Pe̍h-ōe-jī

The first people to use a romanized script to write Southern Min were Spanish missionaries in Manila in the 16th century.[2] However, it was used mainly as a teaching aid for Spanish learners of Southern Min, and seems not to have had any influence on the development of pe̍h-ōe-jī.[7] In the early 19th century, China was closed to Christian missionaries, who instead proselytized to overseas Chinese communities in South East Asia.[8] The earliest origins of the system are found in a small vocabulary first printed in 1820 by Walter Henry Medhurst,[9][10] who went on to publish the Dictionary of the Hok-këèn Dialect of the Chinese Language, According to the Reading and Colloquial Idioms in 1832.[9]

This dictionary represents the first major reference work in POJ, although the romanization within was quite different from the modern system, and has been dubbed Early Church Romanization by one scholar of the subject.[3] Medhurst, who was stationed in Malacca, was influenced by Robert Morrison's romanization of Mandarin Chinese, but had to innovate in several areas to reflect major differences between Mandarin and Southern Min.[11] Several important developments occurred in Medhurst's work, especially the application of consistent tone markings (influenced by contemporary linguistic studies of Sanskrit, which was becoming of more mainstream interest to Western scholars).[12] Medhurst was convinced that accurate representation and reproduction of the tonal structure of Southern Min was vital to comprehension:

Respecting these tones of the Chinese language, some difference of opinion has been obtained, and while some have considered them of first importance, others have paid them little or no intention. The author inclines decidedly to the former opinion; having found, from uniform experience, that without strict attention to tones, it is impossible for a person to make himself understood in Hok-këèn.

— W. H. Medhurst[13]

Frontispiece of Doty's Anglo Chinese Manual of the Amoy Dialect (1853)

The system expounded by Medhurst influenced later dictionary compilers with regard to tonal notation and initials, but both his complicated vowel system and his emphasis on the literary register of Southern Min were dropped by later writers.[14][15] Following on from Medhurst's work, Samuel Wells Williams became the chief proponent of major changes in the orthography devised by Morrison and adapted by Medhurst. Through personal communication and letters and articles printed in The Chinese Repository a consensus was arrived at for the new version of POJ, although Williams' suggestions were largely not followed.[16]

The first major work to represent this new orthography was Elihu Doty's Anglo-Chinese Manual with Romanized Colloquial in the Amoy Dialect,[16] published in 1853. The manual can therefore be regarded as the first presentation of a pre-modern POJ, a significant step onwards from Medhurst's orthography and different from today's system in only a few details.[17] From this point on various authors adjusted some of the consonants and vowels, but the system of tone marks from Doty's Manual survives intact in modern POJ.[18] John Van Nest Talmage has traditionally been regarded as the founder of POJ among the community which uses the orthography, although it now seems that he was an early promoter of the system, rather than its inventor.[10][16]

In 1842 the Treaty of Nanking was concluded, which included among its provisions the creation of treaty ports in which Christian missionaries would be free to preach.[6] Xiamen (then known as Amoy) was one of these treaty ports, and British, Canadian and American missionaries moved in to start preaching to the local inhabitants. These missionaries, housed in the cantonment of Gulangyu, created reference works and religious tracts, including a bible translation.[6] Naturally, they based the pronunciation of their romanization on the speech of Xiamen, which became the de facto standard when they eventually moved into other areas of the Hokkien Sprachraum, most notably Taiwan.[19] The 1858 Treaty of Tianjin officially opened Taiwan to western missionaries, and missionary societies were quick to send men to work in the field, usually after a sojourn in Xiamen to acquire the rudiments of the language.[19]

Quanzhou and Zhangzhou are two major varieties of Southern Min, and in Xiamen they combined to form something "not Quan, not Zhang" – i.e. not one or the other, but rather a fusion, which became known as Amoy Dialect or Amoy Chinese.[20] In Taiwan, with its mixture of migrants from both Quanzhou and Zhangzhou, the linguistic situation was similar; although the resulting blend in the southern city of Tainan differed from the Xiamen blend, it was close enough that the missionaries could ignore the differences and import their system wholesale.[19]

The fact that religious tracts, dictionaries, and teaching guides already existed in the Xiamen tongue meant that the missionaries in Taiwan could begin proselytizing immediately, without the intervening time needed to write those materials.[21] Missionary opinion was divided on whether POJ was desirable as an end in itself as a full-fledged orthography, or as a means to literacy in Chinese characters. William Campbell described POJ as a step on the road to reading and writing the characters, claiming that to promote it as an independent writing system would inflame nationalist passions in China, where characters were considered a sacred part of Chinese culture.[22] Taking the other side, Thomas Barclay believed that literacy in POJ should be a goal rather than a waypoint:

Soon after my arrival in Formosa I became firmly convinced of three things, and more than fifty years experience has strengthened my conviction. The first was that if you are to have a healthy, living Church it is necessary that all the members, men and women, read the Scriptures for themselves; second, that this end can never be attained by the use of the Chinese character; third, that it can be attained by the use of the alphabetic script, this Romanised Vernacular.

— Thomas Barclay[23]

A great boon to the promotion of POJ in Taiwan came in 1880 when James Laidlaw Maxwell, a medical missionary based in Tainan, donated a small printing press to the local church,[24] which Thomas Barclay learned how to operate in 1881 before founding the Presbyterian Church Press in 1884. Subsequently, the Taiwan Prefectural City Church News, which first appeared in 1885 and was produced by Barclay's Presbyterian Church of Taiwan Press,[24] became the first printed newspaper in Taiwan.[25]

As other authors made their own alterations to the conventions laid down by Medhurst and Doty, pe̍h-ōe-jī evolved and eventually settled into its current form. Ernest Tipson's 1934 pocket dictionary was the first reference work to reflect this modern spelling.[26] Between Medhurst's dictionary of 1832 and the standardization of POJ in Tipson's time, there were a number of works published, which can be used to chart the change over time of pe̍h-ōe-jī:[27]

Taiwanese kana used as ruby characters

Competition for POJ was introduced during the Japanese era in Taiwan (1895–1945) in the form of Taiwanese kana, a system designed as a teaching aid and pronunciation guide, rather than an independent orthography like POJ.[37] From the 1930s onwards, with the increasing militarization of Japan and the Kōminka movement encouraging Taiwanese people to "Japanize", there were a raft of measures taken against native languages, including Taiwanese.[38] While these moves resulted in a suppression of POJ, they were "a logical consequence of increasing the amount of education in Japanese, rather than an explicit attempt to ban a particular Taiwanese orthography in favor of Taiwanese kana".[39]

The Second Sino-Japanese War beginning in 1937 brought stricter measures into force, and along with the outlawing of romanized Taiwanese, various publications were prohibited and Confucian-style shobō (Chinese: 書房; pinyin: shūfáng; Pe̍h-ōe-jī: su-pâng) – private schools which taught Classical Chinese with literary Southern Min pronunciation – were closed down in 1939.[40] The Japanese authorities came to perceive POJ as an obstacle to Japanization and also suspected that POJ was being used to hide "concealed codes and secret revolutionary messages".[41] In the climate of the ongoing war the government banned the Taiwan Church News in 1942 as it was written in POJ.[42]

A decree (1955) banning Pe̍h-ōe-jī.

Initially the Kuomintang government in Taiwan had a liberal attitude towards "local dialects" (i.e. non-Mandarin varieties of Chinese). The National Languages Committee produced booklets outlining versions of Bopomofo for writing the Taiwanese tongue, these being intended for newly arrived government officials from outside Taiwan as well as local Taiwanese.[43] The first government action against native languages came in 1953, when the use of Taiwanese or Japanese for instruction was forbidden.[44] The next move to suppress the movement came in 1955, when the use of POJ for proselytizing was outlawed.[42] At that point in time there were 115,000 people literate in POJ in Taiwan, Fujian, and southeast Asia.[45]

Two years later, missionaries were banned from using romanized bibles, and the use of "native languages" (i.e. Taiwanese Hakka, Hakka, and the non-Sinitic Formosan languages) in church work became illegal.[42] The ban on POJ bibles was overturned in 1959, but churches were "encouraged" to use character bibles instead.[42] Government activities against POJ intensified in the late 1960s and early 1970s, when several publications were banned or seized in an effort to prevent the spread of the romanization. In 1964 use of Taiwanese in schools or official settings was forbidden,[44] and transgression in schools punished with beatings, fines and humiliation.[46] The Taiwan Church News (printed in POJ) was banned in 1969, and only allowed to return a year later when the publishers agreed to print it in Chinese characters.[42][47]

In 1974, the Government Information Office banned A Dictionary of Southern Min, with a government official saying: "We have no objection to the dictionary being used by foreigners. They could use it in mimeographed form. But we don't want it published as a book and sold publicly because of the Romanization it contains. Chinese should not be learning Chinese through Romanization."[48] Also in the 1970s, a POJ New Testament translation known as the "Red Cover Bible" (Âng-phoê Sèng-keng) was confiscated and banned by the Nationalist regime.[49] Official moves against native languages continued into the 1980s, the Ministry of Education and the Ministry of the Interior decided in 1984 to forbid missionaries to use "local dialects" and romanizations in their work.[42]

With the ending of martial law in 1987, the restrictions on "local languages" were quietly lifted,[50] resulting in growing interest in Taiwanese writing during the 1990s.[51] For the first time since the 1950s, Taiwanese language and literature was discussed and debated openly in newspapers and journals.[52] There was also support from the then opposition party, the Democratic Progressive Party, for writing in the language.[44] From a total of 26 documented orthographies for Taiwanese in 1987 (including defunct systems), there were a further 38 invented from 1987 to 1999, including 30 different romanizations, six adaptations of bopomofo and two hangul-like systems.[53] Some commentators believe that the Kuomintang, while steering clear of outright banning of the native language movements after the end of martial law, took a "divide and conquer" approach by promoting Taiwanese Language Phonetic Alphabet (TLPA), an alternative to POJ,[54] which was at the time the choice of the majority inside the nativization movement.[55]

Native language education has remained a fiercely debated topic in Taiwan into the 21st century and is the subject of much political wrangling.[56][57]

No.	Diacritic	Chinese tone name	Example listen
1	none	陰平 (yinping) dark level	kha 跤 foot; leg
2	acute	上聲 (sióng-siaⁿ) rising	chúi 水 water
3	grave	陰去 (yinqu) dark departing	kàu 到 arrive
4	none	陰入 (yinru) dark entering	bah 肉 meat
5	circumflex	陽平 (yangping) light level	ông 王 king
7	macron	陽去 (yangqu) light departing	tiōng 重 heavy
8	vertical line above	陽入 (yangru) light entering	joa̍h 熱 hot

The five tone markings used in pe̍h-ōe-jī, representing tones 2, 3, 5, 7, and 8

In standard Amoy or Taiwanese Hokkien there are seven distinct tones, which by convention are numbered 1–8, with number 6 omitted (tone 6 used to be a distinct tone, but has long since merged with tone 2).[73] Tones 1 and 4 are both represented without a diacritic, and can be distinguished from each other by the syllable ending, which is a vowel, ⟨-n⟩, ⟨-m⟩, or ⟨-ng⟩ for tone 1, and ⟨-h⟩, ⟨-k⟩, ⟨-p⟩, and ⟨-t⟩ for tone 4.

Southern Min dialects undergo considerable tone sandhi, i.e. changes to the tone depending on the position of the syllable in any given sentence or utterance.[66] However, like pinyin for Mandarin Chinese, POJ always marks the citation tone (i.e. the original, pre-sandhi tone) rather than the tone which is actually spoken.[74] This means that when reading aloud the reader must adjust the tone markings on the page to account for sandhi. Some textbooks for learners of Southern Min mark both the citation tone and the sandhi tone to assist the learner.[75]

There is some debate as to the correct placement of tone marks in the case of diphthongs and triphthongs, particularly those which include ⟨oa⟩ and ⟨oe⟩.[76] Most modern writers follow six rules:[77]

1. If the syllable has one vowel, that vowel should be tone-marked; viz. ⟨tī⟩, ⟨láng⟩, ⟨chhu̍t⟩
2. If a diphthong contains ⟨i⟩ or ⟨u⟩, the tone mark goes above the other vowel; viz. ⟨ia̍h⟩, ⟨kiò⟩, ⟨táu⟩
3. If a diphthong includes both ⟨i⟩ and ⟨u⟩, mark the ⟨u⟩; viz. ⟨iû⟩, ⟨ùi⟩
4. If the final is made up of three or more letters, mark the second vowel (except when rules 2 and 3 apply); viz. ⟨goán⟩, ⟨oāi⟩, ⟨khiáu⟩
5. If ⟨o⟩ occurs with ⟨a⟩ or ⟨e⟩, mark the ⟨o⟩ (except when rule 4 applies); viz. ⟨òa⟩, ⟨thóe⟩
6. If the syllable has no vowel, mark the nasal consonant; viz. ⟨m̄⟩, ⟨ǹg⟩, ⟨mn̂g⟩

A single hyphen is used to indicate a compound. What constitutes a compound is controversial, with some authors equating it to a "word" in English, and others not willing to limit it to the English concept of a word.[76] Examples from POJ include ⟨sì-cha̍p⟩ "forty", ⟨bé-hì-thôan⟩ "circus", and ⟨hôe-ho̍k⟩ "recover (from illness)". The rule-based sandhi behaviour of tones in compounds has not yet been clearly defined by linguists.[78] A double hyphen ⟨--⟩ is used when POJ is deployed as an orthography (rather than as a transcription system) to indicate that the following syllable should be pronounced in the neutral tone.[79] It also marks to the reader that the preceding syllable does not undergo tone sandhi, as it would were the following syllable non-neutral. Morphemes following a double hyphen are often (but not always) grammatical function words.[80]

In addition to the standard syllables detailed above, there are several regional variations of Hokkien which can be represented with non-standard or semi-standard spellings. In the Zhangzhou dialect, spoken in Zhangzhou, parts of Taiwan (particularly the northeastern coast around Yilan City) and parts of Malaysia (particularly in Penang), there is a final ⟨-uiⁿ⟩, for example in "egg" ⟨nūi⟩ and "cooked rice" ⟨pūiⁿ⟩, which has merged with ⟨-ng⟩ in mainstream Taiwanese.[81] The Zhangzhou dialect also has the vowel /ɛ/, written as ⟨ɛ⟩ or ⟨e͘⟩ (with a dot above right, by analogy with ⟨o͘⟩), which has merged with ⟨e⟩ in mainstream Taiwanese.