The Structure and Nature of the Traditional Yunnan Caravan Trade

Part of Trade Routes

By the mid-19th century the caravans of Yunnanese traders ranged over an area extending from the eastern frontiers of Tibet, through Assam, Burma, Thailand, Laos and Tonkin, to the southern Chinese provinces of Sichuan, Guizhou and Guangxi. Two main routes existed between Simao [Pu’er] in southern Yunnan and northern Thailand, one passing through Burmese territory, and one across Laos. The more westerly, Burmese, route passed through Kengtung [Kyaingtong], and entered the present day frontiers of Thailand in Chiang Rai Province, to the west of Mae Sai. The second, more easterly route entered Laos' Phong Saly Province and led due south across Luang Prabang Province before swinging westwards to cross the Mekong into Thailand in the region of Chiang Khong.

Perhaps the earliest detailed European account of the Yunnan-north Thailand caravan routes is that of Lieutenant T. E. MacLeod, Assistant to the Commissioner of the (British-administered) Tenasserim Provinces of southern Burma, who travelled from Moulmein to the frontiers of Yunnan (by elephant and horseback) in 1836-1837. MacLeod – the representative of an imperial power which had but recently (1826) added Tenasserim to its already extensive dominions in Asia – found the Burmese, Chinese and Chiang Mai authorities all naturally disinclined to facilitate his travels or to provide intelligence with regard to the caravan trade of the region. The Chinese caravaneers, however, were of a different mind and, whilst providing MacLeod with most of his information, expressed their strong desire to open trade links with the British at Moulmein whilst urging him to do all in his power to ensure free movement of traffic on the westernmost caravan route, via Kengtung.

MacLeod arrived in Chiang Mai (which he styles "Zumue") on January 12, 1837. Here the prince and officials attempted to dissuade him from travelling any further towards Yunnan, and specified that – should MacLeod insist on proceeding – he should follow the more easterly caravan route, via Nan and the eastern banks of the Mekong, ‘by the road the Chinese caravans came, which was also open to our merchants’. MacLeod complains that he ‘experienced the utmost difficulty in obtaining satisfactory information about the routes to China’, Fortunately, however, he was able to gather information from Chinese residents of Chiang Mai, thus:

The Chinese merchants residing in the place had told me that the Kiang Tung road [the westernmost route] was the best, that the other [eastern route, via Nan] I should find very difficult, having ranges of high mountains to cross, and that elephants could not travel by it. I should find only scattered hill tribes and no villages for a great distance.

This was further confirmed by the Chinese Haw caravaneers themselves, as MacLeod notes:

On the 27th I was happy to see part of the Chinese caravan arrive, their report confirmed what I had before heard about the road. The chiefs [of Chiang Mai] had assured me that there was a road more to the eastward than the above mentioned one [via Kengtung], along the eastern bank of the Mekhong or Cambodia River, with large towns and villages two or three days apart. These the Chinese informed me did not exist, that they had many years ago been pillaged and destroyed by the Siamese Shans, and the road entirely overgrown with Jangal (sic) and blocked up. They also urged me to try and get the Kiang Tung road, which was by far the best, thrown open.

In fact, both the Chiang Mai authorities and the Haw caravaneers seem to have been emphasising the merits and disadvantages of the eastern Laotian road for their own reasons, the former because ‘the King of Siam had forbidden all communications [with] the Kiang Tung people [who] though not Burmans were subjects of Burma and therefore could not for a moment be trusted’, whilst the latter were anxious to reopen the Yunnan-Kengtung-Chiang Mai route with a view to promoting their trade with Moulmein.

In the end, and after much difficulty, MacLeod was at last able to proceed to Chiang Rung [Jinghong] in Sipsongpanna [Xishuangbanna], only to be effectively denied permission to proceed any further by the wary Chinese authorities, or to remain longer in Kengtung territories by the equally suspicious Burmese. Accordingly, he retraced his steps to Chiang Mai by the same route he had followed on his outward journey. As a result of his expedition MacLeod was, however, able to glean valuable information regarding the trade routes between northern Thailand and Yunnan, which may be summarised as follows:

There were two main Haw caravan routes between northern Thailand and Pu’er in southern Yunnan, other routes ‘being only branches of them and occasionally slightly deviating from them’. The westernmost, by which MacLeod proceeded, was:

... for three days over low hills, then for eleven marches to the frontier village belonging to Kiang Tung [called] Hai Tai, through valleys and occasionally over a few low hills, then over high mountains to Kiang Tung. From Kiang Tung to Kiang Hung [Jinghong] the country is both hilly and mountainous with small rich valleys through which we daily passed, and in which there are numerous villages all well-peopled. These mountains though not passable for carts have good roads and are in every respect easier to pass over than those between this and Zumue, but there is not a spot of ground among them in which an encampment could be formed for a large force. Water is throughout abundant and the country thickly wooded. From Kiang Hung to Muang La [Pu’er region] is five marches, and the road runs over high and barren hills.

MacLeod further noted that Kengtung was a ‘great thoroughfare for the Chinese caravans, being the only safe high road from China to... the Shan states to the westward of the Salween’. As for the more easterly road, which the Chiang Mai authorities initially commended to MacLeod but which he had declined to follow:

The other one is the road by which the Chinese caravans come to Zumue; it separates from the other one [at] the village of Pak Bong [just to the south of Chiang Rai], from whence to the Cambodia [Mekong] River, on which the town of Kiang Khong [Chiang Khong] stands and belongs to Muang Nan, it is six or seven marches. The river is there crossed, the road continues in the Muang Nan district for four or five days, and then enters the Muang Luang Phaban [Luang Prabang] territory and continues through it for two or three days, after which it passes through Kiang Hung [Jinghong] territories to Muang La [Pu’er]. The Chinese describe this road as very mountainous. It occupies them forty days to reach Muang La from Zumue. The road travelled by the Chinese, to Muang Nan separates from the Zumue one at Kiang Khong on the western bank of the Mekhong or Cambodia River.

MacLeod further noted with regard to this eastern road that the Chin Haw caravaneers visited Luang Prabang on an annual basis, whilst they also travelled to Nan and to Phrae to procure cotton which was grown with particular success in those localities.

Some fifty years after MacLeod's pioneering journey to the southern frontiers of Yunnan, W. J. Archer of the British Consular Service in Siam became actively involved in investigating and reporting on the trade routes of northern Thailand. By examining Archer's reports, together with the scattered casual references to be found in the writings of other contemporary travellers – e.g. Bock (1884), Hallett (1890), Smyth (1898) – it is possible to piece together a reasonably detailed picture of the Chinese Haw caravan trade in the Golden Triangle during the last decades of the nineteenth century. Thus, when Archer made his first Journey in the Vice-Consular District of Chiang Mai (1887), he noted an ‘almost total absence of trade’ within the Province of Chiang Saen, commenting that this was all the more remarkable ‘in a province lying on the great northern trade route’. Travelling south-westwards from Chiang Saen towards Chiang Rai he further noted:

Turning westward through some new settlements that show some signs of prosperity, I again passed through Ban Me Khi. This village is an important junction on the routes from Chiengsen [Chiang Saen] to Chienghai [Chiang Rai] and Muang Fang, and on the great trade route from Chiengmai to Chiengtung [Kengtung]. The road from Ban Me Khi to Chienghai is probably the greatest and most important thoroughfare in the whole of north Siam ... This, again, is the route taken by the Haw or Yunnanese traders, on their yearly trading expeditions to Moulmein.

By contrast, the easterly route via Laos to Nan seems to have declined in importance; Archer merely notes that most of the traders in Nan were Chinese, but without specifying their origin; he further comments on the almost complete lack of trade between the provinces of Nan and Chiang Rai.

Haw Traders with their pack horses near Simao, Yunnan, 1910

One year later, in 1888, Archer made his Visit to Chiengtung [Kengtung], on the direct trade route between Simao and Chiang Mai. Besides commenting on the nature of the road and on local productions, he noted that thirteen Haw caravans had passed that way during the previous year (1887-1888).

Two years subsequently, as part of the survey work of the Anglo-Siamese Commission then engaged in exploring and delineating the Burma-Siam frontier, Archer prepared his Report on a Journey in the Me-kong Valley (1890-1891). During his travels in this period, Archer came into more contact with Haw caravans than he had on previous trips; indeed, for a considerable part of his travels he employed Haw muleteers as his carriers. In this report he confirms the status of the Kengtung-Chiang Mai road as the main artery of trade between northern Thailand and Yunnan, whilst drawing attention to the fact that some of the Haw caravans were beginning to enter Thailand via the Fang Road to Chiang Mai, though he had ‘little hope that this route [would] develop into an important commercial thoroughfare’.

As for communications between Chiang Rai and Nan – the latter being an important terminus for the eastern caravan route via Laos – these remained ‘mountainous and bad’. Caravans proceeding directly to Nan via the eastern road, however, might also visit nearby Luang Prabang – where a daily market was held which was ‘certainly the largest and most crowded of any in the north, with the exception of Chiang Mai’ – before proceeding, in some instances, as far as Uttaradit on the Nan River, at the northernmost limits of navigation from Bangkok. Of the Haw caravans themselves, Archer noted:

Besides bullocks and carriers we also met a large number of Chinese caravans. The muleteers are generally Mahommedan Yunnanese, or Haws, but some are Chinese from the borders of Szechuen and Kweichau. The local people distinguish them from the Mahommedans as the ‘Haw Luang’, or men of Greater China, and pork-eaters ... These caravans come either from the northern states of Chienghung or from Yunnan, but one was composed of Chinese of a different type from Yunnanese, and came from as far as Szechuen [Sichuan].

Archer's reports can, for the most part, be confirmed and sometimes elaborated on from the writings of other contemporary and near-contemporary travellers. Thus, we know that most of the Yunnanese trading to Chiang Mai came from the Dali region of western Yunnan, and that they generally proceeded via Pu’er and Kengtung to Chiang Saen, though in the case of the great plains around the latter being flooded, they used the alternative higher road via Fang. Chiang Saen itself was a place:

... admirably situated for purposes of trade, at the intersection of routes leading from China, Burmah, Karenni [the Karen region], the Shan States, Siam, Tonquin and Annam. It forms, in fact, a centre of intercourse between all the Indo-Chinese races, and the point of dispersion for caravans along the diverging trade routes.

From Chiang Saen the Haw caravans found their way to Moulmein via Chiang Rai, Chiang Mai and Mae Sariang; alternatively, they proceeded to Uttaradit via Phayao and Phrae, where they intersected with the southern terminus of the more easterly trade route via Laos and Nan, as well as with the limit of navigation on the upper Maenam-Chaophraya River System. Further eastwards still, the Haw caravans followed the banks of the Song Hong (Red River), via Honghe and Mahao, to Lao Cai and Hanoi in northern Vietnam.

Within Burma – or, more precisely, within the Shan and Karen inhabited regions of northern and eastern Burma – other trade routes ran to Myitkyina, Lashio, and particularly to Bhamo (known to the Chinese as Xinkai, or ‘New Market’), and situated (similarly to Uttaradit on the upper Maenam-Chaophraya) at the limits of navigation on the upper Irrawaddy. Ultimately – and generally after transferral to some other form of transportation, such as river-boat or perhaps ox caravan – many of the goods imported by the Haw reached the great lowland urban centres of Bangkok, Rangoon and Hanoi. Only in the case of Moulmein, at the mouth of the Salween River, did Haw caravans themselves proceed as far as the coast, and then with the primary intention of purchasing goods rather than of selling them. But what were the goods carried by these intrepid Yunnanese muleteers? How large were the caravans, and how were they financed, organised and protected?

Frequency and Volume of the Haw Caravan Trade

Haw caravaneers drinking rice wine at a village in Yunnan, 1937

Embarkation of mules on the Mekong at Notcha Tian-pi, west of Pu'er, c. 1895

Sources, perhaps naturally, vary as to the frequency and extent of the long-distance Haw caravan trade between Yunnan and northern Southeast Asia. Thus, according to the early information of MacLeod, during the third decade of the nineteenth century Haw caravans made two trips each year to Kengtung (the road beyond to Chiang Mai being at this period closed or very difficult because of political differences between the Burmese authorities at Ava and the Thais at Bangkok), or one trip each year to the more distant entrepôt of Luang Prabang, via the eastern (Laotian) road. With the subsequent reopening of the Chiang Mai road and the regular extension of Haw trade beyond northern Thailand to the port of Moulmein in the British-administered Tenasserim, Yunnanese caravans began to visit Chiang Mai on a regular annual basis during the winter and dry seasons.

It seems probable that, despite this extension of the westernmost route to Chiang Mai and Moulmein, regional Haw commercial interests continued trading to Kengtung on a biannual basis; certainly Smyth (1898) reports that Haw caravans often made several trips each year between Chiang Mai and Moulmein, returning northwards with the start of the rainy season. This information is confirmed in more recent sources, thus Seidenfaden (1958) records the arrival of Yunnanese Haw caravans in the markets of north Thailand ‘very cold season’, whilst Moerman (ca. 1960), relying on oral sources, reports that in the ‘Old Days’ (1880-1930) they arrived at Chiang Kham every dry season. In sum, therefore, it seems that the Haw caravans travelled southwards to northern Thailand and to Moulmein in the winter and dry seasons (approximately between the beginning of November and the end of April), occasionally returning northwards as late as the beginning of the wet season (May/June onwards).

In attempting to evaluate the size – and value – of the average Haw mule caravan, the historian must rely on the same disparate sources. Thus, according to Macleod, trade between Yunnan and Chiang Mai was “very limited” in 1836-1837, with only about 300 mules ‘coming down annually (but not one third laden)’. This, however, was at a time of political hostility between Chiang Mai and Kengtung (really, as proxies between Bangkok and Ava), and trade was subsequently to increase. By the late 19th century, Haw caravans to northern Thailand reportedly averaged 50 to 75 pack animals (some with as many as 100), controlled by groups of 10 to 15 muleteers, nearly all of whom were Haw Muslims originating from the Dali region of western Yunnan. Despite a subsequent decline in the trade during the first thirty years of the twentieth century – for reasons which are examined below – Le May was still able to report a caravan of 100 mules in the Chiang Saen area during the early 1920s, whilst in the early to mid-1930s ‘large’ caravans, sometimes ‘half a mile long’, continued to arrive at Bhamo.

Statistics for the total number of Haw caravans visiting northern Thailand annually are similarly disparate. Thus, according to Archer (1888), no more than 800 mules, made up of caravans averaging 50 to 100 mules each, passed through Chiang Mai in 1887; eight years later, in 1896, Smyth reported that an average of 15 caravans comprising 15 to 50 pack animals each passed through Chiang Mai annually to 1893, whilst 12 caravans of Yunnanese mules passed though the same city in 1896. These statistics are approximately comparable to those given to Holt S. Hallett by Chao Oo-boon-la-wa-na, ‘the only sister of the queen, and the daughter of the late king’ of Chiang Mai, during the 1880s. This lady, who ‘had long taken an interest in the currents of trade that passed through Zimme [Chiang Mai] ... had endeavoured to arrive at the number of men and animals employed in the caravan trade’. She duly informed Hallett that:

... from 700 to 1000 laden mules and ponies came yearly from Yunnan, and from 7000 to 8000 from Kiang Tung [Kengtung], Kiang Hung [Jinghong], and other places in the British Shan States ...

Moreover, with regard to regional and local trade:

... 1000 elephants are employed in carrying goods to and from Kiang Hsen [Chiang Saen], chiefly for transhipment to Luang Prabang and elsewhere; 5000 porters travelling into lower Burmah, and 4000 to the neighbouring states, and to the British Shan States lying to the north; 3000 laden oxen ply between Zimme [Chiang Mai] and Lakon [Lampang], and from 500 to 600 to lower Burmah.

It is similarly difficult to arrive at generalised figures for profits from the Yunnan caravan trade. Smyth, writing in ca. 1896-1897, estimated the average value of a Haw caravan (up to 1893) at 3,000 to 3,500 rupees (£2,500). These figures are approximately comparable with the near-contemporary estimate of Dr. Cheek (cited in Hallett, 1890), who considered that a caravan of 60 or 70 mules would ‘ordinarily carry merchandise to the value of 12,000 to 15,000 dollars, occasionally a larger amount’. This latter figure is accepted by Curtis (ca. 1900), where it is stated that ‘a caravan of some dozen or so men with some sixty or seventy mules will carry merchandise amounting to some fifteen thousand dollars in value’. But a better understanding of the problems and profits of the Yunnan caravan trade during the late 19th and early 20th centuries may be gained by an examination of the goods carried; the nature of contemporary (and sometimes potentially competitive) local and regional trade; and finally the nature and composition of the Yunnanese long-distance caravans themselves.

Goods Carried by Long-Distance and Regional Haw Caravans

Turning again to our usual earliest source (pace Ralph Fitch, ca.1590), we find Lieutenant MacLeod reporting an outward (ex-Yunnan) traffic in silks (raw and made-up), copper pots, tinsel and lace, with some Yunnanese caravans making a second annual trip to bring rock salt from the neighbourhood of Pu’er. Goods carried back to Yunnan included (primarily) cotton from northern Thailand and Kengtung, as well as ivory and horn from northern Thailand, tea from Shan State and Sipsongpanna, as well as – perhaps – a share in the regional trade driving cattle and carrying betel-nut between the northern Thai principalities and Kengtung. MacLeod also mentions that a ‘horrible trade’ was carried on with the ‘Red Kareans on the right bank of the Salween’ exchanging cattle for stick lac and slaves, but there is no indication that the Yunnanese caravans participated in this latter traffic.

Roadside inn on trail between Dongchuan and Huili, Yunnan c. 1868

Subsequent sources tend to confirm MacLeod's preliminary observations, while expanding and elaborating on the list of goods sold, bartered or otherwise carried. Thus, according to Colquhoun (1883), tea was grown chiefly in the Laos-Yunnan frontier region about five to nine days march south-east of Simao. The ‘most highly esteemed tea-growing district’ was Yi-bang [Yibang], ‘whence the best so-called “Puerh” tea comes’. Tea, together with cotton, was collected from the various trans-Mekong principalities, after which both were taken to Pu’er, ‘the principal entrepot of tea and cotton’, for redistribution throughout Yunnan. Bock (1884) notes that ‘occasionally a caravan of Haw or Yunnan traders brings a fresh supply of merchandise into the city [Chiang Mai], principally wax, opium... scissors and ironware of various sorts, brass bells, skin jackets, silk cloth, numbers of their characteristic huge straw hats, and a small quantity of adulterated musk’. As for exports from Chiang Mai: ‘The articles that they buy up in exchange most eagerly are young deer-horns, for which they will give fabulous prices [for use as an aphrodisiac]. While I was there I saw forty-five rupees paid for one catty [ca. 80 grammes] of them’.

Bock further adds: ‘The Chengmai people were very careful to “ring” every rupee they received from the Haws, for at Muang Prai, a village to the north of Chengmai, spurious rupees are manufactured wholesale’. Bock's near contemporary, Hallett (1890) lists the following exports from Yunnan: silk, opium, iron, copper utensils and pans (from Kunming), straw hats, bee's wax, walnuts, ox bells, silk piece-goods, fur-lined silk jackets, silk trousers, figured cloth, tea, and (ex-Kengtung, southwards): dah (Shan swords), lead, steel ingots, lacquer boxes, together with more tea and opium. Goods carried back to Yunnan from Moulmein and northern Thailand included: piece goods, ‘general goods’, European goods to sell at Kengtung and, above all, cotton. It is interesting to note that, in this very comprehensive list, tea is represented as an item of ‘outward’ trade in contrast to the ‘inward’ designation given by MacLeod. In fact, the best tea was grown in Sipsongpanna [now divided about 80-20 percent between China and Laos], whence it was exported throughout the region, to northern South-East Asia, India, China north and east of Yunnan, and especially to Tibet.

A further interesting expansion of the list of goods conveyed by Yunnanese caravans is provided by W. J. Archer of the British Consular Service, Siam, who noted on his travels between 1887 and 1891 that the Haw caravans carried (in addition to the goods already listed), hats, shoes and ‘other articles of little value’ from Yunnan, as well as (more specifically) salt from Simao; within northern Southeast Asia they also traded in ponies and mules (a point with which Hallett concurs), whilst importing chiefly cotton and woollen products (the latter presumably of European, and probably British, origin) via Moulmein.

Some further elaboration can be gleaned from the writings of Smyth, Curtis, Le May, Seidenfaden and Moerman. Thus, according to Smyth, the Haw exported sheepskin coats and sandals from Yunnan, whilst importing muslins, cambrics, ‘light articles’ and edible birds' nests from Moulmein. Smyth also provides the interesting intelligence that Haw caravaneers were in the habit of giving ‘excellent walnuts’ to the children in northern Thailand as an ‘advertisement’ or inducement to trade. To Smyth's listing Curtis further adds the export from Yunnan of ‘small wares of China’, together with the import of dye woods, gums, stick lac, gold dust and – reportedly – copper. Le May adds ‘cloth’, clothes and sandals from Kengtung, chillies, ponies, bars of lead and the carriage of silver bullion, as well as mentioning itinerant Haw traders in northern Thailand dealing in ‘betel boxes and a miscellany of articles’. Seidenfaden further lists velvet cloth and chestnuts, whilst Moerman, relying on the recollections of the older citizens of Chiang Kham during the 1960s, adds ‘metal goods’, blue silk ribbon for old-style women's jackets, and rice sickles. The list may be further extended by turning to the trade with Burma, where the Haw delivered ‘felt grass caps’, fresh Yunnanese eggs and other foodstuffs for the Yunnanese settler community, and sold ponies and mules, whilst they carried back with them the usual cotton, ‘broad cloth’ and other European manufactured goods.

Trade with Vietnam is less well documented, but the writings of Prince Henry d'Orléans suggest that by the late 1890s, some small items of English and French manufacture (needles and metal buttons) were entering Yunnan via the Red River Valley, whilst cotton (and subsequently cotton thread of Indian and Japanese manufacture) also entered China via this route. Yunnanese opium certainly entered Tonkin via the Red River (Lao Cai) caravan trail, and tobacco from Canton may have reached this remote region of southwest China via the circuitous overland link through Vietnam.

In sum, therefore, we may list the following items as having been exported or imported by Yunnanese Haw long-distance mule caravans during the latter part of the nineteenth and first part of the twentieth centuries.

The list of goods carried by Yunnanese mule caravans can thus be seen to have been very extensive; yet there is no doubt that the most significant export commodities were tea and opium, whilst as an import cotton was far more important than any other item. By the mid- to late nineteenth century opium was cultivated throughout most of Yunnan; by this period opium-smoking was firmly established in China and imports to China from British India were in decline. Opium was also extensively cultivated in the Shan States around Kengtung. Much of the produce was exported to other Chinese provinces, but Yunnan and Kengtung opium were also regularly carried into Tonkin via Lao Cai, and into northern Thailand via Chiang Saen. Although considerable risks were sometimes associated with this traffic in opium, the profits which could be made were of a commensurate nature.

Tea, that other major export by Yunnanese (and sometimes Tibetan) mule caravan, was. Mainly produced in the Sipsongpanna region, in "Kenghung Territory" (Jinghong), in hills “six to twelve marches” south of Simao; in other words, in a district spanning the present Chinese-Lao frontier in the Sipsongpanna-Phong Saly region. Output was reportedly substantial. Davies, who visited Simao in 1895 and reported the trade as being “principally confined to tea and cotton”, reports that the annual tea production was estimated at 15,000 mule loads, or about 900 tons:

Its price averages about three pounds for a rupee (1s.4d.). The value of the annual export would, then, if these figures are correct, amount to roughly 670,000 rupees, say £45,000.

The tea was, moreover, specially prepared for easy transport and sale:

The tea is made up into disc-shaped cakes some eight inches in diameter and one inch thick, weighing about 12 ozs. These cakes, which are called yuan, are then put together in packets of seven, placed one on top of the other, and done up with strips of the outer bark of bamboo. This packet is called a t'ung and it is in this form that it is usually sold. According to my informants, seven-tenths of the tea goes to Yun-nan, Ssu-ch'uan, and other provinces. A good deal of the remainder is sent to Ta-li and other places in western Yun-nan, and to Tibet.

If tea and opium were the major ‘outward’ commodities of the Yunnan caravan trade at the turn of the century, cotton certainly reigned supreme as the chief and only really important import from northern Southeast Asia. As Davies observed in 1895, the greater part of Yunnan was too cold to permit the satisfactory cultivation of cotton, yet:

... as the whole population is clothed in cotton material, the trade in this commodity is one of the largest and most necessary in the whole province.

A considerable amount of cotton was imported into Yunnan from Sichuan Province, but the movement of Haw caravan trains from the south and west also ensured a regular supply of raw cotton from Burma (via Bhamo) and from Kengtung, northern Thailand, and Laos. The raw cotton thus procured –‘a surprisingly bulky commodity, even when compressed, as witnessed by Curtis, into small packs – was carried to Simao for subsequent distribution throughout Yunnan. According to Hallett (1890), cotton purchased in northern Thailand in this manner cost the Haw five rupees a muen (roll). Despite increasing competition from manufactured cotton yarn exported from India, Japan and Europe in the first decades of the twentieth century, the trade in raw cotton from northern Southeast Asia apparently remained profitable until well into the 1930s; thus Beatrix Metford (1935) reported seeing Haw caravans on the road from Tengchong to Bhamo:

... jingling along the track, the mules unladen save for one or two which carried straw for use as fodder whilst traversing the hills. [In marked contrast] the caravans from Burma were heavily laden with bales of cotton yarn and raw cotton to be woven into cloth by the country people in Yunnan.

Composition and Nature of the Haw Caravan

Yunnanese trader on his pony, late 19th century

As we have already seen, the average Yunnanese caravan at the turn of the century comprised 50 to 75 mules, accompanied by 10 to 15 muleteers and a number of well-trained, but potentially savage, guard-dogs. As a general rule the mules – which might in some instances have numbered as few as a dozen or as many as two hundred – were bred by the Yunnanese muleteers, who were considered to be experts in this field. Similarly, as a general rule the muleteers tended to be Haw Muslims of Yunnan, often from the neighbourhood of Dali in the west of that province.

Muleteers – who were generally, though not always, admired by British and French travellers of the time for their toughness and independence – were usually organised into guilds. A contractor seeking transportation of goods by mule caravan was obliged to enter into a contract with a guild leader by which he hired a specified number of muleteers (together with their animals) to be paid at a fixed daily rate. This rate varied according to the season, the nature of the goods to be carried, the distance and difficulty of the route to be traversed, and, finally, the chances of obtaining a return cargo from the intended final destination. It is clear that the caravaneers also carried goods on their own behalf, and doubtless whether working in the latter capacity or ‘on contract’, the illicit transport of prohibited goods (for example, opium and salt) must have boosted the profitability of many a caravan train.

The Haw muleteers themselves were certainly a hardy and self-reliant group of people. At times, during the late nineteenth century, this self-assuredness combined with the hazards and potential profits of their itinerant trading existence to strike a responsive chord in the hearts of those similarly self-assured disciples of the virtues of hard work and free trade, the Victorian British. Thus, according to Colquhoun:

Many of the muleteers are Mussulmans, fine, strongly built, with that air of defiance which the Mahomedan race wears, more or less. None of our horsemen ate pork. The last set we had came from Tali, and were very pleasant, hospitable, manly fellows, and much more to our liking than the Chinese soldiers and police of a corresponding class whom we saw.

Other contemporary authorities were less impressed, however. Thus d'Orléans found them competent but cruel and unjust, whilst H. Warrington Smyth noted of a Haw caravan at Nan:

They are not ideal neighbours, for they are no more sweet of savour than of face or tongue, and their want of cordiality to strangers is only equalled by that of the ferocious dogs which guard the rows of packs in camp. They have a strange dislike to removing any of their innumerable garments even in the hottest weather, and of cold water they have as great a horror as of telling the truth.

The same writer, who wielded an amusing if vituperative pen, records that on the road to Luang Prabang:

We met another gang of Haws, who made the night hideous by discovering the mules had strayed, and every man and boy among them shrieking, howling, beating gongs, and firing guns by way of attracting them back to the camp. It was a pleasant night, with one of my men raving and shouting with fever until dawn.

A more balanced account is given by A. C. Hanna, who worked as a missionary in Yunnan for many years in the early part of the twentieth century:

The Panthays [i. e. Muslim Haw] are a virile, sturdy and aggressive race ... Caravaneers on the Yunnan trade routes are very likely to be Panthays. The men who guide the long trains of mules and ponies through the wild mountain passes of Yunnan and the Burmese frontier, must be rugged in constitution and resolute in spirit to endure this rough life, filled with hardships and dangers. The scanty and ill-cooked food, the long marches, the exposure to all kinds of weather – drenching rains, scorching heat, or, in season, the bitter frost and raw damps of the mountain slopes would indeed daunt any but men of iron mould. Add to all this the rough hard work of the pack saddle, the occasional bite or kick from a refractory mule, the constant danger of attack by armed robbers, and we are ready to admit that the Yunnanese muleteer is no mollycoddle.

Generally speaking, the muleteers day would start between five and six in the morning. The pack animals, which had been permitted to wander freely in the region of the camp in search of fodder, would respond to a ‘curious cry’ of "Hoi! Hoi! Hoi!", knowing that they would receive a ‘scanty meal of rice or horse beans’ before starting on their days journey. The muleteers, too, ate an early morning breakfast at this time (according to Colquhoun, ‘a hearty meal’, though doubtless this varied according to time and circumstance), before preparing the mules, loading the packs, and loosing the guard-dogs which had been guarding the merchandise during the dark hours of the night. The day's march usually commenced at about seven o'clock; a halt was made for about three hours in the middle of the day, during which time the pack animals were again unburdened and permitted to wander in search of fodder, seldom straying far, ‘never beyond easy finding distance’.

An afternoon march of about three hours followed, completing a journey of 60 to 90 li (30 to 50 km), according to the nature of the terrain which had been crossed—a very fast rate of march, quite unmatched by any other form of transport in the ‘broken, mountainous’ country of Yunnan and northern Southeast Asia.

Lead mule of 19th century Yunnanese trade caravan

Pack saddle from Yunnanese mule caravan, late 19th century

The pack animals employed by the caravaneers were invariably a mixture of Haw-bred mules and ‘Shan ponies’, usually in a ratio of five or six to one. The mules thus carried the bulk of the merchandise, but ponies were also loaded with goods and interspersed amongst the mules as they were less likely to panic in emergencies, and exerted a calming effect on the latter. It is generally agreed by contemporary sources that both mules and ponies were ‘sagacious and well-trained’, though there is some disagreement as to whether they always came when called. The Haw employed an unusual type of pack-saddle, apparently not met with elsewhere in the region. These sat on the backs of the mules and ponies, gripping their sides but shielding their spines from any load. Hallett took the trouble to examine and draw a Haw pack-saddle near Chiang Rai (narrowly avoiding an unpleasant encounter with a Haw caravan dog as a result), and reports:

It [the pack saddle] was ingeniously suited to its purpose, and consisted of a light wooden frame formed to the curve of a mule's back, and had a raised arch in the centre to prevent it from resting on the animal's spine and thus giving it a sore back. Saddles and packs are securely fastened to each other, and are loaded and unloaded together ... At the time of loading, a saddle-cloth is placed on the mule's back, the saddle with the packs attached is lifted by two men, the animal passes underneath, and the saddle is placed on its back and kept in place by a crupper and harness embracing the chest and rump. No belly-band is used, and the whole is quickly adjusted.

This particular arrangement had the advantage of speed in loading and unloading, and also ensured that, in times of difficulty, the mule or pony could throw off its load, thereby diminishing the chances of injury in a fall. Yet despite this relatively insecure means of loading the packs, it was apparently surprising how rarely loads were thrown off. Metford, in particular, seems to have had an admiration for these hardy pack animals:

Nothing seems to daunt the Yunnan mule; he is a wonderful little beast, highly trained for pack work. He hardly ever makes a false step, safely carrying his load up and down narrow slippery mountain tracks, along stream beds full of loose stones and boulders, through thick greasy mud, through deep water. But accidents still happen, and many a time I have seen one of my precious boxes tipped off down the khud [Anglo-Indian: a precipitous slope] ... Sometimes, too, a nervous mule would rush past a narrow place which he did not like, and his load would catch a projecting piece of rock, throwing him off the path. But it is very seldom that they are hurt, even though they roll down a steep hillside.

The ‘Shan ponies’, although sturdy and sure-footed, rarely stood more than eleven hands (44 inches, or 112 cm) high, and – according to Colquhoun – came originally from eastern Tibet.

On the march the ponies and mules, although muzzled to prevent delays through grazing en route, were permitted to walk freely in single file, following the lead pony ‘implicitly’. These leading pack-animals seem to have had an important dual role to play, leading the remainder of the caravan by the best and safest path available, and (in a more spiritual realm), warding off natural disasters and ill luck in general. Thus, according to Metford:

The leader is a large mule, usually dun coloured, and is chosen for his strength and intelligence. Round his neck is hung a collar of bells which rings loudly as he walks along, his bridle is covered with red pompoms and fringes, while other gaudy ornaments are stuck in front of his saddle.

A number of sources comment on this strange – and sometimes very elaborate –decoration. Certainly for Bock (1884):

... The most curious feature in these Yunnanese mule caravans is the extraordinary way in which the leader of the drove is decorated. The head of the animal is covered with a mask made of small cowrie shells, beads or seeds, with two openings for the eyes, and surmounted by a high tuft of feathers from the tail of a peacock or an Argus Pheasant, while an additional supply of bells is hung round the neck and shoulders and also round the hind quarters of the beast, and one tuft of hair is suspended beneath the tail, and another on the left hand side of the neck.

This phenomenon was also commented on in some detail by Colquhoun (1883), who makes the feathers used in the head mask of the lead mule or pony those of the Amherst Pheasant (‘much prized as a head plume’), also noting that the foreheads of many Haw pack-animals were ornamented with ‘small round mirrors, set in ornamental frames’, and that (especially in the north-west of Yunnan), many of the caravans had ‘the following animals decked, as well as the leaders’. According to Bock, ‘the object of all these trappings [was] to protect the caravan from the assaults of evil spirits’, and doubtless this was a indeed a major function of pack-animal face masks and other ornamentation. Decoration of mules and ponies was also an indicator of prosperity, however, as well as a source of pride to (and possibly of competition between) the muleteers. Accordingly, a ‘magnificent panache’ was ‘aspired to by all the caravan drivers for their leading animals’, while ‘the poorer [had] to content themselves with coloured balls of wool and other cheap ornaments’.

In a more mundane capacity, lead pack-animals were also employed on occasion to carry a flag indicating ownership or other affiliation of the caravan. Thus, the lead pack-mule of the caravan employed by Prince Henri d'Orléans to travel between Tonkin and Assam via the Song Hong (Red River) Valley and Yunnan during 1895-1896 bore a red flag with d'Orléans' name on it in Chinese characters; similarly Le May, during the early 1920s, reports seeing a Haw caravan near Chiang Kham, its mules ‘bedecked with red rosettes and bells, and ... a three-cornered flag of red, dotted with white stars, the significance of which I could not discover’.

Besides thus repelling bad luck and evil spirits, as well as providing an indicator of wealth and ownership, the lead pack-animal was also selected for its ability to follow orders and to set an example to its fellows. Thus, turning again to Metford:

He [the lead animal] knows the voice of his master and obeys his orders, and woe betide any other mule which tries to usurp his place at the head of the caravan ... The muleteers' orders... are given in a high shriek, and the mules which have wandered far and wide in search of grazing trot back for supper at the welcome sound. There is something wild and barbaric in the long-drawn-out wailing call. When on the march the muleteer gives a shriek and the mules turn to the right, a yell and they turn to the left, a different note and then they carry straight on at cross-roads.

Colquhoun was less impressed, however:

... The caravan men do not manage their animals as ably as we had been led to expect they would do. Instead of the animals obeying the call or whistle of their drivers, at the midday halt, we always found that they took a long time in hunting them out and driving them in; and we never saw them in any caravan obey all the calls, though doubtless one or two generally did. To tell the truth, we were sadly disappointed with the way the horsemen and muleteers managed their animals. McLeod (sic), who saw them in Laos in 1836, describes them all as answering the call of their masters, and running towards them from any distance! Our experience is at a variance with his, on this particular subject.

Yet in considering the overall performance of Yunnanese mule caravans, one must make allowance for the difficulties of the terrain and the relatively great speeds attained and distance covered by the Haw; this is certainly acknowledged by Colquhoun (in other respects a great partisan of the caravaneers) who records:

The caravans cross the rivers very easily. When they are not fordable, the men take the packs and saddles off, and putting them in a boat [presumably provided by a local entrepreneur or maintained in situ by a Haw associate] swim the animals over the river. On the other side the packs are replaced. It is wonderful how rapidly the operation is performed, and how the animals enter into the arrangement.

In the evenings, as during the midday halt, the mules and ponies were initially permitted to wander freely in the vicinity of the camp in search of grazing. If camped in or near a town or lesser settlement, the pack animals were taken care of in some elementary stall or stable provided by the local inn (often under Haw management). Similar facilities sometimes existed at remote caravanserai established along the various ma lu or ‘horse paths’. Turning yet again to Metford:

At night the mules were tied close together in a row by one foot to a stake, and there was hardly room for them to lie down. There they stood under some open shed munching away at some paddy straw. No matter what the weather, even if it forced to spend the night in the open rain or snow, no covering of any sort was given them. Their food was usually unhusked rice, but when that was unobtainable they ate beans or maize with almost equal relish.

Sources by and large agree that the Haw were less careful in the feeding of their pack animals than they might have been – though this is a common enough complaint in the writings of well-provisioned, cruelty-to-animals conscious Europeans of the time, whether trekking in the Himalayan foothills, or in the desert wastes of Saharan Africa. Thus, according to the experienced eye of Colquhoun:

A little paddy or horse-beans at night and some in the early morning, with what they can graze during the day and night, is their allowance of food and it is usually a starvation limit. In fact, the muleteers seem to work their mules upon the next thing to the ‘one-straw-a-day’ principle, a very penny wise and pound foolish one for the proprietor of cattle [beasts of burden]. Nevertheless, both ponies and mules are active, enduring brutes. The horses are surprisingly hardy and game, and have lots of go in them.

Colquhoun's last remark might equally have been applied to the Haw muleteers themselves for, truly, they were also ‘surprisingly hardy’ and must have had ‘lots of go in them’. Generally speaking, the caravans were organised under the command of a single muleteer, the ma-kuo-t'ou or leader of the caravan. Subordinate muleteers were known simply as ma-fu. In the caravan organised by Prince Henri d'Orléans, the ma-kuo-t'ou received a higher salary than the ordinary ma-fu (d'Orléans is not specific on this point, but indicates that the ma-fu received seven taels a month each, payable through the ma-kuo-t'ou; in exchange the latter was expected to ‘cater for the men, do the same work as they, and act as farrier and vet’. In this latter capacity it should be noted that the ma-kuo-t'ou ‘did not omit to ask for an advance to purchase drugs’ for use on the journey.

Muleteer and Son, Dali, late 19th century

As already indicated, on march the muleteers rose early in the morning, generally at about 5 am, to feed and otherwise prepare the pack animals, and to eat their own breakfast. According to Colquhoun:

The muleteers and horsemen live very well for Yunnanese. We have dined pot-luck with them and can speak from actual experience. They have an air of well-fed swagger which – aided by the gay colours which they are inclined to affect and pistols, dagger or gun – gives them a brigand-like aspect. They have, in fact, much the air of traditional stage-banditti.

The same author indicates that the typical working day of a Haw muleteer commenced with a ‘hearty’ breakfast, and indicates elsewhere that, in the neighbourhood of Simao in southern Yunnan: ‘the muleteers and horsemen, whom we met daily in large numbers, were well-dressed and well-fed looking men; they were active and handsome, with a certain air of independence and swagger which was wanting in the peasantry’. Reports as to the clothing of the Haw muleteer differ (though doubtless, this varied according to religious persuasion, region of origin, relative affluence and contemporary ideas of fashion). Thus, as we have already seen, Smyth described them as a dirty people who had a horror of washing and wore many – indeed, ‘innumerable’ – garments even in the hottest weather. By contrast, according to Colquhoun, they wore ‘gay colours’ and were ‘well dressed’.

Nor do drawings and photographs of the time help greatly in establishing a typical Haw style. Thus, in Bock (1884) the Yunnanese muleteers are represented in miniature lithographs as stereotyped ‘Chinamen’ wearing baggy cotton trousers and jackets, together with wide-brimmed, high-peaked hats or smaller, bent-brimmed caps of some soft material, probably cotton felt. Elsewhere in the text Bock refers to their ‘characteristic huge straw hats’ and notes that they smuggle ‘a great quantity’ of opium ‘under their wide coats and trowsers’. Colquhoun, however, represents them (somewhat more flatteringly) in a variety of styles: in a waistcoat and loosely-wrapped turban (‘muleteer of south Yunnan’); rather dashingly, in a jacket and blouse, sporting a soft –perhaps velvet – cap (‘type of Mohammedan face of south Yunnan’); and, finally, in a heavy, belted Chinese-style robe with a skull cap and shield (‘muleteer and son at Ta-li’).

A more recent source, Scott O'Connor, writing of Upper Burma and the Shan States (1904), has good quality photographs showing a Panthay (Haw) muleteer wearing loose robes and a turban similar to that illustrated in Colquhoun, as well as a Panthay muleteer ‘en route to the ruby mines’ [probably at Mogok] in a loose white shirt and broad-brimmed hat similar to those worn today in Vietnam. Finally Le May, writing specifically of northern Thailand in the 1920s, describes the Haw as ‘a quaint folk, with their trousers and tunics faced with red and their dark blue skull caps with red buttons’. Of their footwear – surely an item of singular importance in the long-distance caravan trade –there seems to be no record.

During the day, some muleteers rode amongst their charges, generally on the backs of sturdy Shan ponies. Bock, in his 1884 study, provides a lithograph of just such a Haw muleteer on pony back, sitting on a semi-fixed saddle (like the pack-saddles, without belly band) and apparently without stirrups. In this way they rode throughout the day's march, dismounting for the mid-day break and to assist the mules in difficult places or at times of potential danger. Other muleteers walked or jogged beside their pack-animals ‘at a kind of half-trot, with ... shouts of command and encouragement’. It is not clear how many muleteers proceeded by foot in this manner, or whether they took turns with their fellows mounted on pony back – but it is unlikely that unmounted muleteers of this sort played a major role in the true long-distance caravan trade (for example, to northern Thailand or Chamdo in Tibet), for as has already been indicated such caravans frequently averaged distances of 60 to 90 li (30 to 50 km) each day.

Transporting a mule across the Mekong c. 1895

En route the caravaneers could expect to encounter numerous hazards, both natural and man-made; an interesting instance of the latter, specifically in Burma, Laos and northern Thailand, was the elephant. Etiquette and indeed established custom apparently dictated that, on hearing a mule or ox caravan approaching, elephant drivers should back their charges into a clearing, or otherwise leave the path, to avoid panicking the mules with their huge beasts and causing a disaster. Another obvious man-made hazard at this time was the endemic banditry which afflicted Yunnan. Caravans were, accordingly, often well-armed. Colquhoun, in the 1880s, reported that in southern Yunnan the muleteers ‘nearly all carried arms of various sorts ... this was even more noticeable than amongst those we had met in the east of the province. Tridents, pikes, lances, huge horse-pistols, and a sort of hammer and axe [combined] were amongst the commoner weapons’. Yet the caravaneers were not without a sense of style, too, when profits permitted. Thus Colquhoun notes:

Some of the wealthiest carried fans, as our escort often did; these were often in fine satin cases, elaborately worked with patterns of dragons, elephants, monkeys, peacocks, flowers and other designs...

There must also have been a strong feeling of comradeship amongst the muleteers – as, indeed, befits members of a guild engaged in a difficult and dangerous profession. Thus, as they travelled through the day, they conveyed information to following caravans ‘by some freshly plucked branch, just as do the gypsies of England. Perhaps one road is impassable through floods, perhaps a caravan is journeying over a new road; the green blocks one path, and warns those behind to follow the other’.

Towards evening the muleteers would begin to gather brushwood – in passing, without halting the caravan – for their own cooking fire. If camping in the open, far from the amenities of a town or even a small wayside inn, an appropriate site was chosen – perhaps a dry knoll close by a stream, where the vegetation was sufficiently sparse to discourage tigers, snakes and other unwelcome visitors and a halt was called, usually by 5 o'clock in the evening. Some muleteers would light a fire or fires to cook their evening meal and to lighten the dark hours, whilst others would unload the pack-animals, standing the packs in a long row to be guarded by the (reportedly ‘ferocious but fair’) Haw caravan dogs. The last pack-animal to be turned loose in search of fodder was, according to Metford, the lead mule:

On arrival at the halting place his load is removed first, but saddle and bridle are left till last; even then he has to do his tricks, wheeling round and round, to the right or left, at the word of command, before his saddle and bridle are taken off and he is set free to join the others in a roll on the grass.

Metford further states that the muleteers were ‘most particular about the shoes of their animals’, and that these were examined twice a day, at the mid-day halt and in the evening. ‘Every shoe’ was examined, and any shoe found to be missing was ‘replaced at once’. Yet the care shown to their pack-animals by the muleteers was clearly often of a very rudimentary nature, for the same author continues:

The stony roads are very hard on shoes. Often these game little animals may be seen limping painfully along behind the caravan, usually as a result of the carelessness of the muleteers in driving in the nails. Saddle galls are also very common, but the muleteers do their best to cure them, cutting away the hair, washing out the wounds, putting on mutton fat.

If camping in the open, the Haw caravaneers were able to make do by sleeping ‘with one or two skins underneath, a blanket rolled around them and their bamboo hat cocked up overhead’. Whether, as Colquhoun suggests, they were ‘very happy’ sleeping thus is another question. Wherever possible, when travelling between the few urban centres of Yunnan, the Shan States and northern Siam, the muleteers took advantage of the hospitality of rural caravanserai. Here they could spend a night indoors, enjoying hot food (often halal) and exchanging information and news with members of other caravans.

By contrast, when halting at major regional urban centres such as Chiang Mai and Luang Prabang, the Haw generally camped outside the city walls, ‘depositing the pack-saddles in a row along the ground, to be guarded by their savage dogs’, whilst straight-away beginning to barter or sell the merchandise carried by the caravan train. Within the towns, the Haw generally had their own hostelries, supply shops, etc., often run by Yunnanese expatriates. Thus, according to Metford, at New Bhamo, the major terminus for Yunnanese caravans in Burma, ‘The Chinese shops seemed to cater mainly for the mule caravans, and could supply everything the muleteer needed. Heavy wooden mule saddles, wooden bells with wooden clappers, cruppers with little wooden reels to prevent friction and sores, thongs of raw hide, bridles of plaited hide, many ornamented with red pompoms and streamers, beads and pieces of mirror glass’.

It is clear that, in the course of their travels, the Haw muleteers were exposed to many hardships – notably attack by wild animals or, on occasion, by bandits; the discomfort caused by leeches or the swarms of mosquitoes (and, more seriously, the malaria which often followed the latter); the vicissitudes of the weather, and so on. Hallett describes a simple but effective means by which wild animals and even bandits might be scared away during the watches of the night. Whilst camping by the banks of the Hlineboay River in the Burmese Tenasserim region, Hallett's party was startled several times during the night by what sounded like the discharge of firearms. As Hallett comments, the Haw caravaneers were ‘not likely to waste powder in frightening off dacoits or wild beasts when they had any simpler, equally efficient and cheaper means’ at their command. In the morning, however, all was made clear. The sound of firearms reports was created by heaping joints of green bamboo onto the camp fires, causing the liquid inside to turn to steam and subsequently to explode.

A vivid example of the hardships often caused by harsh or unpredictable weather may be found in the account of Prince Henri d'Orléans who, during April, 1895, experienced a vicious storm in the region of Meng Li in southern Yunnan, close to the northernmost part of Phong Saly Province in Laos. It merits quotation in full, as it conveys in a very real sense some idea of the hardships which might be encountered by the Haw caravaneers:

On the 1st April, in the afternoon, we had made our customary halt for a bite and a rest, when just as we were about to resume, a tremendous storm, which had lowered for some time in the hills, burst over us. Lightning, thunder, wind, rain, hail, - big guns and mitrailleuses - nothing was lacking; the hailstones were as large as pigeons' eggs. Most curious was the aspect of the caravan, as, cloaked in my ample waterproof with my shoulders stooped to the deluge and my sight half obscured beneath my hat brim, I endeavoured to take in my surroundings.

With ears laid back and tails between their legs the animals scattered, driven by the blast and lashed by the hail, the men running hither and thither in a vain effort to collect them. Others of the mafous cowered beneath their blankets, without which, in sober earnest, the hail would have been dangerous. I felt the stones rattle around my ears, and saw naked limbs receiving a far more lively impress of their string. Soon the faces of the men began to show long lines of red like bleeding scars, the dye was running from inside their caps in streaks on their visages ...

Within a quarter of an hour the little stream that before had trickled was a roaring torrent, and we recognised that this route must be impracticable in the rains. The surface became soft and treacherous, and we had to wade through pools widening over oozy ground in which the animals sank to their girths. Each instant saw a load upset in the mud; the men scarce knew when to give the mules their heads, and, to crown all, the path became so narrow that they had to prick them from behind to make them move forward...

Faced with difficulties of this sort, it is not surprising that some of the muleteers turned to opium to ‘give themselves false and temporary energy’. This was far from universal, however, and elsewhere Reitlinger (1939) records that, despite the cold, the Muslim muleteers in his caravan declined to take alcohol as a restorative during the night halts.

The Decline of the Traditional Haw Caravan Trade

The gradual decline of the Yunnanese caravan trade across those areas which today constitute the Golden Triangle may be attributed to a number of factors.

Firstly, following the British seizure of Upper Burma in 1885 the important caravan centre of Kengtung in the Shan States, formerly a tributary of Ava, passed to the British Raj. Similarly, the important caravan terminus at Bhamo fell into British hands. Trade between British-ruled central and southern Burma effectively dated from the Anglo-Siamese Treaty of 1885-1886 by which it became legally permissible for ‘Europeans, Burmese, and Peguans from British Burma’ to enter Siam and Siamese tributaries for purposes of trade. Until 1885-1886 these categories of person had been legally excluded from the internal trade of Siam.

The seizure of Upper Burma in 1885 had the effect of bringing all Burma and Burmese tributary states under the British flag, including the mountainous north-eastern region bordering Chiang Mai, Sipsongpanna and Yunnan. The effective suppression of banditry and encouragement of trade by the British after 1885 had the gradual effect of drawing trade westwards from Kengtung, towards the Burmese lowlands, and away from Chiang Mai and northern Thailand. Increased settlement of Yunnanese migrants in British-controlled Burma assisted in this process. To be sure, this redirection of trade was a gradual matter, developing over at least two decades. Initially, moreover, the British seizure of Kengtung served to facilitate trade between Yunnan and northern Thailand via this route. Ultimately, however, British control of Kengtung and Upper Burma was to drain trade away from Chiang Mai – a development foreseen by the British Vice-Consul in that city, W. J. Archer, in 1892:

Although the route to Moulmein through Chiang Mai is undoubtedly the easiest route from Southern Yunnan to the sea, I have little doubt that, with the improvement of communications in the Shan States, what trade there is will eventually find its way to Burma.

Secondly, within two decades of the British seizure of Upper Burma and the Shan States – a development effectively bringing the Kengtung Region under lowland British-Burmese control – a similar process was to take place in the hill states of northern Thailand. Thus, as a result of the Shan Rebellion of 1902 and its consequent suppression by the Siamese authorities, the northern frontier districts around Chiang Rai, Chiang Saen and Chiang Kham were brought fully under the control of the central government at Bangkok. As Moerman has pointed out, ‘Overseas Chinese’ shopkeepers followed hard on the heels of the victorious Siamese armies, penetrating for the first time into an area which had previously been the exclusive preserve of the Yunnanese Haw, Shan and local northern Thai traders. Thus:

According to my informants, there were no Chinese in Chiang Kham before 1902. They came to sell to the Siamese officials only after a stockade was built. Until the Siamese conquest, regional trading was done by the Haw, the Shan and the local T'ai.

It should be noted that, over approximately the same period at the turn of the century, the French seized control of Laos and therewith the greater part of the Haw caravan route to Nan and Luang Prabang, and immediately began to extend a virtual French monopoly over the trade of Indo-China.

Thirdly, partly as a result of the extension of centralised Siamese authority into the northernmost border regions of the Kingdom, and partly due to the improvements in communication and fiscal reforms implemented by King Chulalongkorn (Rama V, 1868-1910) and continued under King Rama VI (1910-1925), by the first decade of the twentieth century the northern frontier regions of present-day Thailand were inextricably linked to Bangkok and the Central Thai plains via the Maenam-Chaophraya River system. Some indication of the inevitability of this process may be gleaned from the reports of Archer (1892) and Smyth (1898). Thus, according to the former, by this comparatively early date the trade of Luang Prabang was ‘carried on chiefly by Chinese traders with Bangkok’, whilst the latter, in a short report on the Boat Trade with Bangkok, notes that:

By far the greater part of the imports to Chieng Mai and other Lao States come by this comparatively cheap but very lengthy route. The round journey occupies seventy days in high water, and four to five months in the dry season ... [yet despite this] ... The cost of transport by caravan from Burma to Chieng Mai is four or five times that by boat from Bangkok.

Four years after Smyth penned his report, the Siamese suppression of the 1902 Shan Rebellion presaged a road-building programme which would greatly improve communications between northern Thailand and Bangkok; the completion of a railhead at the northern city of Chiang Mai in 1918 further consolidated this process.

Fourthly and finally, the imposition of European colonial rule on Burmese and Laotian territories to the north-west, north and north-east of Thailand during the latter part of the nineteenth century and first decade of the twentieth century brought about a relatively precise delimitation of international frontiers and, subsequently, the imposition of customs duties and tariffs generally designed to redirect trade and profits to markets dominated by the European colonial powers. The Siamese authorities, too, had an interest in taxing and controlling the northern border caravan traffic and in limiting – in so far as was possible – the illicit movement of such commodities as opium and firearms. This delimitation of international frontiers was confirmed in the post-European colonial period and – while illicit border traffic has never ceased in the notorious ‘Golden Triangle’ area and may today even be on the increase – relatively little officially sanctioned cross-border caravan traffic takes place at present in the Golden Triangle frontier zone.

Text copyright © Andrew Forbes / CPA 2008.

An earlier version of this article may be found in Journal of Asian History, 21, 1 (June, 1987), pp. 1-47, together with relevant footnotes and bibliographical references.

Part of Trade Routes