Shifting sands of Arabic prose

English-speaking scholars of Arabic have been making translations of Arabic poetry and prose available to general readers since the 18th century, when Simon Ockley, George Sale and Sir William Jones all produced distinguished work; but when poetry and prose have been translated together, J. D. Carlyle (1796) and Sir Charles Lyall (1877 onwards) for example, the prose passages have basically been background material. Robert Irwin, whose feel for medieval Arabic has been well shown in his The Arabian Nights: A Companion (1994), has rightly avoided this trap by his decision to include the Koran and adab ( belles-lettres ) and some allied material and so give his anthology a better balance. He might have gone further, for many areas remain uncovered, even though suitable translations are available. Excerpts from the major historians, geographers and travel writers, for example, would have been valuable alternatives to some of the passages that he has selected.

It is hardly surprising that Irwin is at his most expansive and convincing with the genres he knows well. This results in the chapter on later medieval literature, "Servitude and military grandeur", being the longest and best in the book. Here Irwin is in control of his enthusiasms and the reader benefits greatly. Yet, apart from stabilisation of The Arabian Nigh ts, an adventitious masterpiece if ever there was one, this is not a period that produced great literary works.

He himself provides some of the best translations in the book. Elsewhere he sometimes hobbles himself by using translations by those whose grasp of Arabic is not particularly good (for example, A. R. Nykl) or is good but has no literary feeling (for example, Bernard Lewis). He also includes translations of writers of questionable literary worth, such as al-Hallaj, al-Mu'tadid and the pseudo-al-Majriti.

The non-specialist may well find the book readable, but those with some knowledge of the subject will be irritated by mistakes and misjudgements. Many mistakes are minor but confusing: both Umayyad and Ummayad, Abbasid and Abassid, Ibn Khafaja and Ibn Khafaia and so on; and poor Salma Jayyusi is referred to throughout as Jayussi. There are also "spellcheck" errors, such as "different to transcribe" for "difficult to transcribe". Careless phrasing does not help. Thus we are told that the kharja of the muwashshah is usually "in colloquial Arabic or some other vernacular tongue". It is not easy to deduce from that phrase that some kharjas are Romance or bilingual.

Occasionally, the mistakes are horrid. Even though he gives us a translation of Koran 12: 1-45, he fails to recognise the famous Koranic saying that comes only 20 verses later: "Here are our goods returned to us" [12:65], giving us instead, "This is our merchandise. Give it back to us."

The misjudgements follow a similar pattern. Some are petty. To say that Arabic is related to Amharic has about as much point as saying that Latin is related to English. Others lead to distortion. Thus Irwin presses the views of Abbasid literary critics to the effect that poetry was all important in pre-Islamic times. The real situation was that prose was also important and provided the key linguistic registers that were to come together in the Koran. However, much was lost with the coming of Islam and the rest became background material. The critics wrote a quarter of a millennium after the event and made their judgements from what was available in their own times.

Irwin is right to centre his piece on pre-Islamic poetry on the "Mu'allaqa" of Imru' ul-Qays and the "Lamiyya" of al-Shanfara, but his decision to give us comparative versions of parts of the "Mu'allaqa" uses space that could have been devoted to more comment about the thematic content of pre-Islamic poetry and in particular the magnificence of its descriptive passages ( wasf ). We might then have had 'Abid's description of an eagle killing a fox, or some other equally fine celebration of the natural world. As it is, there is no room for even a short fragment from such great poets as Labid and Zuhayr. When we come to the elegy and to women's poetry all we get is al-Khansa', a madwoman of mediocre poetic talent. In general, work by women or about women is under-represented throughout the book.

Only 12 pages are devoted to the Koran, with just five translations. This is hardly enough for the most magnetic and influential work in Arabic. Irwin's use of part of the story of Joseph and of the allegorical "verse of light" is well judged, but where are the pieces of rhetoric or the descriptions of heaven and hell?

The chapter devoted to literature between the death of Muhammad in 632 and the fall of the Umayyads in 750 is also too short and it is not well organised. Thus the translation from Abu Dhu'ayb (died 649) and the key comment that at first Islam made little difference to Arabic poetry is on the last page. Both should have been at the beginning. The chapter also includes passages on philology and grammar that should be elsewhere.

More attention should have been given to 'Umar ibn Abi Rabi'a, a pivotal figure who is inadequately treated, and to Dhu l-Rumma, the last of the great desert poets. 'Umar's poems are said to be "largely dedicated to unrequited love". Yet in the Arab world 'Umar is notorious as a libertine, a reputation based on his poetry.

In the one translation from 'Umar that Irwin prints (an excellent one by Adel Gamal), there is a sketch of a successful assignation. Its other features are light language, a sense of drama and depiction of women's dialogue, all admired characteristics of 'Umar's poetry.

Irwin struggles manfully with Abbasid literature (chapters four and five),and has more successes than failures. Among the latter is the treatment of al-Mutanabbi, most renowned of the poets and provider of the title of the book, who is afforded only one short poem. With the maqama genre of rhymed prose, the problems lie with the choice of maqamas and of translation.

The lengthy Madira maqama of Hamadhani is not as sharp and funny as such shorter pieces as the Baghdad, Isfahan and Mosul maqamas . With Hariri, no attention is paid to the work of Thomas Chenery, by far the most successful translator of this difficult material.

The availability of English translations appears to have been a major factor in determining the contents of the chapter on Andalusian literature.The result is unbalanced, with major poets such as al-A'ma al-Tutili, Ibn Baqi, Ibn Sahl and al-Shustari not even mentioned. Even more peculiar is the omission of Ibn Bassam, the preserver of the Zawabi' of Ibn Shuhayd, from which we are given two passages, and inspirer of the north African al-Maqqari, from whom there are also two extracts. The piece on stanzaic poetry is particularly thin, with only one zajal and one muwashshaha offered.

Despite such obvious shortcomings, the usefulness of the book outweighs the problems. However, Night and Horses and the Desert is a disappointment after Irwin's splendid The Arabian Nights: A Companion , a work full of intuitive insight. That is the book that I shall continue to expect my students to buy.

Alan Jones is professor of Arabic, University of Oxford.