa-yáḥsab(u) / ʔan lan yáqdir(a) / ʕaláyh(i)  / ʔáḥad(un) //
yaqū́l(u) / ʔahlákt(u) / mā́lā(-an) / lúbadā(-an) //
ʔa-yáḥsab(u) / ʔan lam yára-h(ū) / ʔáḥad(un) //

In the definition of the Classical scholars, it is allowed to have the last line of a Saǧʕ Unit be a bit longer than previous lines but not vice-versa. This seems to be going on in Q90:1-4. The Secong Saǧʕ Unit, clearly forms a triad with an áCad/-úCadā/-áCad scheme, but notice that in this analysis, it is one stress unit shorter than the other two.

We also find examples, where a series of rhymes ending in -ā (or -ē, in my analysis) are followed by a series of rhymes in -ā-hā, e.g. in Q79:15-46. The Saǧʕ Unit analysis of that Surah, however feels somewhat difficult.

A more thorough analysis of the enormous amounts of Saǧʕ Units found in the Quran might shed more light on what may or may not be considered a stress unit, but that stress plays an important role in all of this is fairly obvious. Moreover, a more thorough analysis may actually reveal what parts real Saǧʕ and what other parts just have actual rhymed prose. Someone linguistically informed and a passion for poetic analysis may write a really wonderful thesis or dissertation one day, and may reveal essential parts of the Quranic poetic style, and the language's stress system.

The reason why I started this rather long introduction to Quranic rhyme, is because it is relevant. While the traditional Pre-Islamic Poetry (of which we really can't be sure if it is Pre-Islamic or represent a dialect even remotely similar to that of the Quran) is very different from the poetry in the Quran. However actual Pre-Islamic Arabic poetry as found in the En Avdat inscription, and the recently deciphered Safaito-Hismaic Baal Cycle poem present striking similarities with the Saǧʕ of the Quran. Using Ahmad Al-Jallad's transcriptions of Both poems we get the following:

la-hágg(a) / máwt / w-al-lā́ṯ̣ṯ̣ / ṯarā́m //
pa-moyakā́n / ḫálp / layā́lay-oh / wa ʔaywā́m-oh //
wa-hā ʔab-báʕal / yobátt / wa-la-h(ū) bā́tt / wa mā nā́m //

With a 4-4-4 stress unit and a ā́m/ā́m-oh/ā́m triad rhyme it shares many similarities to the Saǧʕ in the Quran.

The En Avdat inscription is, if it is analysed correctly, is a little more difficult to analyse as a form of Saǧʕ, but still shows remarkable similarities:

pa-yápʕal / lā pédā / wa-lā ʔóṯrā //
pa-kon hónā / yabġe-nā ʔal-máwto / lā ʔabġā́-h //
pa-kon hónā / ʔadā́d(a) górḥū (Al-Jallad: górḥo) / lā yórde-nā //

If we take the verb-subject pairs as single stress units, the stress is 3-3-3; The rhyme appears to be a óCCā/aCCā́-h/óCCeCā triad, with  a mismatch in stress and very free rhyme.

A few things should be noted about this inscription: This dialect has clearly not lost its case vowels. The indefinite accusative *-an, has shifted to *-ā and is used in the rhyme (and is written with <ʔ>), the definite nominative *-u and indefinite nominative *-un are marked with <w>, presumably representing -u and -ū respectively. Short a however is not transcribed (lost?). There can be no doubt that this dialect differs from Safaitic which seems to have only retained both the indefinite and definite accusative -ā and -a; The dialect of the Poem above may have even lost final -a, as the final rhyme of this inscription would not work if it had any case vowels left.

The Quran's poetic style seems to stand within a tradition of an already existing Pre-Islamic poetic style, present in the poetic style of two vastly different Pre-Islamic Arabic dialects. With just these two examples it is difficult to decide what to make of all of this, but one hopes that future discoveries will shed more light on this form of Pre-Islamic poetry, and modern research will give a clearer insight into this poetic style in the Quran.