TITLE:

Favorlang songs transcribed in

Southern-Hokkien: Decipherment

AUTHOR(S):

Ochiai, Izumi

CITATION:
Ochiai, Izumi. Favorlang songs transcribed in Southern-Hokkien: Decipherment.
Proceedings of the 51st International Conference on Sino-Tibetan Languages and
Linguistics 2018: 47.

ISSUE DATE:
2018-09

URL:
http://hdl.handle.net/2433/235299

RIGHT:
Favorlang songs transcribed in Southern-Hokkien: Decipherment∗
Izumi Ochiai

Japan Society for the Promotion of Science

(Kobe City University of Foreign Studies)

Abstract
Favorlang is an extinct language of indigenous people in Taiwan (Austronesian). This language
was documented in Bible translation by Dutch missionary in the 17th century. However, the
grammar of this language remains unclear. The only useful literature for the analysis is a Vertrecht’s
wordlist of Favorlang-Dutch written in 1650 which was published in Campbell (1896). Under the
rule of Qing dynasty, a Chinese officer investigated aboriginal customs and documented folklores
from 34 villages (including six Favorlang songs) using Chinese characters (Huang 1722). These
Chinese character should be pronounced by Southern-Hokkien. These songs were provided with
translation in Chinese for each verse, but there were no inter-linear glosses. As a first step to the
deciphement, Florenz (1898) transliterated these songs into roman letters, Ogawa (date unknown)
identified some words, and Sato (1936:410-411) attempted to decipher one Siraya song and one
Favorlang song. This study presents a preliminary decipherment for all six Favorlang songs and
discuss some linguistic features observable in the songs.

keywords: Favorlang, Southern-Hokkien, Dutch, Pazih, philology

1 Introduction
Favorlang, also known as Babuza, is an Austronesian language once spoken in the western plains in
Taiwan. This language was first recorded by Dutch, who ruled over the western plains in the 17th
century (1624-1661) until they were expelled by Koxinga. There were several tribes in the western plains
at the time of Dutch occupation. Among these tribes, Siraya and Favorlang were strongly influenced
by the Dutch. Dutch priests earnestly converted Siraya and Favorlang people into Christianity. As for
Favorlang, Dutch priests studied their language and complied a Favorlang-Dutch dictionary (Happart
1650). They also wrote Bible in Favorlang (Campbell 1986).
Koxinga and his descendant’s occupation in Taiwan lasted about two decades (1661-1663). Then, it
was taken over by Qing dynasty. It was not until the Qing dynasty’s rule that the document on Favorlang
language reappeared.
Huang (1722) in his famous book Taihaishichalu, recorded lifestyles and social customs of people in
Taiwan. There was a section called ൪ଏ࿡ߟ [the customs of Taiwan savages in six aspects], where
Huang discuss resident, food, clothes, marriage, funeral, and tools of the Taiwan aborigines such as
Siraya, Hoanya, Favorlang, Papora and Taokas. In these sections, he added folk songs (൪Վ) of each
tribes as appendices. These folk songs are the subject of investigation.
There are six Favorlang songs*1 , which were recorded in different villages of Favorlang. In some cases,
a few villages shares the same song (section 2.1 and section 2.3). The name for these villages were

∗ This is a handout presented at the 51st International Conference on Sino-Tibetan Languages and Linguistics, Kyoto
University, Kyoto, 27th September 2018.
*1 Ogawa (1999) gave correspondence of a village to the tribal classification. I referred to this to decide on which song
belongs to Favorlang tribe among 34 songs recorded in Huang (1722).
transcribed in Chinese characters with Southern Hokkien pronunciation. These are ೋྛ*2 , അࣳ遴, 䤕ၲ
‫*ׯ‬3 , େಥ (section 2.1), ೆ (section 2.2), ౦ཐ, ੢ཐ (section 2.3), Ѩଋ*4 (section 2.4), ൒ઢ (section 2.5),
and 䤕ໄ拺*5 (section 2.6). The exact village name in Favorlang is not clear in some cases, so these are
represented by Southern Hokkien pronunciation in the following sections, according to Florenz’s (1898)
transliteration.
With the help of the Favorlang-Dutch dictionary and Bible in Favorlang, Ogawa (date unknown) tried
to decipher some Favorlang words in these six songs. Sato (1936) also tried to decipher one Favorlang
song recorded in ౦ཐ and ੢ཐ villages (section 2.3). Florenz (1898) transliterated the songs in Chinese
characters into roman letters using Sothern Hokkien pronunciation also translated the meaning of the
songs into German.
Supplementing Ogawa and Sato’s decipherment and referring to Florenz’s transliteration, this paper
aims to provide a preliminary interpretation for all six Favorlang songs. Section 2 presents the six songs
in original texts and adds the tentative transcription by the author. Section 3 discusses some findings
and observations.

2 Data
The original text in Huang (1722) is presented first. The songs are transcribed in Chinese characters
with Southern Hokkien pronunciation (shown on the left). There are several verses in each song. Each
verse is followed by Chinese translation (shown on the right).
Then, Florenz’s (1898) transliteration of the songs (on the left) and translation of the verses into
German (on the right) are shown*6 .
Finally, the author gives the transcription. The Florenz’s transliteration is segmented into words in
the first line with some modification by the present author. Florenz sometimes showed two variations
of transliteration for a Chinese word that has more than one pronunciation. The author decided which
should be used based on the identified words and its orthography in Happart’s dictionary as well as the
experience in the decipherment of Siraya songs (Ochiai and Shih 2018). The transcriber of the songs,
Shu-jing Huang, was likely to have a tendency to reuse a Chinese character assigning the same phonetic
value. For example, in Siraya songs, there are ຑ; it is transliterated as ba or ma by Florenz (1898). As
the result of our decipherment, it should be read as ma. This reading is applicable to ຑ appearing in
Favorlang songs (e.g. section 2.3).
The second line shows interlinear glossing. Some words were identified with words in the Happart’s
(1650) dictionary. In this case, the fourth line shows the words in Happart’s dictionary. Note that the
orthography is innovated by the Dutch. In most cases the roman letter more or less correspond to the
phonetic symbols in IPA; however, ei might be [i:], oe might be [u], g might be [G] or [x], ch might be
[x] or [h], a consonant geminate might indicate that the preceding vowel was short. For other words
that were not seen in Happart’s dictionary, the author inferred their meaning by preceding and following
words and the position they appear in a verse. Also, the identification words appearing more than once
(see section 3.1) were useful. The third line is the Chinese translation of the songs into English. Fourth
line shows the ideitified words as it appears in Happart’s dictionary with exception of the song in Babuza
village (section 2.6). This song is peculiar in that it has several words that belongs to Pazih. In this case,
the Pazih words in Li and Tsuchida’s (2001) Pazih dictionary, which was compiled based on Naoyoshi
Ogawa’s manuscript, are presented.

*2 Takeis village in Happart (1650).

*3 Batshekan village in Happart (1650).
*4 Asock village (Ino 1996).
*5 It is most likely to be Babuza village.
*6 German translation is not so important in deciphering the songs; however the author decided to show Florenz’s
attempt in translating Chinese into German. Perhaps, some may appreciate his translation.
2.1 “Tax payments” in Jilim, Matsilin, Batsikan and Toatut villages

ʰೋྛഅࣳ遴䤕ၲ‫ׯ‬େಥ࢛ࣾೲᰥՎʱ
㠧ᅵ㠧ୡᏌ㑌ᬠ ʮߞాԂʯ
ଖ䆩໹௝ಹ ʮѪ޷೥‫ܠ‬ʯ
෉ߕഅཷจའ ʮัࣛ‫ڈ‬ʯ
ଖจའ‫ג‬ḍ ʮࣛෆಘҳʯ
‫ࠫཷ׵؁‬ຑྗ‫׵؁ذ‬ ʮқᰥۜಘૣ‫׬‬ᰥʯ
അ໬໹*7 哰໹*8 ଖ㠴ҹ໹*9 ʮՄᬐ࿝໸Ѫ੯ʯ
ᅴ‫ؚ‬ჿ䙮ฏᤈ໹*10 嚎ଖᄺങҳ ʮզ౳ၼိշᒜҿञᬤՎʯ

Drei Stämme der Jilim, Toatut und Matsilin, Zahlung der Abgaben
pa-oan-pa-tat-si-boe-lin “Wir pflügen die Felder,
ki-ma-ya-tin-na Wir lieben den Prospekt eines guten Jahres.
hu-kah-ma-lyu-bun-lan Hirsche zu gangen gehen wir;
ki-bun-lan-thi-li Die Hirschke können uns nicht enfliehen.
kam-hoan/oan-lyu-ba/mo-lek-tsi/ki-kam-hoan/oan Eintauschend [dafür] Abgaben-Silber,
können wir fruüh die Abgaben bezahlen,
ma-yu-ya-lo-ya-ki-im-ya Und werden (können) die Liebe
und Teilnahme des Alten gewinnen.
ui-ham-o-soah-peng-ban-ya-ho-ki-lam-mai-it Wir kommen zurück, vergnügen und freuen uns,
trinken Wein und singen in wein heiterer Stimmung.”

(1) paoan patatsi boelin

go? plow? ?
‘We plough the field.’
(2) kima ya tinna
expect? nom? harvest?
‘We expect a good harvest.’
(3) hukah malyu bunlan
hunt? good deer
‘We hunt deers.’
- mario ‘good,’ binnan ‘deer’
(4) ki bunlan thili
obl? deer inescapable?
‘Deers cannot be escaped.’
- binnan ‘deer’
(5) kamhoan lyusa balek ki kamhoan
exchange? ? money(copper) obl? exchange?
‘We exchange it with money and pay the tax.’
- barieg ‘copper’
(6) mayu ya loya ki liimya
? nom? elder obl ?
We will get the respect and compassion from our elders.

*7 This character as the radial of “mouth.”

*8 This character as the radial of “mouth.”
*9 This character as the radial of “mouth.”
*10 This character as the radial of “mouth.”
 - loya ࿝໸ is a loanword from Southern Hokkien (Florenz 1898:143)
(7) uiham o soah pengban ya hoki lammaiit
drink wine ? ? nom? together? sing
‘We come back, then happily we drink and sing together.
- eicham ‘drink,’ o ‘liquor,’ rummait ‘sing’

2.2 “Drinking together” in Lam village

ʰೆࣾ။ҿՎʱ
㠧࿝ᖯ唭䆩ྲྀ຤໼ ʮߞాԂ۰޷೥ࡀʯ
㠧ࢥࠫཐ㠧ࢥ᫛࠯ ʮᏅಘຑᏅಘถʯ
അཷจའ唭ଧ֏ ʮัಘࣛ‫׌‬ଟʯ
ଧ‫כ‬ଧಸଧ჌ެਃ໹*11 ᇋ ʮ෕ࢠ૆ଙᴡိҿञʯ
টከྲྀ嚎唭ᄺങҳ ʮᓣ‫ݺ‬Վএҝᒜʯ

Stamm der Lam, Versammelt trinken

pa-lo/no-oan-li-ma-lyu-boat-si “Die Felder pflügend treffen wir ein glückliches Jahr an:
pa-su-sa-le-pa-su-lok-so Wir haben Hanf geerntet und haben Reis geerntet;
ma-lyu-bun-lan-ki-ta-hai Überdies haben wir Hirsche zahlreich gefangen.
ta-mau-ta-nai-ta-pau/pū-kong-sim-ya-ts‘ya Vater und Shon, Grossvater und Enkel
kommen allesamt und trinken Wein;
tasu/tsiau-bi/mi-lyu-ho-ki-lam-mai-it Freudig johlend und Lieder singend vergnuügen wir uns.”

(8) palooan ki malyu boatsi

work.in.field? obl good year
‘We plough the field and we have a good year[’s harvest].’
- mario ‘good,’ baas ‘year’
(9) pasu sale pasu lokso
get thread[made of ramie] get rice
‘We get ramie and we get rice.’
- sarries ‘thread, yarn,’ dasso ‘rice’
(10) malyu bunlan ki tahai
good deer obl? many?
‘We hunt many deers.’
- mario ‘good,’ binnan ‘deer’
(11) tamau tanai tapu kongsim ya ts‘ya
father mother all together nom drinking.party?
‘Father, son, grandfather and grandson come togetehr to drinking party.’
- tamau ‘father,’ tanai ‘mother,’ tapos ‘all’
(12) tsaubilyu hoki lammaiit
happy? together? sing
‘We happily sing together.’
- rummait ‘sing’

*11 This character as the radial of “mouth.”

2.3 “Seeing out the year” in Tangle and Saile villages

ʰ౦੢ཐࣾ౓೥Վʱ
㠧Ԃ㠧ୡᏌ㑌ᬠແಹ ʮߞాԂʯ
അྲྀฏ໹*12 ௝ಹຑཹჿ䏢 ʮѪ೥ࡀᏅ੒ʯ
෉ߕഅཷจའ ʮัࣛʯ
‫׵؁‬ຑจ‫ۍ‬ຑྗ ʮқۜ‫׬‬ᰥʯ
ີྛ䆩ྲྀ໹*13 嚎ᰃ*14 ‫ؚ‬ ʮՄ‫ڈ‬ᬳञա೥ʯ

Stämme der Tang-sai-le, Beim Zurücklegen des Jahres

pa-oan-pa-tat-si/su-bae-moa-lin-bu-na “Wir pflügen die Felder,
ma-lyu-peng-ya-tin-na-ba/mo-lyu-o/a-tap Wir lieben die Ernte des Jahres,
hu-kah-ma-lyu-bun-lan Wir fangen Hirsche,
kam-hoan-ba/mo-bun-hin-ba/mo-lek Tauschen Silber [dafür] ein und bezahlen
[damit] die Abgaben.
bit-lim-ma-lyu-ya-ho-hui/ui-ham Wir wollen gehen und Wein brauen,
um das Neujahrsfest zu feiern.”

(13) paoan patatsi boelin buna

go? plough ? field
‘We plough the field.’
- bonna ‘field’
(14) malyu pengya tinna malyu atap
good ? ? good harvest?
‘We are delighted with this year’s harvest.’
- mario ‘good’
(15) hukah malyu bunlan
get? good deer
‘We hunt deers.’
- mario ‘good’, binnan ‘deer’
(16) kamhoan mabun hinmalek
exchange? silver? exchange
‘We exchange it to silver and pay the tax.’
- chummollok ‘to exchange’
(17) bitlim malyu ya ho uiham
brew? good? nom? wine drink
‘We are able to go brewing wine and see out the year.’
- mario ‘good,’ eicham ‘drink’
In (17), cha huicham could be one word, consisting of cha- [xa?] a reduplicated prefix and a stem,
huicham. However, there remains a problem that segments x and h were indistuinguisable if this was
the case.

*12 This character as the radial of “mouth.”

*13 This character as the radial of “mouth.”
*14 This character as the radial of “mouth.”
2.4 “Praising ancestors” in Asok village

ʰѨଋࣾᨙ૆Վʱ
䆩ཹ໹‫כ‬຿䆩㠧ಹ ʮզ૆ԧ࠷༐໠ʯ
ຑཬ຤จའ෍຿຿೹ಹ ʮ۰ࣛೳ‫׆‬ଊʯ
㠧ग़࿊唭ߕങଧট ʮᲁ૸௚ಉԙഅʯ
෍຿຿ຄჿࠫከᬤ ʮ۰ञ៣ҿෆᬪʯ

Stamm der Asok, Lobpries der Ahnen

ma-lyu-ya-mau-bu-ma-pa-na “Unsere Ahnen [waren] im höchsten Grade tapfer;
ba/mo-li-bi-bun-lan-po-bu-bu-pa-na Antreffend den Hirsch, konnten sie [ihn] lebendig fangen;
pa-ts‘ut-li-ki-kah-mai-ta-chau Im Weltlauf [ware sie] vollkommen gleich einem Pferde;
po-bu-bu-pot/put-o/a-sa-bi/mi-ham Antreffend den Wein, wie viel sie auch tranken,
wurden sie nicht betrunken.”

(18) malyu ya mau buma pa na

good nom ? ancestor? ? 1sg.gen
‘My ancestors were the bravest.’
- mario ‘good,’ naa ‘my’
(19) malibi bunlan po bubu pa na
hunt.animal.while.alive? deer ? ancestor ? 1sg.gen?
‘When they hunted, they could catch a deer while it is still alive.’
- boeboe ‘ancestor,’ naa ‘my’
(20) pat‘utliki kamai tachau
run.straight? like? horse?
‘They ran fast and straight like a horse’
(21) po bubu poto sa miham
? ancester neg? ? drink
‘They never got drank.’
- boeboe ‘ancestor,’ micham ‘drink’
There are unknown elements po before bubu ‘ancestor’ in (19) and (21). This may be an prefix attaching
to bubu, probably indicating plurarity. In (18), it is likely that maubuma corresponds to ‘ancestor.’ This
is rather different but has some resemblance to po-bubu in (19) and (21).

2.5 “Drinking together” in Poansan village

ʰ൒ઢࣾᡉҿՎʱ
ᚸ֯෉ߕจೆ ʮัಘࣛʯ
ࢧඋୢჿଧ ʮᏅಘถʯ
ີྛ䆩ྲྀჿ嚎 ʮ၏Լञʯ
อ຿຿ଖѨ঻ᤈॄჿ嚎 ʮࣾऺᴡိᩱፍ။ҿʯ

Stamm der Poansan, Versammelt trinken

tsin-kak-hu-kah-bun-lam “Wir haben Hirsche gefangen,
tsi-pi-sin-o/a-ta Wir haben Reis geerntet,
bit-lim-ma-lyu-o/a-ho Wir haben Wein bereitet;
po-bu-bu-ki-a-syau-ban-sip-o/a-ho [Ihr] Leute des Stammes allesamt kommet,
veranstaltet eine festliche Ceremonie und trinket im Verein!”
 (22) tsinkak hukah bunlam
hunt hunt? deer
‘We could catch deers.’
- tumkah ‘hunt,’ binnan ‘deer’
(23) ki pisin ota
lnk? get? rice
‘We could get rice.’
- adda ‘rice’
(24) bitlim malyu o ho
brew? good wine? wine?
‘Then, we brew wine.’
- mario ‘good,’ o ‘wine’
(25) po bubu ki asyau bansip o ho
? ancestor/people? lnk? play drink? wine? wine?
‘People play and drink wine together.’
- boeboe ‘ancestor,’ o ‘wine’

2.6 “A man and a woman singing in turn in a party” in Babusa village

ʰ䤕ໄ拺ࣾஉ්။ҿጯ౴Վʱ
࣐䤕ᄆല*15 ʮ༮൪੥൪්ઌՎʯ
࣐ୡऒቭᡵ ʮ൪්੥༮൪ઌՎʯ
࣐䤕ྗᬏᬏ༝‫ܥ‬Ҿ࿊‫ޤ‬ຑᙌ*16 ʮ൪ᐌೊ්ਓ‫׌ࣕݡ‬ඒʯ
࣐ୡऒຑୡഅྡྷ唭ॄ֨ ʮ්ᐌೊஉਓӳ༤݉ೳড૸ʯ
࣐䤕ྗᬏᬏූഅ‫ޤ‬ଧ࿝຤᫛Ҿౕࡵ ʮ൪ᐌೊ්ਓࡏՈೳཆ㮦豖ኂೳᬳञʯ
࣐ୡऒୡְְຑҸ౗ჯจೆ‫ޤ‬໢Ռ䆩 ʮ්ᐌೊஉਓ্ࢁೳัࣛຢೳߞాԂʯ
ඒॄՌሇ໹*17 ኩທइࠜല*18 ቭᡵ‫ذ‬Ҿౕࡵ ʮࠓऺࣾօେᓣ‫ت‬࿨Վҿञʯ

Stamme der Babusa (Niaubūsa), Männer und Weiber vereint trinken und halten ein Zwiegespräch
ji/ni-ba/nyau-sin-bok “Der junge Barbar bittet das Barbarenweib
zuerst zu singen:
ji/ni-tat-ja-bu-nau Das Barbarenweib bittet
den jungen Brbaren zuerst zu singen.
ji/ni-ba/nyau-lek-mai-mai-yu-he-in-li-k‘it-ba/mo-yong Der Barbar spricht: Du Weib bist
weise und noch dazu schön
ji/ni-at-ja-ba/mo-tat-ma-lin-ki-sip-kek Das Weib sagt: Du Mann bist ein Held und
verstehst dich zugleich auf das Schnellaufen.
ji/ni-ba-nyau-lek-mai-mai-hu-ma-k‘i-ta-lo-boat-lok- Der Barbar spricht: Du Weib kannst zu Hause
in-lok-sat die Hühner und Schweine pflegen und kannst
Wein brauen.
ji/ni-tat-ja-hek-hek-ba/mo-un/in-to-teng-bung-lam Das Weib spricht: Du Mann steigst
-k‘it-bong-koh-ma in die Berge und kannst Hirsche fangen,
und kannst auch die Felder plfügen.
bi-sip-koh-hai-ya-oan-li-chak/chok-kin- Jetzt sing [wir] Alle im ganzen Stamme
bo/mo-bu-nau/lo-tsi/ki-in-no-sat in grosser Freude und Lust,
und zum Gesang einstimmend trinken [wir] Wein.”

*15 This character as the radial of “mouth.”

*16 This character as the radial of “mouth.”
*17 This character as the radial of “mouth.”
*18 This character as the radial of “mouth.”
 (26) jiba sinbok
woman? first?
‘A young man asks a woman to sing first.’
(27) jitat ja bunau
man? ? song?
‘The woman asks the man to sing first.’
(28) jiba lek maimai yuheinlik‘it mayong
woman ? woman wise? beautiful?
‘The man says the woman is wise and beautiful.’
- Pazih mamais ‘woman’ (Li and Tsuchida 2001:181)
(29) jitat ja batat malin ki sipkek
man? ? young.man brave? lnk? run.fast?
‘The woman says the man is not only brave but also able to run fast.’
- badda ‘child (male)’
(30) jiba lek maimai huma k‘itta loboatlok innosat
woman? ? woman house see(take.care?)? domesticated.animals? wine
‘The man says the woman is not only capable of raising chickens and pigs at home but also brewing
wine.’
- Pazih mamais ‘woman’ (Li and Tsuchida 2001:181), xuma ‘house’ (Li and Tsuchida 2001:325),
inusat ‘wine’ (Li and Tsuchida 2001:131)
(31) jitat ja tathekhek main toteg bunlam k‘itbong kohma
man? ? ? hunt? deer cultivate? field
‘The woman says the man is capable of hunting deers in the mountain but also cultivating land.’
- binnan ‘deer,’ ema ‘field’
(32) bisipkohhai ya oanlichakkinbo bunau ki innosat
? ? ? song? lnk? wine
‘Today everyone in the village is very happy and sings and drinks together.’
- Pazih inusat (Li and Tsuchida 2001:131)

3 Discussion
3.1 Repetitive words
Some identified words appears more than once. These include malyu (mario) ‘good,’ bunlan/bunlam
(binnan) ‘deer,’ ho (o) ‘wine,’*19 uiham/miham (eicham/micham) ‘to drink,’ lammaiit (rummait) ‘to
sing,’ po-bubu (boeboe) ‘ancestor.’ The words in brackets show Happart’s orthography.
Table 7 shows the list of identified repetitive words. The transliteration of the repetitive words followed
by the example number and the original transcription in Chinese character.

Table. 7: Identified repetitive words

Transliteration Example number Chinese character Happart (1650)

malyu (3) അཷ mario ‘good’
malyu (8) 䆩ྲྀ
malyu (10) അཷ
malyu (14) അྲྀ
malyu (14) ຑཹ
malyu (15) 䆩ཷ

*19 In the song in Babuza village (section 2.6), innosat (inusat) ‘wine’ appears twice but this word belongs to Pazih.
 malyu (17) 䆩ྲྀ
malyu (18) 䆩ཹ
malyu (24) 䆩ྲྀ
bunlan (3) จའ binnan ‘deer’
bunlan (4) จའ
bunlan (10) จའ
bunlan (15) จའ
bunlan (19) จའ
bunlam (22) จೆ
bunlam (31) จᄺ
o (7) ჿ o ‘wine’
ho (17) 嚎
o/ho (24) ჿ/嚎
o/ho (25) ჿ/嚎
uiham (7) ᅴ‫ؚ‬ eicham ‘to dirnk’
uiham (17) ᰃ*20 ‫ؚ‬
miham (21) ከᬤ micham *21 ‘to drink’
lammaiit (7) ᄺങҳ rummait ‘to sing’
lammaiit (12) ᄺങҳ
po-bubu (19) ෍຿຿ boeboe ‘ancestor’
po-bubu (21) ෍຿຿
po-bubu (25) อ຿຿

It can be said that Huang’s (1722) use of Chinese character for a particular sound is not consistent
in some cases. For malyu (mario ‘good’), he uses അɹ䆩 or ຑ in the first syllable. The same root അ is
used in the two characters. For the second syllable, ྲྀɹཷ or ཹ for the second syllable. The same root
ཹ is used in two characters.
‘Deer’ (binnan) has two variations; one has a word-final n, the other has a word-final m. It is not clear
whether this difference shows the dialectal variation. For the words with word-final m, the final syllable
is wither ೆ or ᄺ; the root is the same.
For ‘drink’ (eicham), the first syllable is written by either ᅴ or ᰃ (with the radical of mouth).
Including micham (m-icham ‘drink (actor voice)’), the final syllable is written by either ‫ ؚ‬or ᬤ. For
ancestor (boeboe), the common prefix-like element, po, is eitehr written by ෍ or อ.
There are other repetitive words which the author could not find in Happart’s (1650) dictionary.
However, the overall meaning is inferable form the Chinese translation. The following table shows the
list of these repetitive words, the example number, the original Chinese transcription, and the inferred
meaning (Table 8).

Table. 8: Unidentified repetitive words

Tarnsliteration Example number Chinese character Inferred meaning

paoan (patatsi boelin) (1) 㠧ᅵ (㠧ୡᏌ㑌ᬠ) ‘cultivate?’
paoan (patatsi boelin) (13) 㠧Ԃ (㠧ୡᏌ㑌ᬠ)
palooan*22 (8) 㠧࿝ᖯ
tinna (2) ௝ಹ ‘year-end?’
tinna (14) ௝ಹ
hukah (3) ෉ߕ ‘to hunt?’
hukah (15) ෉ߕ
hukah (22) ෉ߕ

*20 This character as the radial of “mouth.”

*21 This shares the root with the preceding two items. In micham, the initial m is probably an actor voice prefix.
*22 This form has an extra segment l compared to the preceding to words.
 kamhoan (5) ‫׵؁‬ ‘to exchange?’
kamhoan (5) ‫׵؁‬
kamhoan (16) ‫׵؁‬
pasu (9) 㠧ࢥ ‘get?’
pasu (9) 㠧ࢥ
bitlim (17) ີྛ ‘brew?’
bitlim (24) ີྛ
hoki*23 (7) 嚎ଖ ‘together?’
hoki (12) 嚎唭
jiba (lek) (28) ࣐䤕 (ྗ) ‘woman?’
jiba (lek) (30) ࣐䤕 (ྗ)
jiba (26) ࣐䤕
jitat(ja) (27) ࣐ୡ (ऒ) ‘man?’
jitat(ja) (29) ࣐ୡ (ऒ)
jitat(ja) (31) ࣐ୡ (ऒ)
bunau (27) ቭᡵ ‘song?’*24
bunau (32) ቭᡵ

3.2 Sound correspondence

In the six Favorlang songs, 23 words were identified with Happart’s (1650) vocabularies as seen in
Table 9. These words are the object of phonological analysis in this section. The author compared the
transliteration from the original Chinese transcription to the orthography in the Happart’s dictionary.

Table. 9: Sound correspondence in the identified words

Transliteration Example number Happart’s orthography

i uiham (7) eicham ‘to drink’
ii miham (21) micham ‘to drink’
iii o (7) o ‘wine’
iv tamau (11) tamau ‘father’
v tanai (11) tanai ‘mother’
vi buna (13) bonna ‘field’
vii hinmalek (16) chummollok ‘to exchange’
viii na (18) naa ‘I, mine’
ix bubu (18) boeboe ‘ancestor’
x malyu (3) mario ‘good’
xi lammaiit (7) rummait ‘to sing’
xii bunlan (3) binnan ‘deer’
xiii lokso (9) dasso ‘rice’
xiv bunlam (22) binnan ‘deer’
xv balek (5) barieg ‘copper’
xvi tsinkak (22) tumkah ‘to hunt’
xvii ota (23) adda ‘rice’
xviii batat (29) badda ‘children [male]’
xix ho (17) o ‘wine’
xx kohma (31) ema ‘field’
xxi sale (9) sarries ‘yarn’

*23 This and the next item both precedes rummait ‘to sing.’
*24 It was Ogawa (date unknown) who analyzed this word as ‘song’ (or ‘to sing?’); however, I do not know which
Formosan language he had in mind. At least, there seems to be no identical word in Favorlang dictionary (Happart
1650).
 xxii tapu (11) tapos ‘all’
xxiii boatsi (8) baas ‘year’

From (i) to (ix), the Florenz’s transliteration and the Happart’s orthography corresponds rather well
except for some vowel differences. In (x), (xi), and (xv), l in Florenz’s transliteration corresponds to r
in Happart’s orthography. In (xiv), l in the transliteration corresponds to n and in (xiii) it corresponds
to d. There is also an l in the transliteration that corresponds to l (vii).
In (xiv), the word-final consonant m in the transliteration appears as n in Happart’s orthography. In
(xv), the word-final consonant k in the transliteration appears as g [G] in Happart’s orthography. In
(xvi), the word-final consonant k in the transliteration appears as h in Happart’s orthography. In (xvii)
and (xviii), the word-medial t in the transliteration appears as d in the Happart’s orthography.
In (xvi), the word-initial ts in the transliteration appears as t in the Happart’s orthography. In (xx),
the word-initial h in the transliteration is missing in Happart’s orthography; similarly in (xxi), the
word-initial k is missing.
In (xxi) and (xxii), the word-final s in Happart’s orthography is missing in the transliteration. Con-
versely, the word-final s in Happart’s orthography appears in (xxiii); however, the segment is not word-
final in the translation. This appears as the onset, and it is followed by an extra vowel i.
In (xiii) and (xx), there are extra consonants, k*25 and h [P] in the transliteration. The extra consonants
appear word-initially.
In (vii), the segments m and i are doubled in the transliteration. First, m appears as the coda of the
first syllable. The same segment reappears as the onset of the following syllable. Similarly, i appears in
final position in the second syllable, and the same segment reappears as the initial vowel in the following
syllable. This kind of doubling method is also seen in (vii); however, the doubled segments are similar
but not identical (n and m).
In short, Chinese character, which contains a syllable, is not the most useful letter to transcribe
Favorlang. There were missing segments and unnecessary segments in the Chinese transcriptions which
lead to the difficulty in decipherment.

3.3 The song in Babuza village: Favorlang or Pazih

The song in Babuza village (section 2.6) is different from others in that it included Pazih words as well as
Favorlang words.*26 Table 10 shows the list of identified words in Favorlang and Pazih. For Favorlang,
the Happart’s orthography is shown in brackets. For Pazih, Li and Tuchida’s (2001) orthography is
shown in brackets.

Table. 10: Favorlang and Pazih words

Favorlang Pazih
batat (badda) ‘child [male]’ maimai (mamais) ‘woman’
kohma (ema) ‘field’ huma (xuma) ‘house’
bunlam (binnan) innosat (inusat) ‘wine’

The location of Babuza village may explain the possibility of language contact with Pazih people.
Among several Favorlang villages, Babuza was situated in the northernmost position (Tsuchida 1982).
Babuza village was situated on the north of Dadu river, which separates Babuza village from other
Favorlang villages which were situated on the south of Dadu river. In the north of Babuza village,
there were a few Pazih villages. Among Favorlang villages, Babuza village was closest to Pazih villages.

*25 However, the initial consonant k in the transliteration could be a retention of the earlier form (Proto-Austronesian
*qumah ‘swidden’ (Blust and Trussel). However, this is contradictory to the description in Li (2003:6) that Proto-
Austronesian *q had disappeared at the time of Happart (1650).
*26 It was Ogawa who first noticed this, which is evident in his notes in the manuscript (Ogawa, date unknown). He
wrote the name of a few Pazih village and Pazih words next to the words he deciphered.
People in Babuza village probably had frequent contact with Pazih people and eventually the language
in Babuza village underwent assimilation to Pazih by adopting Pazih words.

4 Conclusion
This paper tried to decipher six Favorlang songs transcribed by Chinese characters with Southern Hokkien
pronunciation in the early 18th century. This paper showed a preliminary result of the decipherment
(section 2) with the help of Favorlang dictionary and the identification of repetitive words (section 3.1).
Based on this result, phonological observations were discussed (section 3.2). Among six songs, the song
in Babuza village was peculiar in that it contained Pazih words (section 3.3). This may suggest the
language contact between Babuza village and the villages of Pazih tribe.
These Favorlang songs need syntactic analysis in the future. The songs recorded by Huang (1722) still
includes languages such as Taokas, Hoanya, Papora, which have not been deciphered. The observations
made in this study would be applicable in deciphering songs in other extinct Formosan languages.

References
Blust, Robert, and Stephen Trussel .2013. The Austronesian comparative dictionary: A work in progress,
web edition. Retrieved 2018/9/10.
Campbell, William, ed. 1896. The articles of Christian instruction in Favorlang-Formosan, Dutch and
English from Vertrecht’s manuscript of 1650: with Psalmanazarʟs Dialogue between a Japanese
and a Formosan, and Happart’s Favorlang vocabulary. London: Kegan Paul, Trench, Trübner.
Florenz, Karl. 1898. Formosanische Volkslieder nach chinesischen Quellen. Deutsche Gesellschaft für
Natur-und Völkerkunde Ostasiens VII:110-158.
Huang, Shu-Jing 㐋॑䚽. 1722. Taihai Shichalu ᢕւ࢖ᒍ㑚 [In Mandarin].
Happart, Georgius. 1650. Woord-boek der Favorlangsche taal, waarin het Favorlangs voor, het Duits
achter gestelt is (Translated in Campbell 1896). Manuscript.
Ino, Kanori 1996. Taiwan fieldwork diary, vol. 1, translated by Nan-jun Yang. Taipei: Yuan-Liou.
Li, Paul Jen-kuei, ed. 2003. English-Favorlang Vocabulary by Naoyoshi Ogawa. Tokyo: Research Insti-
tute for Languages and Cultures of Asia and Africa.
Li, Paul Jen-kuei. 2002. Introduction: Notes on Favorlang, an extinct Formosan language, in
English-Favorlang Vocabulary by Naoyoshi Ogawa, ed. by Paul Jen-kuei Li, 1-13.
Li, Paul Jen-kuei and Shigeru Tsuchida. 2001. Pazih dictionary. [Monograph Series No. A2.] Taipei:
Institute of Linguistics (Preparatory Office), Academia Sinica.
Ochiai, Izumi and Chao-Kai Shih. 2018. Siraya songs in dormancy: decipherment, presented at When
the West met the East: Early Western Accounts of the Lan- guages of the Sinosphere and their
Impact on the History of Chinese Linguistics, National Tsing Hua University, Sep. 10-11, 2018.
Ogawa, Naoyoshi. 1999. Formosan vocabularies in Taiwanfuzhi (Topographical and Historical Descrip-
tion of Taiwan-fu), Studies of Taiwan Aborigines 4:159-186.
Ogawa, Naoyoshi. [date unknown] Material pertaining to Siraya vocabulary and songs. Manuscript in
the possession of Institute for the Research Institute for Languages and Cultures of Asia and
Africa, Ogawa and Asai’s collection on Taiwan studies, No. OA068.
Sato, Bunichi. 1936. The Songs of the Civilized Aborigines of Formosa as Recorded in the Taiwan
fu-chih or Gazetteer of Formosa. The Japanese journal of ethnology 2:366-420 [In Japanese].
Tsuchida, Shigeru. 1982. A comparative vocabulary of Austronesian languages of sinicized ethnic groups
in Taiwan, part I. West Taiwan. Memoirs of the Faculty of Letters, University of Tokyo 7:1-166.
Tokyo: Faculty of Letters, University of Tokyo.
Tsuchida, Shigeru 1999. Two unpublished papers by Naoyoshi Ogawa [খ઒ঘٛͷະൃද‫ߘݪ‬ೋฤ].
Taiwan Genjūmin Kenkyū 4: 152-158.