Navigation – Plan du site

The art of not being legible

Invented writing systems as technologies of resistance in mainland Southeast Asia
Piers Kelly
Traduction(s) :
L’art de ne pas être lisible


James C. Scott argued that the traditional non-literacy of highland minorities in mainland Southeast Asia may belong to a wider pattern of state evasion whereby lowland practices, including literacy, are strategically rejected. This position ignores the moral and material value attributed to literacy in upland folklore, as well as the many radical messianic movements that purported to bring writing back to the highlands. I review nine such cases of recuperated literacy among Southeast Asian minorities, all of which were created in circumstances of violent conflict with lowland states. Leaders of these movements recognised literacy as an important vehicle of state power, but their appropriation of writing was limited to very specific purposes and domains. In short, the new literacy practices did not mirror the ordinary bureaucratic uses in lowland states. Instead, writing became a symbolic instrument for building state-like institutions of resistance.

Haut de page

Texte intégral

Fig. 1. Map of Zomia’s writings, 2018

Fig. 1. Map of Zomia’s writings, 2018

Map of Siam’s, Tunquin’s, Pegu’s, Ava Aracan’s kingdoms…, 1764 ; made for Terrain ; the green area is the Zomia.

Source : Gallica / Piers Kelly © Bibliothèque nationale de France


1A tragic motif of lost writing resounds throughout the highland folklore of mainland Southeast Asia. In this vast region, today designated as Zomia (van Schendel 2002), the apparent absence of indigenous literacy is lamented through tales in which writing was lost through profligacy, carelessness, treachery or acts of desperation. The Akha, it is said, kept their writing on buffalo skins but were forced to consume them while fleeing the invading Tai armies. The Wa wrote on oxhide that was eaten in hard times, while the Lahu wrote their letters on cakes that met the same fate. The writing of the Hmong was destroyed by the Chinese or eaten by horses, while Karen writing was variously stolen, eaten or left to rot. Similar stories of being cheated out of literacy and its presumed benefits are told by the Kachin, the Chin and the Khmu’ (for summaries see Enwall 1994; Rastorfer 1994b; Scott 2009: 221-222)

2In his influential work The Art of not Being Governed: An Anarchist History of Upland Southeast Asia (2009), James C. Scott speculates that such stories may indeed reflect historical realities. The ancestors of contemporary highlanders, he suggests, once lived in literate lowland communities. Later, as valley states expanded their reach through aggressive regimes of forced labour, conscription and taxation, the proto-Zomians fled to the hills. Entering into “shatter zones” of anti-state resistance they developed state-repelling practices including swidden agriculture, egalitarian structures and fluid cosmologies. These techniques of evasion ensured that they would not be “legible” to the state, to use Scott’s memorable term. However the question of legibility acquires a more literal character in the appropriately named Chapter 6½. Here Scott argues, not without equivocation, that the absence of literacy in the highlands might also have been a political choice to avoid state enclosure. Exclusive orality, after all, confers certain surprising advantages. When freed from written texts, communities are also released from the strictures of literate orthodoxy where deviations from written dogma can be detected and policed. In oral societies, he argued, genealogies can be reconfigured to justify new leadership arrangements while inconvenient official histories are abandoned as easily as a fallow swidden field. In other words, the permanency of a written text is a potential drag on adaptive cultural dynamism. According to this view, it was preferable for literacy to be destroyed than for its practitioners to suffer their own destruction. The hypothesis is reprised with greater confidence in his subsequent work Against the Grain: A Deep History of the Earliest States (Scott 2017). The early states in the Tigris and Euphrates Valley, he argues, applied writing to the task of population control, typically in the form of taxation and enslavement. For this reason, groups on the periphery of states rejected writing and all that it entailed until it was imposed on them by force (Scott 2017, citing C.C. Lamberg-Karlovsky in Algaze 2001).

3The assumption that literacy is the harbinger of violent exploitation is not without precedent in popular and scholarly literature, from Lévi-Strauss (1955), to Derrida (1974) to Diamond (1999), among others. But Scott’s provocative suggestion overlooks multiple examples of “found writing” from within those same Zomian populations. Among upland groups of Southeast Asia, stories of lost writing are occasionally given narrative resolution through inspired re-discoveries of indigenous literacy. Indeed, the documentary record reveals no less than nine reinventions of writing in Zomian Southeast Asia since the 1840s (see fig. 1). In almost all cases, the recuperation of writing took place at the hands of charismatic leaders with radical political agendas founded on the revelation and diffusion of new scripts. Contrary to Scott’s visions of rootless orality, the recuperated scripts were invariably used to reify highland identities and to cement new institutions and orthodoxies. As I will argue, these new state-like configurations exploited a dynamic of “mimicry and rejection” (Rao 2010), permitting them to be recognisable as literate entities, while remaining illegible to the states they opposed. As a resistance strategy, the introduction of rebel scripts gave impetus to the new movements, investing their advocates with authority, and rendering marginalised languages literally visible.

  • 1 For van Schendel (2002), Zomia extends into the contiguous highlands of South Asia, Central Asia a (...)
  • 2 Several Zomian scripts are not included here. Due to a lack of accessible sources I am not discuss (...)

4In this paper I give a brief overview of the new writing systems of Zomia, as territorially delimited by Scott.1 For historical coherence, I divide my analysis according to the imperial of the former British and French empires, examining first four new scripts of British Burma and three scripts of French Indochina.2

The restoration of writing in colonial Burma

5Many of the instances of defiant non-literacy that Scott alludes to are situated within the national borders of colonial Burma, now Myanmar. Yet it is also among Myanmar’s highland minorities that the earliest accounts of recuperated literacy are recorded. In the 1820s, Karen villagers informed the Baptist missionary Jonathan Wade that God had given them writing on leather but that it had become lost and since that tragic event “we have been an ignorant people, without books, without a king, without a government of our own, subject to other kings and other governments, we have been a nation of slaves, despised and kicked about, trodden under foot by everybody like dogs” (cited in Womack 2005: 68). This legend of lost Karen literacy has many versions and what follows is an attempted amalgam of documented accounts (see McMahon 1876; Stern 1968a; Saw et al. [1980] 2006; Womack 2005).

6In ancestral times, the legend goes, the Karen people were personified as the youngest (in some versions, the oldest) of seven brothers who were each gifted with divine writing by the Supreme deity Ywa. A Chinese brother was given a book of paper, a Burmese brother a book of palm leaf and the Karen a book of animal skin or a book of gold. The other brothers looked after their books and studied them carefully. But because the Karen brother was busy tending his field he entrusted the divine text to a “white” brother who assumed possession of it and carried it over the seas in the company of Ywa. To this day, the Karen await the return of the sacred books, Ywa and the white brother. In a variant narrative the Karen brother was distracted by his work and failed to give due importance to the gift. Thereupon he left the inscribed animal skin on a stump or at the base of a plantain tree where it was left to rot over the monsoon. When he came back to it he discovered to his great dismay that it was chewed up by dogs and pigs, and pecked at by chickens beyond all recognition. Since chickens had consumed the text, he supposed that chickens now embodied the wisdom and lost laws it had contained. For this reason, the descendants of the Karen brother initiated the practice of divination through chicken bones. Having thus lost the capacity to read and write, the Karen were easily dominated by the literate Burmese who expelled them from the lowland plains they had once inhabited.

7Following the first Anglo-Burmese War (1824-1826), Myanmar underwent significant political and religious upheavals. The post-war administrative arrangements led to major internal population movements as well as radical reconfigurations of power structures within the Buddhist monastic institutions. In this climate, Buddhist sects with anti-colonial agendas began to proliferate, particularly in Lower Burma. The leaders of these movements were often charismatic prophets who laid claim to supernatural powers, urged rebellion against outsiders, and advocated an ascetic lifestyle in preparation for the coming of Ariya, the reincarnation of the Buddha (Hayami 2011: 1088-1089).

  • 3 Karen language names have multiple synonyms that are confusing and sometimes contradictory. I have (...)

8In what follows I describe three instances of the “return” of writing in two quite separate communities: the Pwo Eastern Karen of Kayin State and the Western Kayah of Kayah State to the north, historically known as the Red Karen.3

The Leke script

9According to its contemporary practitioners, the Leke religious movement and its script trace their history to the 1830s when Baptist missionaries set out to translate the Bible into the language now known as Pwo Eastern Karen (Saw et al. [1980] 2006). The missionaries adapted the Burmese script for this purpose; another Mon-based script was already in use in Buddhist monasteries for transliterating Karen languages but was not widely disseminated until the 1850s. Both of these new systems, the missionary script and the monastic script, would require the addition of novel or repurposed graphemes to represent Karen-specific phonemes.

10As these new scripts were propagated in the highlands, a Karen man by the name of Mahn Thaung Hlya raised objections to both the Burmanised missionary system and the monastic Mon-based script. Accordingly, in about 1844 or 1845, he scaled the sacred Mount Zwegabin to begin a seven-day fast together with six companions. Mahn Thaung Hlya was the only one to complete the full week of fasting, and on the final day he was rewarded with a vision of a figure dressed in white who asked him what it was that he desired. Responding that he yearned for a native Karen script, the figure touched a flat rock with his staff and immediately there appeared 47 symbols that resembled the scratching marks left in the dirt by chickens (Saw et al. [1980] 2006).

  • 4 I have reproduced most of this narrative from a paper published by current members of the Leke sec (...)

11When Mahn Thaung Hlya came down from the mountain he searched for somebody to decipher the symbols and soon found Mahn Maw Yaing in Hnitya village, a man who was known to have assisted Baptist missionaries in translating the Bible. Though suffering from leprosy, a disease that ensured social ostracism, a house was constructed for him in Don Ying village where he continued to study the letters. In time he provided values for the 25 consonants, 13 vowels and nine numerals for what would become known as the leit-hsan-wait, or the “chicken scratch script.”4 But it was not until the early 1860s that a formal spiritual movement coalesced around the new writing system when divine messengers revealed a holy book written in the sacred script to two men of Hnitkya village (Womack 2005: 155). The book, which is said to exist to this day, became the foundational sacred text of the Leke, a millenarian religion that awaits the arrival of Ariya.

12Significantly, Leke practitioners today have added their own codas to the lost literacy narrative mentioned above. They identify the holy book revealed in the 1860s as the misappropriated “book of gold” that has finally been returned to its rightful owners (Stern 1968a). Meanwhile, for the version in which the book is written on animal skin, they relate that the legendary Karen son burnt down the stump on which he had abandoned the text but discovered that he could foretell the future by interpreting the chicken scratches in the ashes. And since chickens were responsible for consuming or scratching away the divine message, the chicken-scratch script represented a distorted but still powerful emanation of the original gift of writing (Saw et al. [1980] 2006).

  • 5 It is possible that this preference rests on a misunderstanding that “chicken scratch” was a term (...)

13Here I refer to the leit-hsan-wait or “chicken scratch script” as the Leke script, in order to remain consistent with the way earlier scholars have referred to it in English.5 Originally designed for the Pwo Eastern dialect of Karen, the Leke script appears to be an alphasyllabary of the Indic type but with a few unusual characteristics that make it look more like a classical alphabet. The complete grapheme inventory comprises 25 consonants, 16 vowels, 3 tones, 10 digits and one end-of-section marker (Fickle & Hoskin 2013). Rates of literacy in the Leke script have never been estimated, though schools for teaching it have been known since at least 1930 (Marin 1943).

14Today’s Leke leaders are strict vegetarians who wear top knots and white robes. Their followers are urged to adhere the precepts of the dhamma and to dress in traditional Karen clothing at the regular Saturday services. In 2011 Hayami estimated that there were about 200 leaders and over 60 places of worship in Karen State and the Karen refugee camps on the Thai border (Hayami 2011).

Two scripts of Kyebogyi

  • 6 The ethnic provenance of this tale is not clear. Cross attributes it to “the Sgau Karens and by th (...)
  • 7 Jonathan Falla associates the fragment of the brass and gold book with a Sgaw Karen story document (...)

15Another Karen narrative of lost literacy relates that an ancestral band of brothers journeyed west in the company of their elder Chinese brothers, but because the Chinese were able to travel more quickly they often left their Karen siblings behind.6 During the journey the travellers stopped at a stream to collect shellfish. The Karen boiled their catch but the shellfish remained hard, so a few of them went ahead to see how their brothers prepared them. After having observed that the Chinese boiled the shellfish until they were fully cooked before breaking them open to eat, they relayed this knowledge to the Karen camp. As the party resumed its journey west the Chinese continued to outpace the Karen and eventually decided to leave them behind. Before they did so, they deposited their younger brothers’ inheritance on the pathway for them to collect. Thus the Karen discovered a bundle with a bridle-bit, a sickle to cut food for a horse and a book inscribed on a plate made of brass and gold. At this very place, the Karen decided to settle and to build villages, cities and a palace. Later, the Burmese came to destroy their civilisation and so the Karen were forced to flee and regroup in the mountains (Cross 1871; Bunker 1872; McMahon 1876).7

Fig. 2. Writing book, c. 1962

Fig. 2. Writing book, c. 1962

This new writing (kayah li) was created to unify and represent all the Karen languages of Myanmar.

Rights : Htae Bu Phae © All rights reserved

16Like the “lost gift” story discussed earlier, this legend has a coda that brings the narrative into the realm of lived history. Of the various articles inherited from their older Chinese brothers, the inscribed metal book was understood as a real object handed down through chiefly lineages, although the writing on it could no longer be deciphered. Indeed, the Karen Baptist pastor Quala physically examined the plate in question and mentioned its existence to E.B. Cross in 1866. Three years later the Baptist missionaries Alonzo Bunker and J.B. Vinton visited the principle Western Kayah village of Kyebogyi where it was held and each of them were permitted to make sketches of it (Bunker 1872; Brown 1879). The sketch made by Bunker is reproduced in fig. 3 below. The village’s founding headman, Kai-pho-gyee, claimed exclusive ownership of the artefact and was also rumoured to own five ivory plates inscribed with the same script though the missionaries were not able to view them.

Fig. 3. Karen plate, 1872

Fig. 3. Karen plate, 1872

The inscriptions of this writing, of divine inspiration, are a trace of the Karens’ one which would have been lost.

Rights : Alonzo Bunker

17At the time of the missionaries’ visit, Kai-pho-gyee had recently passed away, leaving his wife and followers to resume authority over a cult that venerated the metal plate as its central relic. In the missionary account, the plate was invested with power over life and death and of producing famine or bounty. Every year in March, a feast day was held in which people arrived from neighbouring villages to propitiate the plate with offerings of slaughtered livestock and money (Bunker 1872; McMahon 1876). After the 1870s, no trace of this artefact is found in the archives; its physical whereabouts are also presently unknown. There is no cause to believe that this script was ever readable, let alone disseminated, by those who guarded the object.

18Nearly a full century after the Reverend Quala publicised the existence of the so-named Karen Plate Inscription, another script was to emerge in the very same locale. In March of 1962, a university-educated teacher in Kyebogyi village by the name of Htae Bu Phae spent the night inventing a new script that was to represent and unify the Karen languages of Myanmar. Known today as the Kayah Li script, Htae Bu Phae made no attempt to embed his creation within regional folklore, nor to claim any supernatural inspiration. His motive was secular and political: “I thought of the possibility of a special alphabet for the Kayah peoples since my boyhood […] I wanted to show our specificity against the Burman. I must say that none of the existing solutions met my approval. Neither the burmano–karen, nor the romanized offered the complete panel of the Kayah sounds I recognized” (cited in Bennett 1993). He did however, claim that he based it on “ancient Karenni” on the advice of “some old men in the villages” (Rastorfer 1994a).

19The two scripts that had failed to meet Htae Bu Phae’s approval were, by contrast, rooted in religious institutions. A Roman orthography for Western Kayah had, by the 1950s, already been introduced by Catholic missionaries into some Catholic schools and the liturgy, and in 1962 a Karen Catholic from the town of Hpruso created a Burmese-based script with encouragement from government authorities (Bennett 1993; Sproat 2004). A direct derivative of the Baptist orthography for Sgaw Karen, the new Catholic script was used only in Sunday schools, holiday programs and literacy projects. It was likely to have been this script, with its pretensions to Burmese “nativeness”, that spurred Htae Bu Phae into action the same year. There is no evidence that he took inspiration from the Karen Inscription Plate, nor that he was even aware of its existence.

20Htae Bu Phae soon began teaching the Kayah Li script in Kyebogyi, and by 1975 it had generated enough interest to justify the formation of the Kayah Literature Association, an organisation dedicated both to the promotion of the script and the documentation of local folklore. The following year the separatist Karen National Progressive Party, of which Htae Bu Phae would rise to the rank of General Secretary, proclaimed the script as the official and national alphabet of Kayah State.

21The form of the script was designed to be visually distinctive from both Roman script, associated with Westerners, and the Burmese script of the national government. Indeed, Kayah Li is not automatically reminiscent of mayor regional scripts like Burmese, Thai or Shan (Solnit [1984] 1986), even though nine of its characters are either similar or identical to Burmese (Bennett 1993). But while the script is distinctive in appearance, its underlying structure was intended to accommodate the phoneme inventories of all Karen languages and thus stand to embody Karen political unity. In practice, the Kayah Li alphasyllabary is consistent with other Indic scripts in the way it uses a main character for initial consonants, while employing smaller satellite graphs for vowels and tones.

22In the 1980s, officials in Kayah state began to print and circulate school books, dictionaries and magazines in the script but its main sphere of use was to be in refugee camps on the Thai side of the border. The turbulent political context in which the script arose is reflected in two alternative labels for Kayah Li writing: “rebel literacy” on account of its origins within a separatist movement, and “camp script” because of its later prevalence in Karen refugee camps. By 2001, a Thai government survey estimated that in the Karen refugee border camps literacy levels in the Kayah Li script were at 48 per cent, placing total Kayah Li literacy at around 10,000 (Sproat 2004).

The Pau Cin Hau script

23The Chin people, who occupy the highlands of northwest Myanmar and neighbouring Bangladesh, account for their non-literacy with the following story. The ancestral mother Hlinyu laid a clutch of eggs from which 101 brothers hatched. The youngest was the Chin who wandered far away and by the time he returned, the world had already been carved up among his older brothers. In compensation for this misfortune his mother gave him the mountains, along with elephants, horses and other livestock, and placed him under the tutelage of an elder Burmese brother. The brother pretended to instruct him in the art of writing but instead showed him only the back side of the writing slate. Subsequently he used other tricks to cheat the younger Chin out of all of his livestock. The Burmese brother was careful to mark the borders of his territory with stones and pillars, but persuaded the Chin brother to rely on tufts of grass which soon burned away in jungle fires. So it was that the Chin were left destitute, illiterate and stateless in the mountains (Shway Yoe [1882] 1910: 443-444).

24The Chin have been described as existing in a state of perpetual warfare with the Burmese, and in earlier times with the Shan (Banks 1967). Even today, the Chin National Front, founded in 1988, continues a military campaign for regional autonomy. The mythical loss of writing, however, was resolved at the hands of a legendary prophet by the name of Pau Cin Hau. Apprenticed to a prophetess in about 1888, Pau Cin Hau was to suffer a long illness and in 1900 he began having a series of visions that soon elevated him as a prophet in his own right. The visions foretold of new technologies such as railway trains and steamships and of devastating wars between nations. An Englishman appeared in the visions and held a mirror up to his face in order to force Pau Cin Hau to confront the calamities. By 1902, the Englishman returned to Pau Cin Hau to instruct him in the art of writing using stones in the shape of letters. When he had finally mastered the system he woke from the dream and began copying down the letters in a frenzy. The Deputy Commissioner of the Chin Hills was impressed with his invention and encouraged him to publish it, but Pau Cin Hau decided to keep revising the system until the final version was eventually completed in 1931.

25A year after his dream of writing, Pau Cin Hau received another vision which he interpreted as a divine order to reject the practice of propitiating nats (animist spirits) and to worship only one all-powerful god. Thus began the monotheistic Laipian movement, a Christian-inspired religious tradition that continues to this day in the Chin Hills and whose primary rituals are concerned with healing. The consecrated healers are known as paliki and wear official-style uniforms while ministering their services. If the illness does not yield to lower-ranking paliki, higher-ranking officers are called for until the patient recovers on the principle that bad characters are supposed to yield to law enforcement (Bennison 1933: 218). Indeed, the term paliki is derived from the English “police”, via Burmese, and this etymology is recognised by Laipian followers (Banks 1967).

26Several books of Laipian ritual songs are published in the Pau Cin Hau script suggesting that the writing system itself has some role in healing activities (Pandey 2010a). Other manuscript corpora include translations of the Bible (Bennison 1933) and copybooks for learning the system. Pau Cin Hau’s revealed script would proceed through three versions. The first was a logographic script of at least 923 graphemes standing for monosyllabic morphemes. The second version has not survived, but the third and final revision, issued in 1931, is an alphabet of 21 consonants, 7 vowels, 9 coda symbols, and 20 tone marks (Bennison 1933; Pandey 2010). Although the Pau Cin Hau script was designed for Tedim Chin, the inclusion of signs for consonants that are non-phonemic in this linguistic variety is a powerful indication that the system was designed to be used for other more widely spoken Chin languages such as Mizo (Pandey 2010).

27E.P. Banks has claimed that the Laipian religious movement is a product of stimulus diffusion since Pau Cin Hau had no direct contact with Christian missionaries despite its monotheistic character. More provocatively he maintains that the writing system was also a secondary invention, and by implication that Pau Cin Hau was non-literate at the time. Certainly, Pau Cin Hau’s earliest creation of the script in 1902 pre-dates the missionary development of a Roman orthography for Chin languages some eight years later. Moreover, the early logographic script shows superficial similarities with early versions of other stimulus diffusion scripts in so far as it is a syllabary with non-systematic graphic stereotypes. By the third revision, however, it is likely that Pau Cin Hau had familiarised himself a little with Roman and Burmese scripts. A number of graphemes are sourced directly from Burmese and Roman, including Burmese tone markers and punctuation conventions (personnel communication Chris Button). The letter names are also inspired by those used in Brahmic systems, even though the Pau Cin Hau system itself is not alphasyllabic.

Literate rebels in French Indochina

28Although the French and British colonial empires of mainland Southeast Asia may have appeared to be worlds apart, they were united geographically by the massif stretching from the Chin Hills all the way to the uplands of southern Laos. The Zomian fringe of both empires were also remarkably similar in their patterns of cultural and military resistance. As we will see, Zomians under putative French rule were also prone to millenarian uprisings and strategic reappropriations of literacy.

The Khom script

29Just as the Karen describe a past journey in the company of Chinese brothers, the ancestors of all Kha people, including the Jruq, are said to have journeyed from the north. In one account, the Kha set out from Vientiane, led by magicians bearing a magical sword, with the Lave people marching at the front. The Jruq lingered behind until they were overcome by fatigue and decided to settle in their present location. The Jruq, like the Karen, also describe an ancestral appropriation of foreign technology: they learnt weaving from the Alak, house construction via the Lao, and coffee cultivation from the French (Fraisse 1951; Lavallée 1901; cited in Jacq 2001).

30In 1924, military personnel in French Indochina began to intercept insurgent correspondence that was emanating from the Boloven plateau in southern Laos and circulating as far as Vietnam and Cambodia. What unified these messages, over thousands of kilometres of upland terrain, was the unusual script in which they were written. Those literate in Roman, Thai, Chinese or Burmese would have made no headway with the script, though one official supposed that it might be related to Pali. The source of the mysterious messages was none other than the “immortal” rebel leader Ong Kommadam who had risen to prominence among the Jruq and neighbouring groups of southern Laos with a program of rebellion against the colonial state.

31Kommadam received the miraculous script after going into involuntary trances. It is said that he would repeat a sound again and again until its corresponding symbol manifested itself on the bare skin of his chest. An attendant scribe copied it out and waited for the next sound to appear. These sessions left Kommadam exhausted and he would sometimes sleep through the following day (Sidwell 2008). Though Kommadam was to proclaim himself alternatively “King of the Khom” or “Sky God of the Khom,” he was by no means one of a kind. At the turn of the century, the Boloven plateau hosted multiple short-lived revolts by self-styled messiahs that were referred to as Phou Mi Boun, literally “men with merit”; in 1902 Kommadam was merely one among more than 100 such competing holy men (Gunn 1988; Ileto 1992). The majority of these micro-movements either fizzled out or were defeated militarily, but Kommadam, and a handful of others, would rally their followers to lead a unified revolt of the southern highland groups that lasted several decades.

32The causes of the rebellions have been much discussed but most point to sudden changes in economic patterns following incorporation into the French colony (see especially Gunn 1990). After the colonial occupation of the Bolaven plateau in 1893, the Jruq and their neighbours were subjected to higher taxes, while their means for accumulating resources were simultaneously curtailed (Murdoch 1967). With the banning of the slave trade, the introduction of compulsory corvée labour, and the diversion of local trade away from Ubon and Bangkok and towards the colonial centre of Saigon, highland communities came under enormous economic strain. In addition, customary tribute relations to the kingdom of Champasak were disrupted while Lao elites lost their tax rounds and political status to ordinary colonial bureaucrats (Jacq 2001). All this created the conditions for a major political crisis in the highlands. Those who had lost everything with the new regime had nothing to lose in opposing it with all force, bringing a “crisis of authority, power relations, and social relationships” (Murdoch 1967; but see Salemink 1994).

33In 1910, Kommadam had survived a close-range, and somewhat theatrical, assassination attempt on the part of a French official, who, under the cover of peaceful diplomacy, fired at Kommadam’s chest with a pistol secreted in his pith helmet. Kommadam’s reputation for immortality remained intact until French authorities were determined to put an end, once and for all, to the “Kommadam question.” In 1936 he was killed in a raid on the Boloven plateau, staged through an anachronistic combination of aerial bombardments and elephant brigades (Gunn 1988). Use of the script ended after Kommadam’s demise, and French forces allegedly destroyed all surviving manuscripts that they could get their hands on. A local story relates that after Kommadam’s death, his sons transcribed the Khom characters from his tattooed back before he was buried. In another version, one man—perhaps Kommadam—had himself tattooed in the Khom script before fleeing to the Wat Phu temple where he is rumoured to reside as a monk (Jacq 2001: 521).

34The fact that the Khom script is exhaustively recorded in the public documentary record is due to the efforts of Pascale Jacq and Paul Sidwell. In 1998, the pair interviewed Ong Kommadam’s elderly secretary who had retained literacy in the script. They were later to be approached by Kommadam’s grand-niece who has successfully preserved her great uncle’s notebooks and allowed them to be copied. Jacq also managed to obtain samples of the script from French archives.

Four Hmong scripts

35The stories of lost literacy told by the Hmong of Southeast Asia are related in the wider context of a past exodus from China. In its broad outlines the Hmong maintain that they once had their own king, territory and writing system until these were destroyed by the invading Han who forced them to flee south. In an alternative account, the Hmong carried their books with them but were forced to eat them to prevent starvation, or they fell asleep exhausted from their journey only to wake and find their horses had eaten them. A common variant has the Hmong losing their script at a river crossing in the company of their Chinese brothers. Afraid to get their books wet the Hmong ate them while the Chinese carried theirs on their heads. Or a young Hmong brother put the script in his mouth so that he could carry his older Chinese sibling on his back to ford the river, but he stumbled and swallowed the script. In one version the Hmong make an effort to retain their writing by stitching the characters into their clothes, and traces of it can now be detected in contemporary Hmong embroidery. In all such tales, the destruction or ingestion of the script accounts for why the Hmong have a good memory, but its loss is also associated with a loss of territory, political status and the opportunity to become officials. Rounding out the tragic narrative is a prophecy that an authentic Hmong script would fall from heaven into the hands of a prophet who would unite the dispersed populace and usher in a restored Hmong kingdom (Lemoine 1972; Tapp 1989; Enwall 1994; Duffy 2007).

  • 8 The best candidate for documented ancestral Hmong script comes from the early 20th century, when H (...)
  • 9 See Jiang Yongxing (1989) discussed in Enwall (1994, 65). Note that Enwall is sceptical about the (...)

36It is possible that these Hmong stories have at least some historical basis. Numerous “sightings” of ancestral Hmong scripts have been reported over the centuries but none have been positively documented or confirmed (for a summary see Enwall 1994: 59-72).8 Nonetheless, the armies of the expanding Manchu dynasty invaded and displaced thousands of Hmong communities in China and at least one source mentions a decree banning the remaining Hmong from using or teaching writing.9

37The Hmong refugees from China first entered Laos and Vietnam between 1810 and 1820. The first prophet to bring writing back to the Hmong was the rebel Pa Chay Vue who, in 1918, led a revolt against French rule in Laos in the so named War of the Insane (“Guerre du Fou”). Pa Chay asserted himself as the promised Hmong king and sought to establish the capital of the Hmong nation at Dien Bien Phu. Said to have been born carrying a book and a pen, he claimed to have visited heaven where he met four madmen who knew how to write and with whom he communicated by means of a single letter. It was this otherworldly writing system that he later used for writing propaganda tracts urging Hmong to rise up and join his movement. These messages, written with natural ink on processed bamboo bark, also contained accounts of creation to the coming of the Hmong king, and were carried by messengers, who presumably explained their contents to non-initiates, into Hmong villages. Pa Chay also distributed squares of cloth inscribed with his script as amulets for protection in battle and in an echo of the Karen inscription plate, he bequeathed a copper tablet engraved in the script to his wife before he died (Tapp 1989; Duffy 2007). No sample of Pa Chay’s script has survived, and so it is possible that it was a non-linguistic mnemonic code, as opposed to a fully glottographic system. Non-linguistic systems have been recorded among the Hmong of China, including charcoal pictographies used in war correspondence and a system of communication via notched bamboo sticks, whereby the number, location and shape of notches is associated with specific meanings established in advance (Savina 1924; Lemoine 1972).

38Pa Chay Vue was assassinated in 1921 but other Hmong messiahs would later arrive amidst the violent turmoil of the Vietnam War. The first brought two writing systems, one of which survives to this day and which some of his followers believe is a restoration of the lost Pa Chay script. The new messiah was an illiterate Hmong farmer by the name of Shong Lue Yang who had a dream in which two young men asked him if he had already taught his people writing. The farmer insisted that he was illiterate but the young men told him that they already taught him the script. Shong Lue Yang grew to marry and have children, and in May 1959 he heard a loud voice telling him to build a round house temple, make paper from bamboo and ink from the indigo plant and to prepare opium smoking equipment. He smoked until midnight when the two young men of his childhood appeared to him again and gave him two writing systems: one for the Hmong and the other for Khmu’, the language of his mother. Having taught him the scripts, two more visitors appeared wearing the black uniforms of officials adorned with gold medals and bringing offerings to Shong Lue Yang and his teachers. When his infant twins passed away (understood as personifications of his ethereal visitors) he received a post-mortem message from them written in the script, reminding him that Hmong writing had been destroyed by other nations in the past, and prophesied that those nations that continue to destroy it will be destroyed themselves. From that time on, he began teaching the two scripts to the Hmong and Khmu’ people. The Hmong version, known as Pahawh Hmong, went through four progressive revisions between 1959 and 1971 when Shong Lue Yang was assassinated, with each version becoming increasingly compact. The Khmu’ script did not succeed and no manuscripts have survived. Prior to his death, Shong Lue Yang designed a flag for the coming Hmong nation and prophesied the arrival of new technologies (Smalley, Vang & Yang 1990).

39A final Hmong script entered the scene in 1965 just outside the village of Muang Pieng in the Laotian province of Sayaboury. Here a mysterious deity revealed a holy script and several sacred texts to an eight-year-old boy by the name of Nga Va. For seven years Nga Va laboured to learn the script and decipher the texts which were discovered to be founding documents for a future Hmong nation. The eight illustrated volumes, encompassed a guide to learning the writing system, social and philosophical commentaries, instructions for future government and education, as well as moral treatises and blueprints for future flags, currency and government buildings. In contrast to Pahawh Hmong, knowledge of the Sayaboury writing system remains highly regulated, as does access to the sacred texts. It is not known whether the script is still in use today (Culas 2005; Smalley & Wimuttikosol 1998). The literacy of Nga Va, and of his father who assisted him, is also unknown however Smalley and Wimuttikosol proposed that the Sayaboury script is perhaps a kind of substitution-cypher of the standard Roman Popular Alphabet for Hmong languages (Smalley & Wimuttikosol 1998). Shortly before he passed away Smalley reported the existence of three additional Hmong writing systems and two ‘embroidery scripts’ created between 1960s and the 1990s though these did not apparently diffuse beyond their inventors and no samples are available.

Chapter 6¾

40In the preceding discussion I have described nine potent cases of reinvented writing in Zomia. Taken on their own terms, these events complicate Scott’s suggestion that non-literacy in the highlands is a political manoeuvre calculated to evade transcription into literate lowland regimes. Certainly, highland folklore is replete with narratives of ancestral writing that was lost due to the moral failures of its owners or at the hands of domineering lowlanders. Yet as we have seen, these narratives are hardly rallying cries for anti-state resistance and are instead expressed in a register of regret and humiliation. Nevertheless, in the examples reviewed here the stories are seen to hold a proleptic promise: the eventual recuperation of writing is formulated in terms of moral redemption, a realignment of power relations and a restoration of past glory.

41The return of writing among the Karen, Jruq, Chin and Hmong was heralded by charismatic leaders at the forefront of new politico-religious movements. These men strategically reappropriated writing among other markers of statehood to build new visions of society that were recognisable-yet-illegible to their lowland antagonists. Arriving in the wake of the Anglo-Burman war, the revelation of the Leke script to the mystic Mahn Thaung Hlya was one among three new orthographies for the Pwo Eastern Karen language developed in the same tumultuous period. The Leke script, however, held a significant attraction for speakers of Pwo Eastern Karen. It did not dress up the language in a Burmese livery as the missionary and monastic scripts did, and as such it remained illegible to non-initiates. Moreover, its outward form bore traces of its mythical origins presaging the return of the Karen’s legendary lost book in the guise of a new scriptural tradition. Further to the north, and at approximately the same time, a charismatic Karen headman purported to retrieve the gift of writing as it had been bequeathed by ancestral Chinese brothers who had long ‘outpaced’ them in technological prowess. Though indecipherable, the lost book became central to the ritual paraphernalia of the Kyebogyi chiefdom at a time when it was threatened by the Burmese to the south and a rival politico-religious movement to the east. A century later, and under renewed hostility from the Burmese military, another writing system emerged from within the very same village. Though avowedly secular, the Kayah Li script was invested with the responsibility of symbolically uniting Karen linguistic groups while stabilising a volatile separatist institution.

42Beyond Myanmar, the Khom script of Laos was developed in the heat of a brutal war by a charismatic rebel who received it via a trance-like revelation. Despite its creation at the hands of a renowned miracle-worker, the Khom script, as far as we know, was not to become the basis of a sacred written tradition. Rather, its deployment as a medium of political propaganda had a practical application in uniting networks of highland groups in French Indochina in opposition to the colonial state. The Hmong scripts of Laos and Vietnam are among the better-known messianic scripts of Southeast Asia. Conceived in the turmoil of anti-colonial uprisings and the Indochina Wars, these writing systems presaged the return of a Hmong king and a utopian nation-building program.

43All these cases adhere to a similar routine: a) storytellers narrate a critical event in which literacy, and its presumed advantages, was lost; b) a messiah or king is prophesied c) an external upheaval manifests itself as an internal crisis of leadership and allegiance; d) a charismatic individual rises from the chaos attracting followers by calling for a restoration of a past order; d) revelational events bring writing back to the communities in the form of a distinctive script; e) the script is deployed by an emergent politico-religious institution. There is, no doubt, a risk of reductiveness in this formula. After all, of the countless millenarian movements that have come into existence in Zomia in the past two centuries, only a handful made writing part of their ritual or political program. Yet even within non-literate movements, the prophecy cycle is broadly similar, even if the place of writing is occupied by other symbols. Typically a perceived shortcoming lamented in folklore—slowness, sickness, technological deficiency—is redressed via talismans or supernatural gestures, and a lost preeminence is restored (see for example the events discussed in Jacq 2001: 12-13).

44In addition to the general millenarian cycle outlined above, other aspects of recuperated writing across Zomia are remarkably congruent. Virtually all stories explain inter-ethnic conflict as an ancestral condition via the analogy of an unequal sibling rivalry whereby the dominating brother has access to resources denied to his illiterate protégé. A violent crisis challenges these pre-ordained power relations, leading to the intercession of a stranger bearing the gift of writing (a theme reprised by West African inventors of writing). When the oppressed brother takes repossession of his inheritance, and the authority it entails, the balance of power is readjusted and more technologies are predicted to fall into his hands. However the new written tradition that emerges is radically different to those of the lowlands. Zomian prophets never use writing for organising and controlling their followers in bureaucratic institutions, and rarely even ordinary communication. Rather, the raison d’être of a Zomian script is to signal a new location of power. This is achieved in two operations. First, the script publicly identifies its practitioners as members of a specific highland ethnicity. Second, use of the script consolidates and stabilises a new politico-religious institution of nativistic inspiration.

45How does Zomian writing actually realise these ends? Edmund Leach’s seminal ethnography of highland Burma drew attention to the ambivalent nature of ethnopolitical affiliation for Zomian groups (Leach [1954] 1964). It is possible, he observed, for a highlander to affirm membership of more than one language-group, religion and clan lineage simultaneously. Far from undermining claims of identity, Leach argued that these very inconsistencies were essential to the maintenance of the social system. It was only in the act of expressing or performing the idea of cultural difference that such quantum uncertainty is temporarily (and strategically) resolved for specific witnesses in specific contexts (see also Giddens 1984).

46“Unless they speak,” lamented the Kayah official Nyo Min in 1986, “we cannot know the place of origin of our folk now.” Nyo Min’s comment was in response to the gradual loss of traditional dress that once distinguished Karen subgroups from one another, as well Karen people as a whole from other ethnicities. The indigenous writing systems under discussion here might be regarded as an alternative form of national dress that makes language itself a visible and material object of display. (Just as, for Hmong storytellers, the embroidered motifs in their traditional dress are traces of their lost writing.) All that non-literates are capable of “reading” is the expression of difference itself, while languages that are unwritten, or are represented in a foreign system, are doomed to remain heard-and-not-seen. In effect, the identity-marking work of language, familiar from Labovian sociolinguistics, is made visible in such a way that the auditory distinctiveness of spoken language is reproduced in the graphic distinctiveness of a script. The very fact that language and script are so often collapsed in widespread linguistic ideologies is an eloquent demonstration that writing systems partake of the same power of language to mark affiliation. The pervasiveness of this idea is witnessed in the way that the conventional labels for scripts—Arabic, Burmese, Chinese, Thai, etc.—are so often derived from the language of the dominant group that uses them, regardless of what languages are actually being represented in them. This conflation, in part, accounts for why charismatic Zomians chose not to adopt existing scripts that were more readily available to them, even when they may have already been literate in such scripts themselves. It also explains why their scripts were radically distinct from those of their neighbours in their graphic appearance, a phenomenon described by Houston and Rojas (forthcoming) in terms of ‘rupture’ from a dominant model.

47Zomian prophets created new writing systems as irreducible markers of Karen-ness, Chin-ness, Hmong-ness etc., but this act of strategic essentialism was in tension with a contrary need to unify highland communities in utopian movements. The inventors of Karen, Chin and Hmong scripts reached an ingenious solution to this dilemma by expanding the phonological range of their systems. In such a way, their more generous grapheme inventories were able to represent sound contrasts in neighboring languages while still marching under the same flag. Foreign and missionary scripts, on the other hand, remained tethered to the prestige varieties they were originally invented for.

48The standing challenge for any nativist movement is to strike a compromise between a hard ethnocentrism and a more inclusive utopianism. Among peoples that are traditionally egalitarian and acephalous, the task of nation-building is quite literally a Year Zero project. The symbols and technologies that such communities choose to appropriate in these circumstances is therefore especially revealing. That Zomian script inventors desired the power of state bureaucracies is witnessed in the official uniforms of Laipian healers, in the Pahawh Hmong and Khmu’ origin stories, and in the Sayaboury manifestos. Consistent with analogy of ethnic clothing styles elaborated earlier, writing might also be seen to embody the values of institutional power. After all, literacy almost always requires a degree of institutional investment in its creation, acquisition and diffusion. Unlike languages, no script can be acquired naturally through ambient exposure, meaning that literate individuals are marked as having passed through an institutionally mediated process. They are no longer casual members of a society, as native speakers within an ordinary linguistic community, but are transformed into privileged insiders displaying hard-won knowledge. Pierre Déléage, who emphasises the “necessary and constitutive relationship between a form of writing, a discursive genre and an institution” (Déléage 2013: 12), has described how the graphic codes devised by messianic indigenous prophets in colonial North America were bound to specific orally mediated rituals. In reaction to the rival ritual agenda of Christianity, which diffused alongside a powerful printed scripture, the new native discourses were designed to appropriate the prestige of the Bible and to place Native American prophets on a more equal footing with missionary proselytisers. However, the durability of these indigenous codes was contingent on the stability of the ritual institutions to which they were bound. Likewise, Zomian prophets created writing systems that were bound to canonical scriptures (Leke, Sayaboury, Pau Cin Hau), ritual objects (Karen Inscription Plate, Pa Chay) or military-political projects (Kayah Li, Khom, Pahawh Hmong, Pahawh Khmu’). The fortunes of these scripts have risen or fallen on the back of the institutions in which they were propagated. Only one, Pahawh Hmong, has succeeded in extending its range into new and diverse discourses, though Kayah Li shows similar promise.


49What remains of Scott’s suggestion that writing is a dangerous encumbrance for non-state minorities? If true, this might explain the relatively slow pace of literacy diffusion in highland Southeast Asia, beyond the unique instances discussed here. Suspicion of literacy among indigenous populations is not unknown (see for example Brandt 1981), nor is outright rejection of it (Hill & Wright 1988). Yet the power attributed to writing may also be attractive even as it is feared, and whether it is desired or loathed, there are no groups that are indifferent to it. It is incidentally remarkable that literacy in Zomian folklore is represented as universal norm and that it is non-literacy that demands explanation. This is despite the fact that worldwide rates of literacy did not rise to around 50 percent until about 1970 (Roser & Ortiz-Ospina 2018). As I have shown, the movements that promoted indigenous writing systems were responding to changing relationships between highland and lowland communities, mediated by new forms of colonial contact. Outside agents presented a threat to traditional ways of life either through taxation and labour regimes, war, or the introduction of Christianity. All of this took place against a background of rapid globalisation across the nineteenth and twentieth centuries. Under these conditions it is not surprising that Zomians may have felt the need to assert ethnic difference in more concrete terms while resisting “legibility”.

50As we have seen, new Zomian traditions of writing rarely solidified or extended their discursive range. With this understanding, the literacy-rejection hypothesis might have more explanatory power if we pay more attention to different regimes of use, rather than leaning too heavily on a simple binary of literacy vs. non-literacy. Lowland states employ writing in the service of accounting, administration and other mundane communicative concerns. Highlanders, meanwhile, tend to experiment with forms of “restricted literacy” (Goody 1968a, b) that resist universalisation and may be inherently short-lived, consistent perhaps with other cultural modes of evasion outlined by Scott.

51Nevertheless, there is no doubt that both literate and non-literate Zomians understood writing as a political technology with its own material density, capable of cementing identities, relationships and institutions. It is telling that in much of the folklore outlined in this paper, literacy is characterised as a tangible possession that may be gifted and owned, but which is also subject to theft, consumption, decay and rediscovery. Those who succeed in acquiring and protecting the ‘book’ are rewarded with political and spiritual power, while possession of same object by enemies explains their capacity to dominate. Anthropologist F.K. Lehman, whose primary fieldsite was Kyebogyi (where two of these scripts appeared), noted that successful Karen messiahs rise to prominence “by manipulating a real or imagined knowledge of, and experience with, the working of the larger, civilized world” (Lehman 1967: 34). In the same way, Zomian charismatic movements do not proceed from a rejection of statehood and its associated technologies of domination, but in a demonstrated ability to channel these for the benefit of the faithful.

Haut de page


“Initial social complexity in Southwestern Asia: The Mesopotamian advantage”, Current Anthropology 42/2, 199–233.
DOI: 10.1086/320005

BANKS E.P. 1967.
“Pau Cin Hau: A case of religious innovations among the northern Chin”, in Carroll Riley & Walter Taylor (eds.), American Historical Anthropology: Essays in Honour of Leslie Spier, London & Amsterdam, Pfeffer & Simon, 37–59.

Kayah Li Script: A Brief Description, Urbana–Champaign, University of Illinois.

BENNISON J. J., 1933.
Census of India, 1931. Burma: Part I.–Report, vol. XI, Rangoon, Office of the Supt., Government Printing and Stationery, Burma.

“Script as a potential demarcator and stabilizer of languages in South Asia”, in Hugo C. Cardoso (ed.), Language Endangerment and Preservation in South Asia, Honolulu, University of Hawaii Press, 78–99.

—, 2018.
“Writing off domination: The Chakma and Meitei script movements”, South Asian History and Culture, 9/1, 116–140.

“Native American attitudes toward literacy and recording in the Southwest”, Journal of the Linguistic Association of the Southwest 4/2, 185–195.

“On a Karen inscription”, Transactions of the Asiatic Society of Japan 7, 127–129.

“On a Karen Inscription-Plate”, Journal of the American Oriental Society 10, 172–176.

CROSS E.B., 1871.
“On the Karens and their language”, Journal of the American Oriental Society 9, xixii.

Le messianisme hmong aux xixe et xxe siècles. La dynamique religieuse comme instrument politique, Paris, CNRS Éditions.

Inventer l’écriture. Rituels prophétiques et chamaniques des Indiens d’Amérique du Nord, XVIIe-XIXe siècles, Paris, Les Belles Lettres.

Of Grammatology, Baltimore, Johns Hopkins University Press.

Guns, Germs and Steel: The Fates of Human Societies, New York & London, W. W. Norton.

DUFFY JOHN M., 2007.
Writing from these Roots: Literacy in a Hmong-American community, Honolulu, University of Hawaii Press.

A Myth Become Reality: History and Development of the Miao Written Language, t. 1, East Asia Monographs, Stockholm, Institute of Oriental Languages Stockholm University.

“Revised proposal to add the Leke Script in the SMP of the UCS”, Universal Multiple-Octet Coded Character Set International Organization for Standardization, 1–14.

“Les villages du plateau des Bolovens”, Bulletin de la Societé des Études Indochinoises 26, 53–72.

The Constitution of Society: Outline of the Theory of Structuration, Cambridge, Polity Press.

GOODY JACK, 1968a.
“Introduction”, in Literacy in Traditional Societies, Cambridge, Cambridge University Press, 1–26.

—, 1968b.
“Restricted literacy in northern Ghana”, in Literacy in Traditional Societies, Cambridge, Cambridge University Press, 199–264.

“Sambran (The White Python): The Kha (Lao Theung) revolt of 1936-39”, Sojourn: Journal of Social Issues in Southeast Asia, 207–216.

—, 1990.
Rebellion in Laos: Peasant and Politics in a Colonial Backwater, Boulder, Westview Press.

“Pagodas and prophets: Contesting Sacred space and power among Buddhist Karen in Karen State”, The Journal of Asian Studies 70/4, 1083–1105.

“Time, narrative and ritual: Historical interpretations from an Amazonian society”, in Jonathan D. Hill (ed.), Rethinking History and Myth: Indigenous South American Perspectives on the Past, Urbana & Chicago, University of Illinois Press, 78–105.

“Sourcing novelty: On the ‘secondary invention’ of writing”.

“Religion and anti-colonial movements”, in Nicholas Tarling (ed.), The Cambridge History of Southeast Asia, Cambridge, Cambridge University Press, 197–248.

Description of Jruq (Loven): A Mon-Khmer language of the Lao PDR, Canberra, Australian National University.

“Notes ethnographiques sur diverses tribus du Sud-Est de l’Indochine”, Bulletin de l’École française d’Extrême-Orient 1/4, 291–311.

LEACH E.R., 1964 [1954].
Political Systems of Highland Burma: A Study of Kachin Social Structure, London, The London School of Economics and Political Science.

LEHMAN F.K., 1967.
“Burma Kayah society as a function of the Shan-Burma-Karen context”, in F. K. Lehman, Richard Downs & Toshinao Yoneyama (eds.), Contemporary Change in Traditional Societies, Urbana, Chicago & London, University of Illinois Press, 3–104.

“Les écritures du Hmong”, Bulletin des Amis du Royaume Lao 7-8, 123–165.

Tristes tropiques, Paris, Plon.

Ethnologue: Languages of the World, Dallas, Texas, SIL International.

MARIN G., 1943.
“An old Pwo-Karen alphabet”, Man 43, 17–19.

MCMAHON A.R., 1876.
The Karens of the Golden Chersonese, London, Harrison & Sons.

“The 1901-1902 ‘Holy Man’s’ Rebellion”, Sciences 5, 78–86.

OLLONE H. M. G., 1912.
Écritures des peuples non chinois de la Chine, Paris, Ernest Leroux.

“Preliminary proposal to encode the Pau Cin Hau script in ISO/IEC 10646”, ISO/IEC JTC1/SC2/WG2 N3781: L2/10-080: 2010-02-28.

PURSER W. C. B. & AUNG TUN, 1920.
A Comparative Dictionary of the Pwo-Karen Dialect, Rangoon, American Baptist Mission.

RAO RAHUL, 2010.
Third World Protest: Between Home and the World, Oxford, Oxford University Press.

“The development of Kayah national identity”, in Uta Gärtner & Jenz Lorenz (eds.), Tradition and Modernity in Myanmar: Proceedings of an International Conference Held in Berlin from May 7th to May 9th, 1993, Berlin, Fakultätsinstitut fur Asien- und Afrikawissenschaften der Humboldt-Universität zu Berlin, 101–128.

—, 1994b.
On the Development of Kayah and Kayan National Identity: A Study and Bibliography, Bangkok, Southeast Asian Publishing House.

“Literacy”, accessed 5 April 2018,

“The return of the python god: Multiple interpretations of a millenarian movement in colonial Vietnam”, History and Anthropology 8/1-4, 129–164.

Histoire des Miao, Paris, Société des missions étrangères.

“A brief outline on the traditional background of the Lehkai (Ariya) religious sect [and] its Leit-Hsan Wait”, Karen Heritage 1, 10–25.

“Geographies of knowing, geographies of ignorance. Jumping scale in Southeast Asia”, Environment and Planning D. Society and Space 20/6, 647–668.

Against the Grain: A Deep History of the Earliest States, New Haven & London, Yale University Press.

—, 2009.
The Art of not Being Governed: An Anarchist History of Upland Southeast Asia, New Haven & London, Yale University Press.

SHWAY YOE, 1910 [1882].
The Burman. His Life and Notions, Londres, Macmillan.

“The Khom script of the Kommodam rebellion”, International Journal of the Sociology of Language 192, 15–25.

Mother of Writing. The Origin and Development of a Hmong Messianic Script, Chicago, University of Chicago Press.

“Another Hmong messianic script and its texts”, Written Language and Literacy 1, 103–128.

SOLNIT DAVID B., 1986 [1984].
A Grammatical Sketch of Eastern Kayah (Red Karen), Berkeley, University of California PhD dissertation.

“Language use and policy in a linguistically fragmented refugee community”, mémoire non publié de l’université Macquarie, Sydney.

“Ariya and the Golden Book. A millenarian Buddhist sect among the Karen”, The Journal of Asian Studies 27/2, 297–328.

—, 1968b.
“Three Pwo Karen scripts: A study of alphabet formation”, Anthropological Linguistics 10/1, 1–39.

Sovereignty and Rebellion. The White Hmong of Northern Thailand, Singapore, Oxford University Press.

Literate Networks and the Production of Sgaw and Pwo Karen Writing in Burma, c. 1830-1930, thèse de l’université de Londres.

Haut de page


1 For van Schendel (2002), Zomia extends into the contiguous highlands of South Asia, Central Asia and China, while Scott locates it in mainland Southeast Asia.

2 Several Zomian scripts are not included here. Due to a lack of accessible sources I am not discussing the Hanifi Rohingya script devised in Myanmar in the 1980s by Maulana Mohammed Hanif and the Rohingya Language Committee. I also exclude scripts developed in the diaspora, such as Nyiakeng Puachue Hmong, or those created by outsider missionaries such as the Pollard script created for the Miao languages. For an account of scripts invented in South Asia see Brandt (2014, 2018). I am indebted to Leke activist Sher Nay Moo, also known as Paung Mla Mla, for his expert advice on the Leke script. Thanks also to John M Duffy for access to rare sources on lesser-known Hmong scripts. I am grateful to the late, and much beloved, Nick Tapp for introducing me to the history of Hmong messianism and writing systems.

3 Karen language names have multiple synonyms that are confusing and sometimes contradictory. I have defaulted to the terminology preferred by Ethnologue (Lewis, Simons & Fennig 2015). The overarching label Karen designates a heterogeneous and dispersed population of communities in Myanmar and Thailand. Linguistically, there are 21 Karenic languages in Central, Northern, Peripheral and Southern subgroups. Even within subgroups these languages are not always mutually intelligible.

4 I have reproduced most of this narrative from a paper published by current members of the Leke sect residing in a refugee camp on the Thai-Burma border (Saw et al. [1980] 2006). However, the version provided by Womack differs in several particulars. In Womack’s account there were two Karen men by the name of Pu Nai Thayat and Pu Maung Tawdut. While on Mt Zwegabin the men were visited by multiple nats (Buddhist spirits) who gifted them with the script. Later, two other men would systematise this script into an alphabet. In the early 1860s, the same cohort of spirits returned to reveal the foundational holy book of the Leke sect to Pi Mike Kali and Pu Ti Thuang Tawt.

5 It is possible that this preference rests on a misunderstanding that “chicken scratch” was a term of ridicule used by those outside the Leke movement (see for example Purser &Tun 1920; Stern 1968b). This is contradicted by Womack (2005, 115) but more importantly by contemporary Leke adherents who take pride in the term and relate it to the script’s origin story (Saw et al. [1980] 2006).

6 The ethnic provenance of this tale is not clear. Cross attributes it to “the Sgau Karens and by the eastern Bghais” while Bunker associates it with “the Red Karens [Kayah], or Eastern Bghais”. Ethnonyms have changed so frequently since the nineteenth century that it is perhaps safer to regard it as a “Karen” tale.

7 Jonathan Falla associates the fragment of the brass and gold book with a Sgaw Karen story documented by MacMahon, in which there were seven Karen brothers who each received a portion of a bamboo bucket from their parents. The eventual restoration of these fragments was to coincide with a time when the Karen would live together again in harmony. Falla, however, misattributes the story to the Bwe Karen and mistakes the source as Bunker (1872). I believe it is safe to conclude that the bamboo bucket and the gold and brass plate are unrelated motifs in the Western Kayah folkloric tradition.

8 The best candidate for documented ancestral Hmong script comes from the early 20th century, when H.M.G. Ollone elicited an independent script of some 338 characters apparently in use among a Hmong community in China (Ollone 1912). Although the Hmong scholar Jacques Lemoine later demonstrated that most of the characters were distortions of Chinese, he nonetheless isolated several that were original to the script (Lemoine 1972).

9 See Jiang Yongxing (1989) discussed in Enwall (1994, 65). Note that Enwall is sceptical about the reliability of this source.

Haut de page

Table des illustrations

Titre Fig. 1. Map of Zomia’s writings, 2018
Légende Map of Siam’s, Tunquin’s, Pegu’s, Ava Aracan’s kingdoms…, 1764 ; made for Terrain ; the green area is the Zomia.
Crédits Source : Gallica / Piers Kelly © Bibliothèque nationale de France
Fichier image/jpeg, 554k
Titre Fig. 2. Writing book, c. 1962
Légende This new writing (kayah li) was created to unify and represent all the Karen languages of Myanmar.
Crédits Rights : Htae Bu Phae © All rights reserved
Fichier image/jpeg, 1,4M
Titre Fig. 3. Karen plate, 1872
Légende The inscriptions of this writing, of divine inspiration, are a trace of the Karens’ one which would have been lost.
Crédits Rights : Alonzo Bunker
Fichier image/jpeg, 614k
Haut de page

Pour citer cet article

Référence électronique

Piers Kelly, « The art of not being legible », Terrain [En ligne], 70 | octobre 2018, mis en ligne le 06 novembre 2018, consulté le 16 novembre 2018. URL : ; DOI : 10.4000/terrain.17103

Haut de page


Piers Kelly

Minds and Traditions Research Group, Max Planck Institute for the Science of Human History, Jena

Haut de page
  • Logo CNRS
  • Logo Université Paris Lumière
  • Logo Ministère de la Culture
  • Logo Université Paris Ouest
  • Logo MAE - Maison Archéologie Ethnologie
  • Logo FMSH-Diffusion
  • OpenEdition Journals