This post details the orthography of Namari, as well as the main romanisation systems in use. Namari is primarily written in kana, although in the past Han characters have also been used (their use in the present day is largely limited to disambiguation). In terms of romanisation, the most common system is a modified version of the Hepburn system used to transliterate Japanese, although other systems are also in use.
The orthography of Namari is not entirely phonemic (although kana-only Middle Namari orthography was phonemic). This is due to changes in how vowel sequences are pronounced:
- Sequences ending in /i/ or /u/ monophthongise.
- As a result, the distinctions between /eː/ and /ei/, and /oː/ and /ou/, are lost.
- Sequences ending in /e/ or /o/ then fill the void that was once occupied by the former diphthongs.
The end result is that Namari orthography is more complex than Japanese orthography. Vowel sequences may not be pronounced as they are written:
- /ai/ → [ɛː]
- /au/ → [ɔː]
- /ae/ → [ai̯]
- /ao/ → [au̯]
- /iu/ → [ʲɯː]
- /io/ → [iu̯]
- /ui/ → [yː]
- /ue/ → [ui̯]
- /ei/ → [eː]
- /eu/ → [ɤː]
- /eo/ → [eu̯]
- /oi/ → [øː]
- /ou/ → [oː]
- /oe/ → [oi̯]
As a result, Namari orthography can be said to be morphophonemic, because the sound changes also affected inflected forms (e.g. present kaku [kà.kú], past kaita [kɛ̀ː.tá]). I should mention, though, that short versions of the additional vowels listed above exist; however, they are extremely rare in native words and absent in Sino-Namari words.
In general, there are two different sets of kana used. The main set used in Namari is hiragana, while katakana is mainly used for emphasis. There are a number of collation orders used; however, the gojūon ordering (known natively as esone within Yaezora and isone outside) is the one most commonly used. The other orderings one may find are the iroha and ametuchi orderings, which I’ll also mention below.
The gojūon order splits the syllabograms into groups of five, each representing a separate initial consonant (in the order (null), /k/, /s/, /t/, /n/, /p/, /m/, /j/, /r/, /w/). Within each group the syllabograms are sorted based on their vowel, in the order /a/, /i/, /u/, /e/, /o/. These are listed below, for hiragana:
- (null): あ い う え お
- k-: か き く け こ
- s-: さ し す せ そ
- t-: た ち つ て と
- n-: な に ぬ ね の
- p-: は ひ ふ へ ほ
- m-: ま み む め も
- y-: や – ゆ (𛀁) よ
- r-: ら り る れ ろ
- w-: わ ゐ – ゑ (を)
- Extra: ん
The moraic nasal ん does not fit into the categorisation, hence it is simply placed at the end. The character 𛀁 is archaic; it represents /je/, which has largely merged with /e/ in the modern language. Similarly, the character を is only found in mixed writing, and solely as an accusative marker; it represents /wo/, which has merged with /o/.
The same list, but for katakana, is listed below:
- (null): ア イ ウ エ/(𛀀) オ
- k-: カ キ ク ケ コ
- s-: サ シ ス セ ソ
- t-: タ チ ツ テ ト
- n-: ナ ニ ヌ ネ ノ
- p-: ハ ヒ フ ヘ ホ
- m-: マ ミ ム メ モ
- y-: ヤ – ユ (エ) ヨ
- r-: ラ リ ル レ ロ
- w-: ワ ヰ – ヱ (ヲ)
- Extra: ン
The character エ is normally used for /e/; however, when transcribing ancient texts, it instead represents /je/ and 𛀀 is used for /e/ instead.
(A side note: the archaic kana 𛀁 ye and 𛀀 e are found in Unicode 6.0, dated to October 2010; if you do not have a relatively recent font, then you will not be able to see these kana.)
The other two orders are based on Japanese pangrams. The iroha ordering is as follows (listing the straight transliteration, Japanese reading and Namari reading):
- いろは にほへと (iropa nipopeto, iro wa nioedo, eyoha nioedo)
- ちりぬるを (chirinuruwo, chirinuru o, chīnuyō)
- わか よ たれそ (waka yo tareso, waga yo dare zo, waga yo taezo)
- つね ならむ (tune naramu, tsune naran, tune nayan)
- うゐの おくやま (uwino okuyama, ui no okuyama, uino okuyama)
- けふ こえて (kepu koete, kyō koete, keu koete)
- あさき ゆめ みし (asaki yume mishi, asaki yume miji, asake yume miji)
- ゑひも せす (wepimo sesu, yoi mo sezu, eimo sezu)
The kana 𛀁 and ん do not appear in the poem, and are simply appended in that order.
The ametuchi order is as follows:
- あめ つち ほし そら (ame tuchi poshi sora)
- やま かは みね たに (yama kapa mine tani)
- くも きり むろ こけ (kumo kiri muro koke)
- ひと いぬ うへ すゑ (pito inu upe suwe)
- ゆわ さる おふせよ (yuwa saru opuseyo)
- えの 𛀁を なれゐて (eno yewo narewite)
Again, ん is simply appended to the end.
As mentioned above, Namari orthography is morphophonemic, rather than phonemic. I explained earlier which vowel sequences correspond to which long vowels; however, this time I will list how the short versions of those long vowels are written:
- /kɛ/ = かぃ
- /kɔ/ = かぅ
- /ky/ = くぃ
- /kɤ/ = けぅ
- /kø/ = こぃ
It is important to note that the second graphemes are small, rather than normal-sized. Also note the absence of [ɯ]; this is actually an allophone of /u/ after palatals in the Yaezora dialect, meaning short [kʲɯ] is simply represented as if it were written kyu.
An -i kana followed directly by a small y- kana (or small ぇ for ye) indicates that the mora is pronounced with the consonant palatalised (e.g. きゃ for kya [kʲa]). Any consonant other than /j/ can be palatalised, although palatalised /w/ is archaic and has since been lost.
The kana く ku followed directly by a small w- kana indicates that the consonant is a labiovelar (e.g. くゎ for kwa [kʷa]). However, labiovelars have disappeared in most Namari dialects, becoming labials in the Yaezora and Haguya dialects and plain velars in the Chiyohara dialect.
(A side note: the small kana 𛅐 wi and 𛅑 we, and their katakana equivalents, are found in Unicode 12.0, dated to March 2019. As of writing, few, if any, fonts support these characters. In a pinch, ぃ can be used for wi and ぇ for we; however, the former creates ambiguity with /y/.)
Finally, the last small character used is っ. It simply indicates that the following consonant is geminated, and thus unlike the other small kana represents a full mora.
In Namari orthography, two diacritics, the dakuten and the handakuten, are used. The dakuten indicates that the consonant is voiced (e.g. が ga), while the handakuten represents aspirated consonants (e.g. こ゚ ho).
In romanisation, as I’ve mentioned, the system in most common use is a modified version of the Hepburn system. I’ve detailed the rules on the wiki, and they are still accurate, except with respect to aspirated consonants. The rules for aspirated consonants are below:
- /x/ is written as -h-
- /θ/ is written as -th-
- /ɸ/ is written as -f-
- Palatalised /θ/ is written as -sh-, not -ch-
- What would be written as fa, fi and fy- are instead written as ha, hi and hy- respectively, due to mergers.
- Even though tu and thu are pronounced identically, they are still written differently.
The other romanisation systems in use are a modified version of the Nihon-shiki system, and what is known as the Blylandic system.
The modified version of the Nihon-shiki system largely resembles the original, except with the addition of aspirated consonants and modifications to adhere to Namari phonology (e.g. -p- instead of -h-, the same as with the modified Hepburn system). Like with the modified Hepburn system, this modified system uses -h-, -th- and -f- for the aspirated consonants, except that they do not change at all when palatalised or in certain phonetic environments (with the exception that the topic marker is still written -ha).
The Blylandic system somewhat resembles the modified version of the Nihon-shiki system, except for the following differences:
- /j/ and palatalisation are written as -j-, not -y-.
- /θ/ is represented using the letter þ.
In all three systems, the extra short vowels are represented as follows:
- /ɛ/ = ä or è (the latter only in Hepburn)
- /ɔ/ = å or ò (the latter only in Hepburn)
- /y/ = ü
- /ɤ/ = ë
- /ø/ = ö
That will be it for now.