Revitalising Naxi dongba as a ‘pictographic’ vernacular script

Abstract

A script can be a window into a language and all the culture contained within it. China’s minority peoples have a multitude of scripts, but many are in danger of falling out of use, a decline spurred by the adoption and promotion of standard Chinese across the country. Nevertheless, efforts are being made to preserve minority writing systems. This article reveals how the primarily logographic Naxi dongba script (often labelled the world’s ‘last living pictographs’), used in China’s southwestern Yunnan province to record the Naxi language, can be practically used as a modern writing system alongside its more widely known traditional role as a means of recording religious rites, and what exactly separates these two styles of writing. The efforts that have been made to achieve the goal of modernisation over the past decades are reviewed, including the longstanding attempts at Unicode encoding. I make some suggestions for the future development of the script, and employ plenty of examples from recent publications, alongside phonetic renderings and English translations. It is hoped that overall awareness of this unique script can be raised, and that it can develop into a vernacular script with everyday applications.

Keywords

Dongba script minority writing Naxi pictographs Unicode

Introduction

There is a need for diversity in written scripts just as much as there is a need for diversity in spoken languages, and yet we only rarely hear about efforts to save scripts.¹ A script can be a window into a language and all the culture contained within it. China’s minority peoples have a multitude of scripts, but many are in danger of falling out of use; a decline spurred by the adoption and promotion of standard Chinese across the country. Standardisation brings with it many benefits, but there must surely be room for people to be literate in multiple languages and scripts, to possess an extensive and flexible repertoire of writing resources, as opposed to a limited and inflexible one.

This journal was founded with the mission to promote an understanding of ‘ideographic writings’ that represent ‘pearls of human wisdom with manifold structures and types’, and that act as ‘rich sources for historical and cultural investigations’ (Zang, 2017: 3). Saving logographic scripts² and encouraging their use in the modern world must surely be in the interests of all of us who study writing systems. It is the purpose of this article to reveal how the primarily logographic Naxi dongba script (sometimes referred to as the world’s last ‘living pictographs’), used in China’s Yunnan province to record the Naxi language, can be practically used as a modern writing system alongside its more widely known traditional role as a means of recording religious rites, and what separates these two styles of writing. The efforts that have been made to achieve this goal of modernisation over the past decades will also be reviewed, and some suggestions for future work (notably Unicode encoding) will also be made. I hope the examples provided herein will be evidence of the feasibility of a larger modern Naxi script reader to be published in the near future, and that overall awareness of this unique script can be raised.

It is true that the Naxi dongba script has largely been preserved in countless Chinese and English publications, in the form of parallel-text translations (see Dongba wenhua yanjiusuo, 1999; Rock, 1939). But this is not really a ‘living’ preservation: the script is not ‘alive’ in the sense that it is used daily for purposes of communication. What we have now is akin to a form of sterile museum preservation, the inevitable conclusion of which is a kind of cultural living death (the graphs are preserved, but nobody actually uses them). Undoubtedly, the dongba script still has high visibility in the Naxi cultural capital of Lijiang. If you visit Lijiang, you will quickly spot the dongba script adorning shop and road signs throughout the town centre. Even Starbucks has a translation in Naxi dongba (see Figure 1). These translations are very rarely read, however, serving instead as aesthetic, cultural tokens. They have little real-life impact and are not an organic part of Naxi life.

Figure 1.

Starbucks shopfront, Lijiang, China, October 2018.

In this picture, the Naxi graphs for ‘Starbucks’ can be seen. They appear above the Chinese and English names (not pictured). Starbucks is translated as gee bbaq kee, . The first character means ‘star’ (gee, depicted as three stars), the second and third graphs being phonetic loans (rebuses – where a word difficult to express in picture-writing is expressed by means of a homonym), the flower (bbaq) and the dog (kee) together approximating the sound of the English ‘bucks’; it is a combination of literal and phonetic translation, and one that is an exact analogue of Starbucks’ Chinese translation Xingbake 星巴克, a literal ‘star’ followed by two phonetic syllables. ‘Coffee’ is rendered with two phonetics (as in Chinese), kafv , represented by the graphs for ‘bitter’ (something bad going into the mouth) and ‘hair’. But these are just a few characters on a shop sign. What of real, syntactic writing?

The Naxi dongba script

The Naxi name for their logographic (a traditional classification, although they are known as pictographs in China because of their presumed heritage as picture-writing) script is ser jel lv jel , literally ‘wood record, stone record’,³ which scholars suggest points to the earliest materials used for recording the characters. It is known as dongbawen 东巴文, literally ‘dongba script’. The Naxi religion and its priests are both known as Dobbaq, Chinese ‘dongba’, which is either derived from the Tibetan ston pa, meaning ‘teacher’ (according to Rock, 1963: 87), or the Tibetan Bonpo, from whom the dongba ritualists likely inherited many ritual manuscript traditions. The dongba religion shares elements of the Tibetan Bon tradition, but also draws elements from Tibetan Buddhism and Taoism. There are several thousand graphs, which can be combined in numerous ways to create noun and verb phrases. Joseph Rock’s (1963) dictionary, for example, lists over 3000 graphs.

The majority of the extant Naxi ritual texts are written in the dongba script, a script that has stuck very close to its ideographic roots – direct representation of the thing itself. The Naxi do have another script however, a syllabary known as ggebbaq, which can be written in dongba like this: , read in Chinese geba 哥巴, and which He Zhiwu (1976) claims means ‘disciple’. There are fewer manuscripts written in geba, but geba characters are often used alongside the dongba script as supplementary phonetics. Scholars have dated the script to at least the 13th century (Mathieu and Ho, 2011), though such efforts are complicated by the lack of dates on the ancient manuscripts.

Modern Chinese scholarship has gone some way to offering further insights into the nature of the script. In the preface to his authoritative dictionary, Naxi xiangxing wenzipu 納西象形文字谱, the respected Naxi scholar Fang Guoyu wrote that ‘the Naxi pictographs are symbols representing language’, and went on to label them as ‘primitive’ characters at the stage of proto picture-writing (see Fang and He, 1981: 56). Fang identified 10 different categories into which the characters can be classified, which include the following: basic pictographs; pictographs that focus on a particular feature; pictographic representations of abstract ideas; compound ideograms; and phono-semantic compounds. Scholars have also managed, largely successfully, to categorise the script in accordance with the six types of character classification first advocated by Xu Shen. He (1976) has shown the difference between graphs that are clearly derived from pictograms, such as bi , sun, and ji , cloud; graphs that are combined ideograms, such as zzo , bridge, a wooden plank over a river; compound graphs that are composed of multiple individual graphs that represent either a phrase or a whole sentence, such as bbiq gv rua na , a mythical flying steed, composed of a horse, a horse’s leg and two wings; phono-semantic compounds such as aiq , cliff, with a chicken’s head peeking out from the bottom corner of the cliff that is used as a phonetic (in Naxi, chicken is pronounced aiq); simple ideograms such as ceiq , ten; polyvalent graphs that have multiple readings and meanings that are connected metaphorically, such as la bbei , axe, that can also be read as shuq, iron; and finally, phonetic loan graphs such as iu , meaning either monkey or ancestor.

In the West however, scholarship on, and therefore general understanding of, the Naxi is understandably spottier. DeFrancis’s brief description of the Naxi script is now badly out of date. He says, ‘An intriguing aspect of the Naxi pictographs that has been widely noted is the fact that they sometimes represent sounds’ (1989: 235). The simple fact is that all Naxi graphs represent sounds, because the script is glottographic writing. DeFrancis quotes Naxi scholar (and ‘leading specialist’ on the script) Fu Maoji in tracing a sequential development of the Naxi scripts as beginning with a stage of pictographic writing called dongba. After dongba comes the syllabic geba writing, created by simplifying dongba or Chinese graphs. The final stage is using the pictographs for their phonetic value, under the influence of the syllabic writing (see DeFrancis, 1989: 236).

This sequence is questionable. Even as DeFrancis claims that the Naxi have made little use of rebus symbols (i.e. the latter stage of development), we know that this is patently false: the rebus principle is at play everywhere in Naxi texts, classic and modern. The graph of the tiger’s head, which begins most traditional manuscripts, is a metonymic rebus for the oral formula, ‘long, long ago’. It serves the phonetic role as the ‘la’ in the phrase ‘ a la me sher ni ’, and as Rock’s examples of early manuscripts have this formulaic opening, it can be thought to be a rebus as old as the ritual manuscripts themselves. One final quote from DeFrancis (1989: 237) shows the general feeling that the script is no longer a going concern: ‘the limited use of the rebus principle by the Naxi seems to have led nowhere’. It is the exact aim of this article to show where in fact the rebus principle has led the Naxi script, because it certainly is somewhere – to its possibility for use in recording the modern vernacular, to electronic typing via entering the phonetic value of a graph (as in Chinese pinyin input), in short, to becoming a viable modern-day logographic script. In short, there is no firm evidence that the rebus was a later invention in dongba script, as all extant manuscripts use it.

Père Desgodins of the Missions-Etrangères de Paris, one of the first Westerners to encounter the Naxi script, began the tradition of denying it the status of ‘real writing’. In an 1882 letter to Terrien de Lacouperie that was published in de Lacouperie’s 1894 book, Desgodins writes:

These hieroglyphics ... are not, properly speaking, a writing, still less the current writing of the tribe. The sorcerers or Tong-bas alone use it when invited by the people to recite these so-called prayers, accompanied with ceremonies and sacrifices, and also to put some spells on somebody, a speciality of their own. They alone know how to read them and understand their meaning; they alone are acquainted with the value of these signs, combined with the numbers of the dice and other implements of divination which they use in their witchcraft. Therefore these hieroglyphics are nothing else than signs more or less symbolical and arbitrary, known to a small number of initiated, who transmit their knowledge to their eldest son and successor in their profession of sorcerers. Such is the exact value of the Mo-so manuscripts; they are not a current and common writing; they are hardly a sacred writing in the limits indicated above. (Terrien de Lacouperie, 1894: 47–48)

Even today, this view holds firm. American anthropologist Erik Mueggler calls dongba textual practice ‘recitational’, before affirming that the writing system is incomplete:

Dongba writing could rarely maintain syntactic or semantic coherence separate from its vocalization ... a verse might be transcribed in many different ways; each dongba created his own combinations of graphs; many single graphs represented long names or even verb phrases. For these reasons, dongba writing was more like a complex of mnemonic tools than a full writing system. (Mueggler, 2011: 98)

In 2010, a similar view was officially espoused by the International Committee for Information Technology Standards (INCITS) in a response to a revised proposal for encoding the Naxi dongba script in the Supplementary Multilingual Plane (SMP):⁴

The USNB [US National Body, i.e., INCITS) does not feel Naxi Dongba should be encoded, as it is closer to ‘pre-writing’ rather than a writing system that was used systematically to represent words in running text. This can be demonstrated in particular by the fact that there is no way to consistently specify names for all of the Naxi Dongba symbols. (INCITS, 2010)

Western commentators, beginning in the 19th century and continuing all the way through to the present decade, have denied the dongba script the status of true writing, as if they were the gatekeepers to the written word. Desgodins remarked that the dongba script was ‘not a current and common writing’, but various usages suggest that it was at least current up until the establishment of the People’s Republic of China, even if it was never common to the lay people. The research of Yu Suisheng 喻遂生 (2008) and others has shown that the script has been in use for hundreds of years as a secular tool for the notation of accounts, land contracts and personal correspondence, amongst other uses. The ephemeral nature of such documents has, however, led to few surviving examples.

All this shows that the dongba script is a field rife with misunderstanding. How do you write in pictures, and can such a script ever be used in modern times? One of the many barriers to the development of a written vernacular in Naxi dongba has been its historical situation: the script was used solely by the religious practitioners, passed down from father to son and not accessible to lay Naxi, and as such (as we have seen) has not been seen as a full writing system. Times have now changed, of course, and the script has been opened up for anybody to learn, but there are very few incentives to do so in the current cultural climate. I believe that past analyses of the script have focused too readily on its traditional usage. What I am suggesting here is that there is in fact a linear form of dongba writing that maintains syntactic and semantic coherence, and we should recognise it as a form distinct from its traditional religious usage. Once this recognition has taken hold, it will be possible to more widely develop modern vernacular usages of the dongba script. I would like to posit a modern form and a traditional form, the distinction being usage (ritual vs everyday vernacular) and typographic style, phonetic representation (number of characters written vs mnemonic), and linearity; that is, written in straight lines left to right rather than a jumbled arrangement. We could call these distinct forms traditional dongba script (in Chinese chuantong dongbawen 传统东巴文) and vernacular dongba script (baihua dongbawen 白话东巴文), categories not dissimilar to hieratic and demotic Egyptian. I will proceed to break down two of the more easily identifiable differences between these forms.

Relationship between written word and spoken word. In its most traditional, ritual form, dongba script is purely mnemonic. That is to say, the graphs on the page are visual cues designed to aid in the recall of entire story episodes, only parts of which are actually written down. In this way, a single graph can, in extreme examples, stand metonymically for several metrical lines. Let us take an example from a manuscript that tells the Naxi origin story, part of the myth that tells how the male and female ancestors of the Naxi tribe came together to beget the Naxi people.

(Recreated from Fang and He, 1981: 497)

Simply looking at the four characters here and reading them one by one, we get: lee, a fir tree; jil-to, a small goat; me, not; and req, (a single blade of) grass. When read by a Dongba, however, the four graphs could potentially represent the following nine speech units :

Naxi: Lee naq zzer keeq hai jilto nee bbeiq deeq chee hai hai

English: Beneath the great fir tree, the little golden mountain goat was bleating

Naxi: ‘Nee bbeiq siuq bbeiq ddeq she zil?’

English: ‘Why do you bleat so?’

Naxi: ‘Nee bbeiq ee me bbeiq,

English: ‘I bleat not out of leisure,

Naxi: Nge jjil rheeq req herq yel

English: When I was only a kid, I was fed green grass.

Naxi: Ddeeq nge req me ddeq

English: Now I am older, there is no grass.

Naxi: La ler req ceil herq

English: Lush green grass and leaves,

Naxi: Sseiq cil sseiq nv hee me ddoq?

English: Where have you gone, no longer to be found?

Naxi: Req cil req lei meiq

English: It is lost and must once again be found

Naxi: Nge bbeiq tei bbeiq muq zil.’

English: And this is why I bleat so.’

I say ‘could’ represent the above, because the spoken text is not fixed. It could be cut down or even further embellished upon, depending on any number of factors. When faced with a situation like this, where one graph, such as grass, can be embellished to tell an entire story episode in itself, we can understand how the early commentators might have suggested that the Naxi dongba script was not a full writing system. If you do not know this story, you could not guess it from just these few graphs alone. This is therefore not a style of writing that is meant to be read and comprehended in the way we would usually understand it, but writing that is meant merely to jog the memory of a story.

Layout of the graphs on the page. Traditional dongba script is read largely from left to right, top to bottom, but sometimes the characters are placed in irregular order. Take for example this extract from the manuscript Salaq-Abaq Ngvl (the passing on of Salaq-Abaq):

(Naxixue ziliao congbian, 2014: 3)

This formulaic segment often found at the beginning of ritual manuscripts is read ‘mee tv ddiuq ku rheeq, bbi tv leiq tv rheeq, gee tv ssaq tv rheeq’ (in the time when the sky appeared and the earth was split, when the sun and moon appeared, when the stars and comets appeared). The numbers represent the rough order in which the graphs are read; but it is even more complex than this, as the verb tv (graph number 2), to appear, is read out five times but only written once, and the adverbial time marker, rheeq, is not written at all. In traditional dongba script this becomes almost a collage of graphs that come together to form a cosmic scene, with the sky and the earth and the sun and moon and the stars in-between.

The non-fixed nature of the traditional script means that a similar extract could be written with a different number of graphs, and still be read out in the same way (or indeed embellished upon). For example, the 7 graphs above could become 11 graphs as below:

(Lijiang shi guchengqu wenhuaguan, 2017: 3)

The same elements are present, but the verb ‘to appear’ (a bucket, used as a rebus) is repeated several times.

Aside from its primary function of recording the dongba religious scriptures, the Naxi script was also used to write such things as medicinal prescriptions, accounts, contracts, notes and letters – what can be labelled secular documents. These secular usages employed the vernacular style, that is, linear, left to right writing with no variance in height. The following examples are from the cover of a land contract written in Naxi dongba. Photos of the original manuscript were published by Naxi scholar Yu Suisheng (2008: 157–159), and it was the first Naxi-language land contract to ever be published. It was donated to Yu in 2003 by an agricultural family from the Baidi region to the north of Lijiang, and was alleged to be already at least six generations old at the time it was donated; dating the document to the early 20th century at the latest. The contract is written on both sides of a single sheet of traditional dongba paper some 26.5 cm long and 20.5 cm wide. One side has the title and a blessing, and can be understood to be the cover. The title of the contract on the cover reads as follows:

Naxi: laqsheq leel leel qi oq mei

English: For the sale of land in the area of Lashi.

What we have here is a linear sentence, with grammatical particles and every syllable recorded with a graph. The Naxi place name, Laqsheq, is depicted by a hand (a phonetic loan) and a geba phonetic graph (this is the village of Lashi 拉市, just west of Lijiang centre). There is a compound Naxi graph for the place, Lashi, and it is a tiger (la) with lean meat (shee) above land (ddiuq): . The two insects (lice) act as phonetic loans to stand for ‘place’, the first attached to Laqsheq (the area of Lashi), the second indicating the land to be sold. The thorn, read ‘qi’, has the same pronunciation as the Naxi verb, ‘to sell’, and we end with the copula ‘oq’ and a modal particle used as end emphasis. The cover page also contains a familiar blessing in Naxi that reads as below:

Naxi: ssee sher ha’iq hoq mei

English: Long life of plenty.

Here we have grass, which is a rebus for ‘life’, followed by seven, used as a rebus for ‘long’, a compound graph meaning ‘have rice’ and two grammatical particles. It has been suggested that modern vernacular uses of the Naxi dongba script should be in this linear style: ‘the primary concern of the modern user community is to use Naxi dongba in a linear, practical fashion to represent the language’ (Unicode, 2017: 1). But this linearity is not a strict marker of modernity: even in historical sources we can see the divide between vernacular and traditional writing. This suggests that a divide between modern and traditional forms of the script along these lines is not an accurate classification, and instead of ‘modern’ we would be best served to say ‘vernacular’ for this particular linear, word for word form of the script.

New, published examples of everyday usage

But to suggest that Naxi dongba could be used as a modern script means we should look to more recent examples of its usage. The past decade has seen a number of publications that purport to help the reader to learn Naxi and the dongba script, and I will quote from several of them here. Let us start with an example from the textbook, Gen WoXueshuo Naixihua (‘Learn Naxi with me’) that features a number of sentences in the dongba script, followed by Chinese translations. I have selected a short dialogue (‘Where are you going today?’) from the text as an example:

(Wu and Hu, 2014: 148–149)

Naxi: Cheeni sseiqggai heebbee naiqvf

English: Where do you want to go today?

Naxi: Cheeni ngvlv bbeijjuq bbee

English: Today [I’m] going to Ngv lv snow mountain.⁵

Naxi: Ngvlv bbeijjuq jjaif al qil lei

English: Will Ngv lv snow mountain be very cold?

Naxi: Meeceel dal qil meq, ayi jjaif me qil

English: Only in winter, it’s not very cold now.

These are simple sentences that can easily be written down in the dongba script, and show how it could be used in everyday situations, or on instant messaging applications. The vernacular script can also be used to depict loanwords, such as ‘world heritage’ in the following phrase (presumably coined in Naxi for the publication in which it appears):

(He, 2007: 242)

Naxi: Sseehuaq shuashua, gguqbbei rhee lvq, Dobbaq tei’ee; Sheeljial-yiqcai see siuq zzeeq mei Yiggvdiuq

English: The Three Parallel Rivers, Lijiang Old Town and Dongba Manuscripts: [the place where] three world heritages come together is Lijiang.

The phrase ‘world heritage’ is written as Sheeljial-yiqcai in Naxi, a direct borrowing from the Chinese shijie 世界 (world) yichan 遗产 (heritage). The original text says that jail is a phonetic loan taken from the Naxi jai jue ‘three forked’, used to stand for the Chinese ‘jie’, world. Note here that Fang Guoyu’s dictionary lists this graph as a geba character, pronounced ee. There is still much discrepancy in the various usages of geba characters. Also of note is the fact that here the word ‘dobbaq’, dongba, is written with the traditionally di-syllabic graph, dobbaq , followed by a flower, bbaq: . This example therefore consciously uses one graph per syllable, the second syllable in the dobbaq graph is not read, as this syllable is provided by the flower graph. This approach turns the Naxi script into a kind of graphically complex syllabary, and moves it away from a logographic script (for in a logographic script, by definition, each graph should represent a word).

The dongba script is as flexible as Chinese in that even foreign names can be transcribed with logograms (although as I conclude later in this paper, I believe the geba syllabary may be more suitable for this purpose). What follows is another example of a dialogue written out in the dongba script in the original. From the booklet ‘Dongba Dialog’ published in Lijiang that aims to teach tourists some dongba script, this extract, titled ‘Borrowing and lending things’ specifically invokes Joseph Rock, the great Naxiologist:⁶

(Dongba zhifang, n.d.: 10)

Naxi: Tei’ee chee cai ngvl leiguq

English: I’m returning this book to you.

Naxi: Jjaiq liuq chul kee mai

English: You’ve read it so quickly!

Naxi: Nge lei Loke tei’ee berl gge ddee cai ni bbee

English: I also want to borrow a book about Joseph Rock.

Naxi: Cheegvq ddee cai jjuq ye

English: There’s one right here.

The dialogue is interesting because it shows fully syntactic constructions in Naxi, all using dongba graphs. Once more, we have a single geba graph mixed in with the dongba graphs: ‘lei’ (again). Of particular interest is the transcription of Joseph Rock’s name into Loke, seemingly via the Chinese transcription of his surname, Luoke 洛克, making this a secondary transcription. This is also an example where the logographic for book, ‘tei’ee’, is written with one single compound graph (a flag atop a book, the flag being a phonetic component indicating the pronunciation, ‘tei’).

Electronic input – a new horizon?

There have already been several attempts to modernise the script and make it accessible to modern computer uses, the first major one being the creation of an input method editor with an accompanying true type font that was developed by Chinese company Zmnsoft in the early 2000s. This was based upon graphs found in Fang Guoyu’s dictionary. Following Fang, the input editor contains 1562 dongba graphs and some 661 geba phonetic graphs. The majority of Naxi graphs used in this article were inputted using this software.⁷

The second attempt was the creation of a stylised, coloured script. An inventory of 176 graphs was created by Chen Nan of the Huadong daxue meishu xueyuan, and published as an open-access design tool in Japan by Gijutsu-Hyohron in 2004. These graphs were combined into fully syntactic phrases in vernacular Naxi. For example:

(Azuma tomoe moji aikō kai, 2004: 45)

Naxi: Meeruq gge kv chee sei / Jjiqna heegee kv

English: On summer evenings, there will be dark clouds and rain.

Note here the usage of one geba phonetic, sei , a sentence end particle, at the end of the first line. It is also often written with the dongba graph for the mountain goat, sei , but the geba phonetic is much simpler and easier to write.

These graphs are somewhat over-stylised (and perhaps over-coloured; the colouring of graphs was never done to such extremes, and is seen by dongba practitioners from traditional northern schools such as Baidi as a Lijiang affectation), with questionable inclusions such as the cat’s face. The cat’s face stands out as a ‘new’ graph as in terms of graphic design; it does not match the other animal graphs of the dongba script. While it is true that the heads of animals are used metonymically to represent the whole creature, the heads are always written in profile, never face on (e.g. Naxi kee , dog). The traditional dictionaries do not provide a graph for ‘cat’. Nevertheless, this limited inventory does make it possible to create simple sentences in vernacular dongba, albeit by painstakingly combining individual images. The limited inventory of only 176 graphs does, however, mean that their use can only ever be limited to simple sentences.

Unicode encoding is the last great barrier to practical electronic representation of the Naxi script in the modern setting, but a barrier that is about to be scaled, with a proposal for inclusion into the Unicode standard made by Chinese scholars that has now been recommended as a stable basis for a repertoire of graphs that can be balloted. In fact, a proposal to encode Naxi dongba script in the Unicode standard has been in evidence for around a decade, but it has been mired in a slow process of many revisions. In September 2017 however, the revised proposal was finally accepted for balloting.

A total of 1188 graphs have been identified by Naxi experts from the Lijiang Dongba Culture Research Institute, and initially ratified by a Unicode panel. It must be stressed that the motivation for Unicode encoding is to be able to portray vernacular dongba script, and official documents have said as much: ‘the encoding is intended to address its modern, revivalist use, and not the traditional use’ (Anderson, 2014: 1). These 1188 graphs will not be enough to represent traditional ritual manuscripts. If (or more hopefully, when) these 1188 graphs are added to the Unicode standard, it will be possible for all to read and write Naxi electronically. The benefits to Unicode encoding are enormous. Any text written in Naxi pinyin could be easily converted into Naxi dongba. Take for example this piece, written in Naxi pinyin for the inaugural edition of the Lijiang Naxi-language newspaper (Liljai Bal) in 1985: ‘Na Hei tei’ee bei soq vf / Tei’ee soq sei see see nei gv ta mei’ (see He, 1985; meaning ‘Naxi and Han culture should both be studied / Only by reading books, can one become knowledgeable and wise’) can be represented in Naxi as follows:

So far in the above examples we have encountered two ways of writing tei’ee (book) in Naxi dongba: first , the flag and the svastika, and second a flag (tei) as a phonetic over a book, but there is a third way that will simplify the issue. We can simply use the single graph for book as a polysyllabic graph, : tei’ee. Rather than the first example above, the phonetic rendering of flag plus swastika, this graph has the visual connection to the old Naxi manuscripts, the long, rectangular pages divided into three horizontal lines, each broken down into sections by vertical lines (that become diagonal in the graph). It is also simpler than the second, phono-semantic graph.

There should be some room for flexibility in phonetic representation: where possible, it is not necessary to make the script monosyllabic. I believe that for the modern script to be a worthwhile venture it must retain some of its logographic idiosyncrasy; the single graphs representing compound names and verb phrases, for example, can be kept. If we are simply phonetically transcribing each syllable then not only do we lose the logographic nature of the script (it devolves/evolves into a syllabary), but much of the cultural content may be lost. One single graph can encode so much information. Naxi scholars have, for example, written at length about the cultural content of the name of the Naxi culture hero and pre-flood ancestor Coqsseileel’ee, information carried in the graph itself (see Poupard and Li, 2016).

As another example, one traditional name for the Jade Dragon Snow Mountain, Bber shee jji nv lv, is depicted in a single compound graph: . This means the silver-rock mountain covered by clouds above the village of Baisha. Bber shee jji nv lv could also be written syllable by syllable, using rebuses for the majority of the syllables: (as seen in Guo and Yang, 2006: 25). Here we have foot (bber), lean meat (shee), cloud (jji, a logogram, not a rebus), silver (ngv) and stone (lv). And written out in this way, all the visual sense of the mountain capped by clouds is lost.

Somewhat worryingly, the graph for ngv lv , snow mountain, used almost exclusively to refer to Lijiang’s Jade Dragon snow mountain, is not present in the current Unicode proposal, suggesting that the word would have to be written with the two phonetic elements, ‘silver’ and ‘stone’: . This is the drawback to a cut-down character set: you may be able to recreate words phonetically via combining syllables, but the original logographs are in danger of being lost. This brings the script further away from the traditional form, making someone literate in vernacular dongba script unable to access the wealth of cultural knowledge found in the traditional dongba texts. In addition, the currently proposed Unicode encoding has no single graph for the word dobbaq (dongba). A similar graph of a dongba reciting is present under the pronunciation biuq. Dobbaq could still be encoded with two phonetic loans: , a board and a flower. In his dictionary, Rock lists two ways of writing ‘dongba’. The first is the more pictographic graph , read dobbaq, representing a dongba ritualist wearing the five-lobed crown (only three are visible, as it is a front-on view). Rock does also give a second method, using two rebuses, the first being the wooden board, the second being a man suffering from a goitre, also pronounced bba . Rock says that the more common method was the literal depiction of a seated dongba: ‘phonetics ... have been used to write the term, usually that for the figure of a Dto-mba is used’ (Rock, 1963: 87). The problem comes when, as in the example used earlier, we have the seated dongba appended with a phonetic, such as the flower, as this means there is a danger the two graphs could be read as ‘dobbaq bbaq’ by those familiar with the di-syllabic reading of the first graph. We should ensure that the script remains a logography and does not become a syllabary, for a syllabary would be of less interest in terms of long-term cultural impact.

There are also some further points of consideration, the first being the graphic inventory. We need an agreed-upon number of base graphs that can be included in dictionaries, and with the Unicode proposal this is something that is being worked towards, and that can be built upon. Anything that can be said in Naxi should be able to be written in vernacular dongba script. Potential should also be left for the creation of new graphs. The current inventory of characters does allow for the creation of new compound ‘huiyi 会意’ graphs, such as the Lijiang bus company’s new graph for bus, which combines a seated man, woman and wooden vehicle to convey the idea of public (see Figure 2).

Figure 2.

Detail from the back of a Lijiang municipal bus, July 2018.

While individually the three graphs could be read as sso mi ce (man, woman, vehicle), this new compound graph would be read as ‘gu-ja-ce’, as a phonetic loan from Chinese gongjiaoche 公交车, bus, according to the Hail-Naqxi siciq sufyujif (Chinese-Naxi collection of new words and terminology) published in 2015 (Zhao, 2015). Turning the script into a syllabary would remove the opportunity for such graphic invention, and reduce the possibilities of logographic translation. Many of the examples provided herein are not punctuated, and it is also possible to separate sentences with vertical lines as seen in the traditional manuscripts, but I believe it would be straightforward and beneficial to adopt modern Chinese punctuation for Naxi dongba script (punctuation that has been adopted already for Naxi pinyin romanisation).

The geba syllabary?

As I have previously stated, there is a second script in use in the Naxi areas. In several of the examples here we have seen geba phonetics mixed with dongba graphs in modern usage: in proper names, or as grammatical particles that mark tense. They could also function as graphs for syllables that have no established traditional dongba graph; or in transcriptions of foreign names, much like Japanese katakana, as the proposed Unicode graph inventory is already so limited. This would make it easier to see which words are transcriptions of foreign names. Currently, geba is not being considered for Unicode inclusion alongside Naxi dongba: ‘The ad-hoc recommends that Naxi Geba should be left for further study as a separate script as it is not relevant to the modern use of Naxi Dongba’ (Unicode, 2017: 2).

Geba is listed on the Supplementary Multilingual Plane roadmap, but a detailed proposal for its inclusion has not yet been written. As I have attempted to show above, the Naxi geba syllabary is in fact very relevant to modern usages of Naxi dongba; they are frequently employed for grammatical, semantically empty markers, and cannot be said to be separate for this very reason. In fact, certain dongba graphs have geba components, such as the graph Sei bi rai ngv lv (Wenbi Mountain), which contains the geba sei within the mountain as a phonetic. Therefore to claim the geba phonetics are not relevant to modern Naxi dongba is ill-informed (or only irrelevant if such compound graphs were considered unfit for inclusion in the Unicode standard). Furthermore, I believe there may be certain lexical areas where the geba syllabary could prove very useful. Take, for example, the Naxi for ‘Book of Mark’ (i.e. the gospel of Mark) from the Elise Scharten translation (1932, British and Foreign Bible Society). This translation was made using a specially created Naxi script heavily inspired by the Fraser Lisu script. In this script, the title reads: ; in Naxi pinyin: Ma ko tei’ee. Literally, this means ‘Mark’s book’. A rendition in dongba script would read:

The first two graphs are phonetic loans for the syllables Ma and ko, the final graph being di-syllabic, tei’ee (book). Of course, the foreign name poses a problem. If we were to take the graphs at face value (as is always the danger with graphs derived from pictures), literally this might mean smear-horn book, from the image of a person smearing butter on something, the image of a head with two horns, and of course the book. The only way we would have any idea a name is involved is if we were already familiar with the phonetic translation of the English name ‘Mark’. The first two syllables could, however, be written in the geba script, giving us the title as follows: . Using geba phonetics such as the above it would be more immediately obvious that this is a loan word. Of course, this issue mainly becomes relevant in the creation of modern texts or the translation of foreign works into dongba script. I believe that a simple inventory of geba graphs should be adopted for the Naxi syllabic inventory (this would be somewhere in the region of 250 geba graphs), as phonetic loans allow for the easy coining of names and foreign terms.

Conclusion

The examples I have provided of vernacular dongba script thus far have come from textbooks, historical sources and shop signs. There is, however, a legitimate question that could be raised: How many people are actually using the script now? The answer, as one might imagine, is very few. But there is the potential for the script to develop, and there are nevertheless still a small number of people actively using it. Figure 3 below is an image taken from a notebook belonging to a native Naxi student in Lijiang, and represents vernacular dongba script, recording the lyrics to a popular Naxi song, Dong zhi lian 冬之恋 (Winter’s love).

Figure 3.

Naxi song lyrics in dongba script from a Naxi student’s notebook, July 2018.

Interestingly, the punctuation here is traditional, with speech units marked by vertical lines instead of full stops or other marks.

Yang Zhengwen, a Naxi writer from Shuijia village, Sanba township, writes stories in modern dongba (see previous image). In the interests of space I have not provided a phonetic breakdown of this text, but a somewhat edited-down translation would be as follows:

Ddobaq sherl-ler’s parents were both descendants of the gods. When his mother had been pregnant with him for nine months and thirteen days, he shouted out from the womb:

‘Mum! Which path do you want me to take to get out of here?’

His mother said, ‘Son, go via the path that humans take!’

To which Ddobaq sherl-ler replied, ‘That path is too dirty, I don’t want to go that way!’

Having said that, he somehow managed to pop out from his mother’s armpit. Within moments, he was able to speak. ‘Mum, dad, you’ve had a tough time of it!’

Examples such as these reveal that the script is a going cultural and intellectual concern, and that it can be used for literary creation as well as convenience. It is by no means relegated to historic, ritual use as it was in the past, or indeed limited to aesthetic, promotional use in the present day. The linguist Alexis Michaud has called for scholars to develop Naxi studies by making databases of Naxi texts and documents. Upon the publication of Thomas Pinson’s (2012) Naxi-English-Chinese trilingual dictionary, the work was hailed as a major step forward for research into the Naxi language. ‘The “three treasures”... that linguists can contribute on the basis of in-depth fieldwork are (i) a dictionary, (ii) a reference grammar, and (iii) a collection of glossed texts. The first treasure is now available’ (Michaud, 2013: 135).

We should take up this call and build on the work of Yu Suisheng and other modern scholars of Naxi to create a corpus of texts in vernacular Naxi script. Of course, as Michaud (2013: 136) says, we are working with limited resources:

Texts in Naxi remain relatively few in number and narrow in terms of genre, consisting mostly of glossed editions of Naxi ritual texts. There is clearly room for new editorial projects, for instance creating parallel-text editions of the small set of monolingual books in romanised Naxi (transcribed ballads, songs and tales) published in the 1980s (see also Yang 1988), or recording texts anew and making the recordings available together with glossed transcriptions and translations.

There are very few texts available in romanised Naxi, let alone in the dongba script. We must first stress the possibility of transferring any romanised Naxi into the logographic script (such as the newspaper Liljai Bal, quoted earlier); literacy in the dongba script should be promoted alongside literacy in Naxi pinyin.

Nevertheless, I hope this article has gone some way to showing first and foremost the very real usages of Naxi dongba as a way of recording the modern Naxi vernacular, and the diversity of text types that use it. Moreover, the future certainly holds the potential for an exponential increase in its usage. I propose a reader in vernacular dongba script, with annotated short texts from a variety of modern sources. The examples used here could be expanded with word-for-word readings and vocabulary lists. Beyond this, we do not simply need a modern reader in the Naxi script, but a dictionary too. I believe that Pinson’s (2012) dictionary represents an excellent lexicographic framework from which to build upon, and dongba graphs could be provided at both the headword and example levels. The difficulty would be in decoding these graphs: in Chinese, we can look up unfamiliar graphs via the usage of radicals, and perhaps it would be of benefit to identify a radical framework for any finalised inventory of dongba graphs accepted by Unicode.

If we work towards making the dongba script more readily accessible, by helping students to acquire literacy in it, and of course by implementing the Unicode standard, then there is hope that the script can be further developed and will take on new life. I envisage a future where young Naxi, and indeed any who speak the language, can send each other messages and make social media posts in Naxi dongba (the more pictographic and ideographic dongba graphs are themselves possible stand-ins for emoji), and those with the inclination can write new literary works in the language, via the medium of a revitalised vernacular dongba script.

Declaration of conflicting interests

None declared.

Funding

This research received no specific grant from any funding agency in the public, commercial or not-for-profit sectors.

Footnotes

1. The British Library’s Endangered Archives programme ( has aided in the preservation of Yao and Yi manuscripts in recent years, but this only indirectly helps with the preservation of traditional scripts.

2. Although it is not strictly true to suggest that all Chinese or Naxi graphs represent distinct words (they are not fully logographic systems), I purposefully move away from the quoted use of first ‘pictographic’ (in the title) and second ‘ideographic writings’, as the graphs in Chinese and Naxi dongba represent language, not pictures or ideas.

3. Here the abstract verb is depicted via a rebus: although we have two graphs derived from pictographs, a tree and a lump of stones, the verb jel, to record, is portrayed with the graph for bracelet, jjiuq, as a rebus.

4. The SMP is used primarily for historical scripts, and as such I would argue that a better classification might be the Unicode Basic Multilingual Plane, as these graphs are still in use to record a living language – Naxi.

5. Literally, ‘silver-rock snow mountain’, known in English as the Jade Dragon snow mountain, from the Chinese Yulong xueshan 玉龙雪山.

6. Joseph Rock was the first man to take Naxiology as a serious discipline, living intermittently in the Lijiang region for over two decades before the Communist revolution, and producing voluminous monographs and articles on the subject. His greatest works are a two-volume ethnographic study of the Naxi people, The Ancient Na-khi Kingdom of Southwest China, and a two-volume dictionary, A Na-khi -English Encyclopaedic Dictionary.

7. The software is still available to download and purchase at the time of writing (August 2018): .

References

Anderson

(2014) Feedback on Naxi Dongba Supplement document (N4633). Report. Script Encoding Initiative.

Berkeley, CA

University of California, Berkeley, September

Azuma tomoe moji aikō kai 東巴文字愛好会 ATMAK (2004) Pasokon de tanoshimu ai to yūjōno tonpa moji パソコンで楽しむ愛と友情のトンパ文字 [Enjoy the Love and Friendliness of Dongba Writing on Your PC].

Tokyo:

Gijutsu-Hyohron.

DeFrancis

(1989) Visible Speech: The Diverse Oneness of Writing Systems.

Honolulu:

University of Hawaii Press.

Dongba wenhua yanjiusuo (东巴文化研究所) DBWHYJS (1999) Naxi dongba guji yizhu quanji 纳西东巴古籍译注全集 [Annotated collection of Naxi Dongba manuscripts]. 100 vols. Kunming: Yunnan renmin chubanshe 昆明: 云南人民出版社.

Kunming:

Yunnan People�s Publishing House.

Dongba zhifang (东巴纸坊) DBZF (n.d.) Dongba dialog

Fang

( 方

国瑜

) He

( 和

志武

) (1981) Naxi xiangxing wenzipu 纳西象形文字谱 [A Dictionary of Naxi Pictographic Characters].

Kunming:

Yunnan renmin chubanshe 昆明：云南人民出版社. Yunnan People’s Publishing House

Guo

( 郭大

烈

) Yang

( 杨一

红

) (2006) Naxi xiangxingzi dongbawen yingyong 纳西象形字东巴文应用 [The Application of Naxi Pictographic Dongba Script].

Kunming:

Yunnan daxue chubanshe 昆明: 云南大学出版社 Yunnan University Press.

( 和宝

林

) (2007) Changyong dongba wenzi mingyan suyu xinshang 常用东巴文字名言俗语欣赏 [Appreciation of Popular Dongba Sayings].

Kunming:

Yunnan minzu chubanshe 昆明云南民族出版社 Yunnan Nationalities Publishing House.

( 和万

宝

) (1985) Na Hei tei’ee bei soq vf [Naxi and Han Writing Should Both Be Studied]. Liljai Bal December 26, 1.

10.

( 和志

武

) (1976) Naxizu de xiangxing wenzi he dongbajing 纳西族的象形文字和东巴经. Kunming:

Yunnan daxue lishi yanjiusuo 昆明: 云南大学历史研究所 Yunnan University Institute of Historical Research.

11.

International Committee for Information Technology Standards (INCITS) (2010) Comments on ‘Revised proposal for encoding Naxi Dongba pictograph in the SMP’. Available at: http://unicode.org/L2/L2010/10462-naxi-dongba.pdf (accessed September 2018)

12.

Lijiang shi guchengqu wenhuaguan (丽江市古城区文化馆) LJSGCQWHG (2017) Heibai zhanzheng: Lianhuanhua 黑白战争：连环画 [The War Between Black and White: A Comic].

Kunming:

Yunnan renmin chubanshe 昆明：云南人民出版社. Yunnan People’s Publishing House

13.

Mathieu

] Ho

(2011) Quentin Roosevelt’s China: Ancestral Realms of the Naxi.

Stuttgart:

Arnoldsche Art Publishers.

14.

Michaud

(2013) Review of: Pinson Thomas M. 2012. A Naxi-Chinese-English Dictionary (NaqxiHabaq-Yiyu Ceeqdiail). Linguistics of the Tibeto-Burman Area 36(2) 129–137

15.

Mueggler

(2011) The Paper Road: Archive and Experience in the Botanical Exploration of West China and Tibet.

Berkeley, CA:

University of California Press.

16.

Naxixue ziliao congbian (纳西学资料丛编) NXXZLCB (2014) Naxizu Dongbajing diancang jingxuan 纳西族东巴经典藏精选 [Selected Naxi Dongba Classics]vol. 2.

Kunming

Yunnan minzu chubanshe 昆明：云南民族出版社 Yunnan Nationalities Publishing Hous

17.

Pinson

(2012) A Naxi-Chinese-English Dictionary (Naqxi-Habaq-Yiyu Ceeqdiail).

Kunming:

Yunnan minzu chubanshe 昆明: 云南民族出版社 Yunnan Nationalities Publishing House.

18.

Poupard

(2016) Translating a Naxi hero’s name: From ethnic to world literature. Journal of Ethnology 7(4) 26–33

19.

Rock

(1939) The romance of K’a-mä-gyu-mi-gkyi. Bulletin de l’Ecole française d’Extrême-Orient 39(1) 1–152

20.

Rock

(1963) A Na-Khi – English Encyclopedic Dictionary.

Rome:

Instituto italiano per il Medio ed Estremo Oriente.

21.

Terrien de Lacouperie A (1894) Beginnings of Writing in Central and Eastern Asia or Notes on 450 Embryo-Writings and Scripts.

London:

Nutt.

22.

Unicode (2017) Results of the ad-hoc meeting on Naxi Dongba in Hohhot. Available at: https://www.unicode.org/L2/L2017/17337-n4895-naxi-dongba-adhoc.pdf (accessed September 2018).

23.

( 吴

芳

) Hu

( 胡凤

丽

) (2014) Gen wo xueshuo Naxihua 跟我学说纳西话 [Study Naxi with Me].

Kunming:

Yunnan chuban jituan 昆明: 云南出版集团 Yunnan Publishing Group.

24.

( 喻遂

生

) (2008) Naxi dongbawen yanjiu conggao vol. 2 纳西东巴文研究丛稿第二辑 [Collected Naxi Dongba Research (Vol. 2)].

Chengdu

Bashu shushe 成都：巴蜀书社 Bashu Publishing House

25.

Zang

(2017) Inaugural editorial: Promoting the understanding of ideographic writings. Journal of Chinese Writing Systems 1(1) 3

26.

Zhao

( 赵庆

莲

) (2015) Hail-Naqxi siciq sufyujif [Chinese-Naxi collection of new words and terminology].

Kunming:

Yunnan minzu chubanshe 昆明: 云南民族出版社 Yunnan Nationalities Publishing House.