The British-Yemeni Society

News and events


Journal articles

Book reviews

About the Society

Society officers

Annual reports


Annual appeal



Traditional Arab sailing ships


This is an abridged version of the illustrated talk given to the Society by the author on 14 May. James Taylor served as an engineer in the Indian Army before independence, and later in East Africa and the Arabian Gulf His interest in dhows was aroused during a voyage by troopship to India in 1944, and developed when he was stationed in Mombasa and Bahrain. In retirement he has studied Arabic language and classical literature. He first visited Yemen in 1988, returning there on the Society’s tour in October 2002.

About twenty years after the migration of the Prophet Muhammad and his followers from Mecca to al-Madinah that marks the beginning of the Muslim calendar, the second Caliph of Islam, ‘Umar bin al-Khattab, famously refused to sanction the invasion of Cyprus by his governor of Syria, Mu’awiya bin Abi Sufyan, on the strength of the letter that he received from ‘Amru bin al-’As, the Arab conqueror of Egypt. ‘Amru wrote, ‘The sea is a boundless expanse whereon great ships look like tiny specks; naught but the heavens above and the waters beneath; when calm, the sailor’s heart is broken; when tempestuous, his senses reel. Trust it little. Fear it much. Man at sea is but a worm on a bit of wood (dud ‘ala ‘ud), now engulfed, now scared to death. ’

One consequence of this incident is that some western scholars have dismissed the peoples of central Arabia as reluctant mariners, ignoring the significant role played by descendants of immigrants from the heartlands of the Arabian peninsula in the fleets of Kuwait, Bahrain, ‘Uman and al- Yemen that, along with ships from Basrah, carried the trade and faith of Islam to Africa, India and the Far East. I am not going to say much about those whom ‘Amru dubbed ‘worms’, but I am going to tell you a little about some aspects of the ‘bits of wood’ on which they ventured forth, namely the traditional Arab sailing ships that we in the West call dhows. This, however, is not what the Arabs call them; for, although the word is spelled as if it might be a transliteration from the Arabic, you will not find anything like it in the modern Arabic lexicon. So where does it come from?

Alas! The absence of any definitive evidence has left the field wide open to those who dabble in the black arts of speculation and intuition. Hence, at one time or another, Persian - Basque - Marathi - Swahili - and Chinese have all been suggested as its source. From the sound patterns of the languages spoken in the countries surrounding the Indian Ocean I suspect it to derive from the Hindi/Portuguese patois that evolved in India during the Portuguese hegemony. Indeed, there is a very similar word, padao, in use today for a type of Indian sailing ship. Among themselves, Arabs generally refer to dhows merely by the Arabic expression for ‘sailing ships’ (marakib/sufun shira’iyah), except when they are knowledgeable enough to use the technical terms for the different types of dhow. These are said to exceed 200 in number.

To talk of differences of dhow type is to talk of difference in hull shape, for this is the main criterion by which one type of dhow is distinguished from another. Hence we see (Fig 1) that each of such names as boum - sanbuq - zaruq - and baghlah - is associated with a characteristic form of stem and stern, whilst there is very little variation in the sail plan, which always consists of a single, large, triangular sail, which we call a lateen, hoisted on each mast. The number of masts is not significant and so one may encounter a sanbuq, for example, with one, two or even three masts. This is in marked contrast with the system of nomenclature adopted in Europe and America, where the main criterion which distinguishes one type of sailing ship from another is the number and arrangement of the masts and sails, without specific reference to hull form.

In addition to differences in hull shape, some types of dhow are further distinguished by characteristic painted or carved embellishment of the stem or sternpost or, in the case of the batil (Fig 2), both. Apart from a replica built in Kuwait, batils are now extinct but were much in favour as warships and slave ships during the 19th century because of their speed and manoeuvrability. Captain Colomb, who commanded a Royal Naval ship engaged on the suppression of slave trading in the Indian Ocean during the 19th century gives a succinct description: ‘If a pear be sharpened at the thin end and then cut in half longitudinally, two models will have been made resembling, in all essential respects, the ordinary slave dhow’ The inboard facing emblems at bow and stern echo the ships of Ancient Egypt, Rome, and Byzantium in their arrangement whilst the stern of the batil recalls the ferocity of a Viking figurehead, although it is said to be of Phoenician origin. Here again,

the speculative school of marine historians has had a field day theorising about the significance of these decorations. It has been suggested that the menacing figurehead at the stern of the batil was to scare off the malignant spirits that were believed to haunt certain headlands and dangerous places such as Ras Fartak and Jabal Kadmal in Southern Arabia. This is denied by the distinguished Yemeni nautical historian, Dr Hassan Sabab Shihab, who says that Arab seamen sought to conciliate the local demons by offerings of food and drink set afloat in buoyant cooking vessels rather than to confront them. According to him, the function of the ferocious looking figurehead was to frighten enemies, a theory that accords well with the role of the batil as a warship.

The famous Omani navigator Ahmad ibn Majid an-Najdi wrote in his 15th century treatise on navigation and seamanship that the Arabs learned the art of shipbuilding from the prophet Noah who was in turn instructed by God, through the mouth of the angel Gabriel. Ibn Majid went on to say that the outline of Noah’s Ark (safluat an-Nuh) was delineated by 5 stars of the constellation Ursa Major (The Plough), a profile that closely approximates to a Kuwaiti boum, a batil, a baghlah/ghanja and a mtepe from the island of Lamu off the Kenya coast; this suggests that the lines of the archetypal Arab sailing ship were firmly established by the time that the legend was set down in writing in the last years of the 15th century CE, and probably much earlier. We can safely assume that the first primitive ship-wrights drew inspiration from their observation and experience of the world around them and the raw material which its resources offered. Thus the sight of a fallen tree borne on the flux of some torrential stream, perhaps with some hapless creature squatting terrified on its trunk or clinging to its branches, may well have been the germ of the log raft or the dug-out canoe or huri (Fig 3) which I photographed in the nineteen fifties when it was serving as a tender with the remnants of the Bahrain pearling fleet. Additional planks have been put on to raise the freeboard and so increase the load carrying capacity of the canoe. This practice is almost universal in the construction of the huri. Perhaps the sight of a floating reed, papyrus stem, or palm leaf inspired the reed boat, and the sight of some floating carcass bloated by intestinal gases the Assyrian kelek. Strabo (xvi. 4. 19) tells us that the Arabs of South Arabia used float-supported rafts to sail across the Red Sea to the coast of Africa in order to trade with the inhabitants and, in the accounts of Mahmud of Ghazni’s invasion of India, we are told how his advance guards fought their way across the river Ramaganga supported on inflated skins, plying their bows as they swam. Even in comparatively recent times military engineers have not disdained to use such devices.

The abundance of reeds in the marshes of southern Iraq, close to one of the epicentres of old world civilisation, the persistence of boats made of palm leaf stalks in Oman and the Gulf today and the well publicised activities of Wilfred Thesiger and the late Thor Heyerdahl led some scholars to believe that Arab navigation began on top of a bundle of reeds. Although I can accept this theory in so far as it applies to canoes and small fishing boats, I find it hard to swallow the idea of porous and flimsy reed boats carrying bulk cargoes from Sumer to the Indus valley, or hauling copper from Oman via Bahrain. However, my main objection to the idea is that, by 2500 B. C. , when there is sound evidence for the existence of a regular trade between Sumer, Dilmun (Bahrain) and the Indus valley, the art of shipbuilding had advanced beyond the reed bundle boat. Indeed, a silver model of a canoe from a Sumerian grave of the same period appears essentially the same as the wooden tarrada of the type used by Wilfred Thesiger in the Iraqi marshes about fifty years ago (Fig 4). Of course, it is just possible that the Sumerian tarrada was built of reeds and plastered with bitumen, but its sharp angles and plane surfaces lead me to think otherwise. Thus, in my view, chronology alone rules out the probability that the reed bundle ship played a significant part in the Sumer - Dilmun - Indus valley sea-borne trade.

At this point in time the trail of the Arab ship runs cold and remains so for about 3000 years.

Although there are numerous descriptions of the rich sea-borne trade of the Arabs with India, East Africa, China and the East Indies during this period, there is no mention of the ships involved until the middle of the 6th century CE, when one of the seven great pre-Islamic poets, Tarafa bin al-’Abd, includes a couple of verses in his mu‘allaqa in which he likens the movement of the camel-borne litters in which the Bedouin women used to ride, winding their way around the stones of a dry watercourse, to the zigzag passage of khaliya safin in the sea. Later Arab scholars tell us that khaliya safin were ‘great ships’, or that they were ‘ships that travel without seamen to make them move’, which, in the language of the time, probably meant that they were sailing ships.

According to al-Jahiz, in the last decade of the 7th century CE, al-Hajjaj ibn Yusuf ath-Thaqafi, the iron handed Marwanid viceroy of Iraq, tried to introduce flat-bottomed, nailed ships like those of the Mediterranean to the waters of the Arabian Gulf. The experiment failed because experience had taught Arab seamen that the ships they were used to, in which the planks were fastened together with coir ropes and daubed with grease, were better equipped to withstand the frequent groundings and collisions with the sandbanks and submerged reefs that abound in the inshore waters of the Red Sea and the Gulf.

This last piece of information has largely been overlooked by western nautical historians. This is a pity because it sheds a different light on the sudden change from stitching to nailing in the construction of Arab ships that coincided with the appearance of the Portuguese in the Indian Ocean at the end of the 15th century C. E. In the past, it has generally been assumed that the change was merely one of the improvements in shipbuilding techniques introduced by the Portuguese. However, the knowledge that nailing had been tried and rejected by the Arabs 800 years previously opens the field to an alternative and, I think, more credible explanation. Prior to the advent of the Portuguese, the tactics of sea fighting in the Indian Ocean consisted of boarding and hand-to-hand fighting, mainly in skirmishes with the pirates that infested some waters. Indeed, Pliny reports in his Natural History (vi. 173) that the piratical activities of some Arab tribes living on the coast of the Red Sea forced the Romans to carry guards on their merchant ships and the Arab geographer al-Muqadassi warned, in the last decade of the 10th century CE, of the need to carry armed men and throwers of Greek Fire when navigating the waters of southern Arabia. The sudden arrival of the Portuguese with their ship-mounted cannon changed all that. The Arabs had to adapt, or, quite literally, go under. Nailed ships had the strength to bear the weight of the cannon that the Arabs now felt obliged to carry. Moreover, they were better able to withstand the impact of shot and shell.

Al-Muqadassi was one of the first of a long line of travellers to mention the construction of an Arab ship, which was probably of a type called a jalbah or jalabah, and the terrors of travelling in it. In those days, the frontier between the Crusader Kingdom of Jerusalem and Fatimid Egypt stretched from Aylah at the head of the Gulf of Aqaba to Asqalon on the Mediterranean coast, blocking the overland route for Muslim pilgrims from Andalusia, North Africa and Egypt; so pilgrims used to make their way overland to Aidhab, a port on the Red Sea coast of Egypt, where they crossed over in a jalabah to Jeddah, en route to the two Holy Cities of Mecca and Medinah. Their descriptions of the ship tally exactly with the Lamu mtepe (Fig 5) in almost every essential respect. Stitched fastenings, a single mast, a square sail made of palm leaf matting, no deck; everything agrees, except that the mtepe is steered with a tiller; whereas al-Muqadassi mentions hand lines.

Like us, the Arabs are pretty casual in their use of ship type names and so one finds the same name being applied to vastly different types of ship at different periods in history, in different parts of the Arab world and, frequently, by different individuals. For example, we have seen the name tarrada, derived from the Arabic verb meaning to chase - hunt - drive away, still used in Iraq for the war canoe of Sumerian provenance. It is also used in Arab navies for a modern battle cruiser and was, in the past, used for both a light sailing war vessel and, according to Ibn Battuta, for a ship ‘in the shape of a barrel for conveying horses and cavalrymen’. It has also been confused in print with the Portuguese word terrada, meaning of the land or country, which was used as a general term for European/Asiatic hybrids created by Indian and Arab shipwrights. Moreover, the Arabs were given to categorising a ship by its port of origin, hence Arab scholars believe the ‘aduliyyah of Tarafa’s mu’allaqa was a ship from the port of ’Aduli on the eastern seaboard of Arabia. Also like us, the Arabs frequently recycle the name of ships that have fallen into disuse.

All this inclines me to suspect that, when the jalabah of the medieval travellers finally disappeared from the scene, the name re-emerged as the jalbut (Fig 6) which shares with the shu’i the distinction of being the most popular types of dhow plying their trade in Oman and the Gulf today. Others believe the name derived from the English ‘Jolly boat’ and a great deal of ink has been spilled in arguing the subject. The present popularity of the jalbut and the shu’i comes from their long straight keels, their transom sterns and their generous quarters, which make them ideally suitable for conversion to mechanical propulsion. Back in the fifties, when I was living in Bahrain, mechanisation was firmly established, although sailing ships were still very much in evidence in the remnant of the pearling fleet. Travelling around the coasts and harbours of the Arab world, one cannot fail to notice that each area seems to have its favourite type of ship or ships. Thus we see that the jalbut and shu’i predominate in Bahrain, the shu’i and sanbuq in Oman, the za’ima, zaruq and, less frequently, the sanbuq in the Yemen.

The modern sanbuq (Fig 7) is recognised by its transom stern and spoon shaped bow although the exact shape of the bow can vary quite considerably. The similarity in their appearances has led many scholars to suspect that the sanbuq is derived from the Portuguese caravel. This is by no means impossible because (1) The transom stern is a late innovation in Arab shipbuilding accredited to the Portuguese (2) The whole profile of the modern sanbuq differs from the archetypal Arab ship profile exemplified in Ibn Majid’s description of safinat un-Nuh (3) There was a stitched ship known as a sanbuq already in existence which is believed to have been more like the zaruq than the modern sanbuq.

The zaruq, which takes the form of a large canoe closely resembles a number of archetypes such as the Viking longship.

Finally we come to the first known picture of an Arab sailing vessel, the so-called Hariri Ship (Fig 8). It appeared in 1237 CE. , about 250 years after al-Muqadassi described the manner of steering used by the ship upon which he circumnavigated the Arabian Peninsula and about 50 years after Ibn Jubair’s description by the jalabah, so it is quite possible that the picture was based on the descriptions of these and other travellers. It was drawn by Yahya bin Mahmud al-Wasiti for an illustrated copy of a book entitled Al-ma qamat written by Abu al- Kasim bin al-Hariri, a part time grammarian and man of letters whose day job was sahib al khabar; or head of the intelligence department of the court of Basrah. The book consists of 50 stories in rhymed prose teaching various recondite aspects of Arabic style, philology and grammar through the adventures of a master conman named Abu Zaid who, in tale after tale, tricks his way out of trouble, or into a sum of money, through his mastery of the finer points of Arabic.

The picture comes from the 39th or Omani rnaqamah set on a ship bound for Surat and shows a number of interesting features: stitched planking, a grapnel anchor, the continuous baling mentioned by various travellers in stitched ships, a hull shape resembling the archetypal Arab ship profile of safinat un-Nuh and a central, stern mounted rudder, which may have been the first of its kind.

Al-Hariri’s work does not translate well so his writings are known in the West only among Arabists but, ironically, his name is perpetuated amongst Western nautical scholars because of this picture, drawn by another hand a century or more after his death.

August 2003