Home > China Facts > China Overview > Chinese Language > Evolution Of The Chinese Script

Evolution Of The Chinese Script

Last updated by fabiowzgogo at 2015/5/21

The Chinese script has evolved dramatically over the years, as indicated in the introduction, from a simple, ideo- and pictographic script to a highly stylized script that eventually underwent a change from the perhaps overly complex – albeit very artistic – to the highly simplified before making the shift to a Latin-based (Romanized) script that would not only facilitate the Chinese government's efforts to eradicate illiteracy, but which would make the Chinese language much more accessible to a foreign audience, since the Romanized script made it possible to indicate pronunciation with the help of diacritical markings familiar to several Western languages.

Chinese Script Evolution Diagram

The table diagram below, replete with supplementary information, depicts the progression of the Chinese script from its earliest, ideo- and pictographic forms to its final form prior to the leap to a Romanized script, bypassing the somewhat dead-end attempt to create a simplified phonetic script that did not rely on Romanization, namely, the Zhuyin Fuhao script (aka Bopomofo) that never quite caught on (see the section on indigenous, Latin-inspired – but not Romanized - transliteration systems below).


Comments to the above Chinese script evolution diagram:

The Oracle-Bone script – the earliest known Chinese script and which was highly pictographic in nature, is dated to the Shang (BCE 1700-1027) Dynasty period.

The Bronze script – is dated to the late Shang Dynasty/ early Western Zhou (BCE 1027-771) Dynasty period. Its creation stems from the fact that a more graphic type of script was needed for inscriptions on metal objects (tools, weapons and utensils), which objects were principally of bronze during the period in question, hence the name subsequently given to the script.

The Large Seal script – is dated to the Eastern Zhou Dynasty (BCE 770-221), which is subdivided into the Spring and Autumn Period (BCE 770-476)and the Warring States Period (BCE 475-221) . It was also used primarily on bronze objects, and, seen in retrospect, represents an intermediate stage beteen the older script styles of the preceding era and the newer, "cleaner" scripts that would be introduced with the Qin Dynasty (BCE 221-207) , China's first Imperial dynasty.

The Small Seal script – stems from the Qin (BCE 221-207) Dynasty. Just as Qin Shi Huang unified China on the heels of the troubled Warring States Period of the Eastern Zhou Dynasty, declaring himself the emperor, Qin Shi Huang also undertook to standardize the Chinese script, and thus was introduced the Small Seal Script (in fact, the number of achievements of this short-lived dynasty is rather amazing, but as any psychologist will tell you, zealousness (and Qin Shi Huang was apparently a very zealous man... perhaps a micro-manager type in today's parleur) often leads to failure, or, put slightly differently, one's greatest quality oftentimes proves to be one's Achilles heel).

The Clerical script – owes its roots to the Small Seal script, according to historians. That said, it bears little resemblance to that script, at least to the layman. In fact, one can safely say that the Clerical script marks a turning point in the evolution of the Chinese script away from the pictographic and toward the more stylistic.

The Clerical script made its appearance during the 1st century BCE, or during the Western Han Dynasty (BCE 206 – CE 009), but came into prominence first during the Eastern Han Dynasty (CE 25-220). It has also been nicknamed the "Breaking Wave" script due to the bold, sweeping curvature of its downward-sloping strokes (seen in the above diagram in the characters for "horse" and "to see"). Note also that the Simplified script would revive this "breaking wave" feature almost to the point of exact replication.

The Grass script – aka Cursive script, was, in all likelihood, conceived sometime in the 3rd century CE, i.e., during the Eastern Han (CE 25-220) Dynasty. There is some murkiness over the exact period of the emergence of the Cursive script, just as there is murkiness surrounding the exact period of the emergence of the Semi-Cursive script (see below), with some sources claiming that the latter script preceded the former, though most sources – and these, in contrast to the more dubious sources, cite references – claim that the Cursive script arrived first.

The confusion is understandable, for with the cursive scripts (cǎo​shū​), we have for the first time in Chinese history a script type that served a dual purpose: the usual means of writing and, additionally, a means of artistic embellishment.

Sifting through all of the evidence that I could dig up without too much trouble, I have come to the conclusion – based mainly on the historical record, but also partly on common sense and intuition – that the first cursive script to see the light of day (zhāng cǎo) served the usual function as a medium of writing, and that it was not very fanciful, being instead a hurried, or shorthand, way of writing the main script of the day, the Clerical script, but that on seeing this more flowing, more comely script, others – being more aesthetically than practically oriented – saw in it a vehicle for artistic expression.

Thus the later, more fanciful cursive script that would emerge (jīn cǎo) possibly existed parallel, at least for a time (this less fanciful cursive script would be superceded by a more deliberate, in-between script whose function was primarily if not solely that of writing, namely, the Semi-Cursive script described further below), with the less fanciful version.

It would seem, then, that the less fanciful script, the zhāng cǎo script, emerged during the latter part (around the turn of the 3rd century CE) of the Eastern Han Dynasty, while the more fanciful variant, the jīn cǎo script, came into prominence during the Kingdom of Wei (CE 220-265, aka Cao Wei Dynasty) Period, one of the famous three parallel kingdoms of the Three Kingdoms Period (CE 220-280) that followed on the heels of the Eastern Han Dynasty.

Note that the Chinese language non-cursive versus cursive comparison to printing versus handwriting in Western languages is entirely valid as long as we are speaking of ancient China, when each word consisted of a single syllable-character. The building blocks of, say, the English language are the letters of the alphabet, while the building blocks of the ancient Chinese language are the strokes of the alphabet, as it were (see the diagram above depicting the 12 standardized strokes). Since the strokes of the Cursive script flow together just as do the letters of the English word, the comparison is quite apt as regards a single Chinese syllable-character, though the comparison breaks down as regards polysyllabic Chinese words, since each character-syllable stands alone, is not connected.

The Regular script – aka Standard script, since it became the standard that is still used today to represent the traditional Chinese character, made its appearance either toward the end of the Eastern Han (CE 25-220) Dynasty, or during the subsequent Three Kingdoms (CE 220-280) Period.† I offer both these possibilities because there is much confusion on the subject of the chronology of the ancient Chinese scripts. The Regular script is considered the most readable (i.e., legibile) style of traditional Chinese script hitherto produced, and is the traditional Chinese script that is adapted to – i.e., made compatible with – all commonly used modern Chinese fonts, including the font used to write this article.

† Note however that many sources suggest that the Regular script first appeared later, during the Kingdom of Wei (CE 220-265) Period, while at the same time they indicate that the Semi-Cursive script (see below), which they indicate originated during the Eastern Han (CE 25-220) Dynasty, was created as a compromise between the cumbersome (slow to write) Regular script and the near-illegibile Cursive script, whose first variant appeared also during the Eastern Han Dynasty, yet the chronology of this defies logic, since this would mean that the Regular script, which is supposed to have inspired the creation of the Semi-Cursive script, came after the Semi-Cursive script!

The Running script – aka Semi-Cursive script, was, according to most sources, developed during the Eastern Han (CE 25-220) Dynasty. While most insist that it was developed as a compromise between the cumbersome Regular script and the illegibile Cursive script, they suggest a chronology for the creation of these three scripts that defies logic (as per the note above).

My belief (my best, most educated guess) is that the Semi-Cursive script was not at all created in response to either the illegibile Cursive script (the jīn cǎo script) or the cumbersome Regular script, but that it was simply a further extension of the earlier, prototype Cursive script, the zhāng cǎo script, whose sole function was writing (i.e., requiring legibility), not embellishment – albeit, speedwriting, as it were.

If we take this route, then we avoid entirely the trap of illogical chronology. It is also a much more satisfying explanation, since it links back to the original raison d'être of the very first cursive script, namely, a rapid form of writing. (Imagine it for a moment... a shorthand form of writing is developed, after which it is taken over by a group of artistic souls (calligraphers) who care not a whit about legibility, and what was supposed to be a shorthand form of writing has instead become an altogether different animal!)

Given this actual – not conjectured – background, it is more than plausible that someone interested in writing as a medium for verbalization, rather than being interested in the art of calligraphy, decided to revive the original shorthand form of speedwriting, perhaps with improvements, and thus the Semi-Cursive script was born. (I leave it to the reader to decide which is the most plausible explanation – the consensus opinion with its illogical chronology, or my conjectured explanation which is at least logically sound besides being plausible.)

Note however, that calligraphers eventually also cast themselves over even the Semi-Cursive script! In fact, one of the most famous Chinese calligraphers, Wang Xizhi, who lived during the Han Jin Dynasty (CE 265–420), wrote his much celebrated literary work, Preface to the Poems Composed at Orchid Pavilion, in the Semi-Cursive script. So the Semi-Cursive script ended up serving at least a dual function, meaning that in spite of being employed by calligraphers, it could still be read!

The Simplified script – is a modern script that stems from 1949 and can be seen as a halfway step (in this case, we are on much safer ground, as we know this to be a fact!) toward the creation of the later-to-come Romanized Pinyin script. While it was quite short-lived, it at least provided the impetus, as indicated, for developing a thorough script system that would make it easy for even uneducated people to learn to read and write. In fact, one can reasonably claim that had the Simplified script not been developed in 1949, Pinyin would most likely not have been developed as early as 1958, if at all!

Many of the ancient scripts depicted above are still current, even if only for very specialized uses. For example, the Large Seal script is commonly used today for shop signs and is additionally seen in certain calligraphic applications, while the Small Seal script is used in smaller applications such as for name tags and seals, and for logos on stationery, but also for certain calligraphic applications. The Clerical script is used both for shop signs and logos, but also on the headers of stationery.

The Grass script, aka Cursive script, quite understandably remains one the most popular Chinese scripts for calligraphic applications. The same applies to a large extent to the Semi-Cursive script, while the Regular script remains to this day the most common script for the purpose of writing traditional Chinese characters, and, as we have seen above, it is now incorported into all of the common Chinese fonts used in software applications.

Latinization and Its Influence on Chinese Orthographic Conventions

With the introduction of hanzi characters during the Han Dynasty – if not earlier – the Chinese language was written from right to left, and in columns, not in rows. Hanzi characters were designed to occupy an equal amount of two-dimensional space, regardless of their actual size. Thus hanzi characters were like blocks that were stacked one upon the other, with the initial character in each column located at the top of the column. Reading an ancient Chinese text thus consisted of reading first the righthandmost column from top to bottom, then the column immediately to the left of this from top to bottom, etc., etc.

Contact with the outside world, in particular, contact with Europeans, whose language was written from left to right and in rows, not columns, came to influence Chinese orthography, simply because it was not practical to include borrowed, Latin-based words – some of which could be quite long – in the standardized columns reserved for Chinese characters.

Moreover, since each Latin-based word is written from left to right, including such a word in a traditional Chinese text that was written right to left, albeit, in rows, was a bit jarring, therefore traditional Chinese texts eventually came to be written in the same manner as Latin-based text: in rows, from left to right. These considerations notwithstanding, the traditional method of Chinese orthography, i.e., of writing in columns from right to left and from top to bottom, is still practiced in Taiwan, though if the practice reflects a resistance to the language conventions of mainland China for ideological reasons, it will likely change in time, simply for practical reasons.

Foreign-Introduced Latin-Based Transliteration Of Traditional Chinese

It is generally agreed that the first to attempt to transcribe Chinese syllable-characters to a "Western" (Latin/ Roman) alphabet, using phonetic symbols (diacritical markings, etc.) to capture the tonal sounds of traditional Chinese syllable-characters, was the Portuguese Jesuit missionary duo, Matteo Ricci and Michele Ruggieri. While this is apparently true, there is an even earlier effort to transcribe the Chinese syllable-character to a foreign language...

The first Buddhist monks to arrive in China from India and who commenced shortly thereafter to translate their Buddhist texts from Sanskrit to Chinese quickly discovered certain structural regularities in the spoken Chinese language, such as an "initial sound", a "final sound" and a "suprasegmental tone" that characterized the pronunciation of the Chinese syllable-character.

It is conjectured that this knowledge formed part of the foundation of the later (but not much later, for we are still in China'a ancient, Three Kingdoms (CE 220-280) Period), universal phonetic system called Fanqie, which is a system that illustrates the pronunciation of unknown orthographies (any "foreign" (to the audience in question) language) with the help of snippets – often resorting to rhymes as a pronunication aid – of known (to the audience in question) orthographies.

As a modern-day example of Fanqie, imagine that a Danish speaker wished to explain the pronunciation of the family name of the former Danish government minister who chaired the 2009 UN clima conference in Copenhagen, Connie Hedegaard (now EU Commissioner for Climate Action), to an English speaker. The Danish speaker might say that the "Hede" of "Hedegaard" is pronounced like the "heath" of "heather" (which is decidedly not pronounced like the "Heath" of "Heathcliff", the protagonist in the 19th century English novel Wuthering Heights!).

However, the structural regularities of spoken Chinese discovered by the Buddhist monks who arrived in China as part of the cultural "commerce" that spread from East to West and from West to East along to the Silk Road breaks down when applied to later-era, polysyllabic spoken Chinese, such as Mandarin (in ancient China, as indicated above, most words consisted of single syllable-characters).

The aforementioned Portuguese Jesuit missionary duo, Matteo Ricci and Michele Ruggieri, developed the first consistent system of transcribing Chinese "words" into Latin in preparation for their joint work on a Portuguese-Chinese dictionary. The Latinized transcription of Chinese and the subsequent compilation of the Portuguese-Chinese dictionary were undertaken in the period 1583-88, though the manuscript was never published in book form, but instead somehow ended up in the Jesuit Archives in Rome, where it remained until it was rediscovered in 1934, though it would first be published in book form as late as 2001.

A reverse dictionary, from Chinese to Portuguese, making use of a comprehensive system of diacritical markings as a guide to tonal pronunication, was compiled by Ricci and another Jesuit colleague, Lazzaro Cattaneo, during the winter of 1598. Unfortunately, the manuscript to this work is lost, though it may yet surface one day.

Fortunately, Cattaneo, who was the driving force behind the comprehensive transcription system of the aforementioned Chinese-Portuguese dictionary authored in partnership with Matteo Ricci – but which unfortunately has disappeared – produced a separate manuscript of his comprehensive transcription system, a manuscript that would later be used by the Polish Jesuit missionary to China, Michał Boym, who, together with two Chinese assistants, would translate the first Latinized version of the Tang (CE 618-907) Dynasty period Nestorian Stele (the Nestorian Tablet) as it appeared in China Illustrata in 1667, an encyclopedic work that was compiled by the German Jesuit scholar, Athanasius Kircher.*(5)

The most famous Chinese name in history, Confucius (sorry if you thought it was Mao Zedong!), is just such a Latinized transcription that can be broken into four component parts: Con, a Western sound bastardization of the Chinese family name, Kong, common in the ancient province of Canton (Guangdong); fu (the same as in the other Latinization transcription systems); ci (tzu, zi or dz in other Latinization transcription systems); and us, which is strictly a Latin suffix indicating a masculine noun. Since fu-ci means "grand master", Confucius can be translated as "Grand Master Kong". Another famous Confucian, Mencius, is the Latinized version of "Master Meng".

Another well-known Latinization system for the Chinese language is the Wade-Giles system, named after the British diplomat and Sinologist, Sir Thomas Francis Wade who, in 1859, compiled The Peking Syllabary, a Romanized transliteration system, replete with tonal representations, designed specifically for the Peking (Beijing) dialect, and Herbert Allen Giles, who, a generation later (in 1892), would extend and expand The Peking Syllabary to a full-fledged transliteration system that rendered Mandarin Chinese into a Romanized alphabet.

The main drawback of the Wade-Giles system, which was the first widely accepted Chinese-to-Romanization transliteration system, is that in its diacritical markings, it relied partly on numerals rendered in superscript, while many smaller publishing houses were not sophisticated enough to be able to handle text with superscripts, with the result that the numerals were either omitted or were rendered in normal script, which of course defeated the purpose of the numerical superscripts as a guide to pronunciation, and which therefore made the Wade-Giles system less appealing as a universal transliteration medium, even if it is still widely referenced today.

Without going into detail, other Chinese-to-Romanization transliteration systems include: the 1902 EFEO (Ecole Francaise d'Extreme Orient) system; the so-called Postal system, which borrowed perhaps from the EFEO system since it was based on the French style of Romanization (it was used exclusively to render Chinese place names into a Romanized alphabet precisely for postal purposes... one should remember that much of China after the middle of the 19th century was under foreign influence, a result of the many international treaties involving Chinese trade and territorial concessions (think: Hong Kong and Macau, as well as parts of Shanghai, Qingdao and other larger port cities) that were forced upon the Qing government), and the Yale system that was developed at Yale University in the US as an aid to communication between the Allied military and Chinese resistance forces during the WWII-era Japanese occupation of much of China.

The Protestant movement in China also developed very comprehensive transliteration systems to deal with the many local languages that Protestant missionaries encountered, since in most rural areas of southern China, the people were not familiar with either Mandarin Chinese or Cantonese. The Protestants also made use of the Legge Romanization system developed by James Legge, the Scottish Sinologist and Protestant missionary who served in China during the latter half of the 19th century. The aforementioned Wade-Giles system built in fact on the advances inherent in the transliteration system developed by Legge.

The Chinese eventually developed their own systems of transliteration using a Latinized alphabet (generally referred to as a Romanized alphabet at this point in time, though earlier – for example, during the period when Portuguese Jesuit missionaies were just beginning to arrive in China – Latin was still an active written language) or a Latin-inspired alphabet. For certain, all of these systems of transliteration, including indigenous Chinese transliteration systems, built upon the advances that had preceded them.

Indigenous Latin-Inspired Transliteration Of Traditional Chinese

The first indigenous Chinese Romanization system was the Qieyin Xinzi ("New Phonetic Alphabet") that was developed by a certain Lu Zhuangzhang in 1892. The Qieyin Xinzi system was specifically developed to render the phonemes of the Xiamin dialect of the Minnan ("Southern Min") language to a Romanized alphabet. Here again, earlier work surely influenced the author, for it would have been very odd indeed had Lu not been aware of/ had not seen, the Romanization systems of Cattaneo, of Wade and Giles, etc.

Lu, together wih Wu Jingheng (originator of the so-called beansprout alphabet) and Wang Zhao (who had developed the Guanhua Zimu ("Mandarin Alphabet") in 1900) were all three part of a commission tasked with developing a special Mandarin-specific phonetic system that did not strictly belong to a Romanization system, even if it mimicked such a system in its tonal structure, and which came to be known as Zhuyin Fuhao, or Bopomofo, the latter a reference to the system's first four character-names: ㄅ,ㄆ,ㄇ and ㄈ.

In 1923, the Republic of China's own Ministry of Education set up the National Language Unification Commsission that consisted of five renowned scholars, including the aformentioned Chinese-American linguist, Zhao Yuanren, author of the humorous, single-syllable tale, The Story of Shi Devouring Lions, in which Zhao used only the syllable-character "shi​". The commission, after a year of deliberations and investigations into the task involved, took the decision to develop a Romanized transliteration system, and in 1928 presented their result, the Gwouyeu Romatzyh system.

This system differed from previous Romanized transliteration systems in that it relied neither upon diacritical marks or numerals in superscript, but instead offered variations on the Romanized alphabetic spellings of the roots of the syllable-character in order to reflect the inherent tonal variations of different Chinese character-syllables that are spelled the same in other transliteration systems except for the addition of diacritical marks or numerals in superscript (think of the near-homophones as an example).

A direct result of this is that the Gwouyeu Romatzyh system could be written using a standard QWERTY keyboard, which also made it accessible to all publishing houses, regardless of the level of sophistication.

In spite of these seemingly promising features – and in spite of its creators' hope of seeing it replace traditional Chinese entirely – the Gwouyeu Romatzyh system never quite caught on among Chinese users except where it was employed in dictionaries as an aid to the pronunciation of certain difficult Chinese characters. Perhaps the Gwouyeu Romatzyh system was simply ahead of its time.

Not surprisingly, Soviet Russia, a decade or so after the Russian Revolution (1917) attempted to develop a generic Romanized system for all of the Chinese languages – the Latinxua Sinwenz system – which was completely free from tonal aspects. Like the Gwouyeu Romatzyh system, the Latinxua Sinwenz system had great ambitions with regard to eventually entirely replacing traditional Chinese as well as the other Sinitic languages. The Latinxua Sinwenz system was developed in 1931 by a joint team of Russian and expat Chinese scholars living in Moscow in order to bring literacy to the large population of Chinese speakers who lived in the eastern region of the Soviet Union.

Latinxua Sinwenz achieved some popularity, having been embraced by leading Chinese intelletuals such as Lu Xun and Guo Moruo. It was introduced into the Soviet-controlled provincial regions of western China (the Shaanxi-Gansu-Ningxia region) during the period 1940-42. In spite of the adoption of Latinxua Sinwenz by the railway system in northeastern China as late as 1949, the Latinxua Sinwenz system had already been dealt the death knell in 1944, when the Chinese Communists were handed back the portions of western China (the Shaanxi-Gansu-Ningxia region) that the Soviet Union had seized during the period when China was under the rule of the Republic of China (the Kuomintang, or KMT), and warlordism flourished especially in western China.

The Chinese Communists, wishing to ingratiate themselves with the peoples of the area who had been living under Soviet Russian rule and who in fact resented the imposition, by the Soviet Russians, of the Latinxua Sinwenz system and its "foreign" orthography, halted the spread of the Latinxua Sinwenz system, reinstating traditional Chinese and its regional variants.

However, the Chinese Communists themselves would soon thereafter develop a similar Romanization system, the now widely accepted Pinyin system that would not only borrow heavily on the preceding Gwouyeu Romatzyh and Latinxua Sinwenz transliteration systems, but would also include some of the key linguists who had helped to develop those earlier Romanization/ Latin-inspired systems.

Wu Yuzhang, Ni Haishu and Lin Handa, all co-developers of the Latinxua Sinwenz system, as well as Li Jinxi and Luo Changpei, both co-developers of the Gwouyeu Romatzyh system, together with Zhou Youguang, a distinguished economist, administrator and former banker (Zhou lived and worked in Japan and the US until he returned to China in 1949) who was also a self-styled hobby-linguist, developed the Pinyin transliteration system during the period 1955-58. Zhou Youguang is generally regarded as the "Father of Hanyu Pinyin".

It must be said, however, that the committee of linguists assembled under the direction of Zhou Youguang made every effort – perhaps understandably so – to develop a Latin-inspired transliteration system that did not directly rely on the Latin alphabet, but as the former Premier, Zhou Enlai, in a 1958 speech announcing the creation of Hanyu Pinyin put it, the committee had wasted three years trying to avoid the direct use of the Latin alphabet but since "no satisfactory result could be obtained... the Latin alphabet was then adopted".

The former premier went on to say that "[i]n future, we shall adopt the Latin alphabet for the Chinese phonetic alphabet. Being in wide use in scientific and technological fields and in constant day-to-day usage, it will be easily remembered. The adoption of such an alphabet will, therefore, greatly facilitate the popularization of the common speech".

It is claimed that Chairman Mao (Mao Zedong), like many others who had earler eagerly embraced a Latin-based transliteration system, firmly believed that Hanyu Pinyin (simply referred to as Pinyin today) would completely replace all other forms of Chinese, including traditional Chinese.

As we now know, that did not happen. In fact, during the Cultural Revolution (1966-76), Pinyin disappeared altogether from the masthead of the People's Daily newspaper and from the Red Flag (Hong QI) Journal, where Pinyin had been used since its introduction in 1958 as translated subtitles to the main, or headline, titles that continued to be rendered in traditional Chinese characters on all official newspapers and magazines/ journals.

An attempt to refine Pinyin further (the "refinement" was completed and published in 1977) met with utter rejection after a few years, with the result that Pinyin reverted back to its original 1949 form in 1986. In the meantime, Pinyin had been recognized by the ISO (International Organization for Standardization) in 1982 as the official standard for transcribing traditional Chinese. Curiously, Pinyin came to incorporate an adaptation of the 'initial sound, final sound and suprasegmental tone' paradigm that characterizes the earliest form of Chinese – consisting of single-syllable words – as it was spoken when the first Buddhist missionaries (monks) arrived in China from India.

Today, other, albeit informal, Latin-based or Latin-inspired transliteration systems are used in order to render lesser Chinese languages such as Cantonese into a more modern, user-friendly form. One such effort is the rendering of colloquial Cantonese into a written language. It has become quite popular among Hong Kongers and other Cantonese speakers as an online chatting and instant-messaging language, though its use is restricted to informal, non face-to-face settings.

Similarly, other informal renderings of colloquial ethnic minority languages exist in written form, such as the Nüshu script of the local Chinese dialect, the Xiangnan tuhua, spoken by the people of the Xiao and Yongming River region of northern Jiangyong County, Hunan Province. The people of this region are bilingual, being able to converse in a dialect of Mandarin that is common to southwestern China, but when they feel the need to write in their southwestern Mandarin dialect, they invariably write it using traditional Chinese characters, not using the Nüshu script of their local Chinese dialect.