Guidelines:Romanization

See also: Template:Translit/Documentation, Template:Lang/Documentation, Module:Language/doc, Module:Language/Subtags

To retain site-wide consistency, Megami Tensei Wiki recommends certain romanization schemes for languages with non-Latin scripts. These schemes should be kept to in most cases, except for instances where another scheme would be required for contextual purposes.

Basic summary

Summary table

Language	Recommended scheme	Resource	Tool
Arabic	DIN 31635:2011-07, with modifications: Render the consonant ⟨ج⟩ as ⟨j⟩ instead of DIN 31635's ⟨ǧ⟩. Render the glottal stop at the beginning of words if spelled with a hamza or madda. Render the diphthongs ⟨au⟩ and ⟨ai⟩ as ⟨aw⟩ and ⟨ay⟩. Always render geminated semivowels (⟨w⟩, ⟨y⟩) as the consonant twice regardless of which vowel they follow (e.g. ⟨iyy⟩, not ⟨īy⟩ or ⟨ī⟩). Generally follow Standard Arabic, using pausal form of nouns and adjectives at the end of the sentence (or at the end of a standalone noun phrase) instead of always.	On-site
Avestan	Hoffman
Belarusian	Łacinka
Chinese	Hanyu Pinyin
Coptic	ALA-LC	[1]
Egyptian	Leiden Unified Transliteration Due to lack of font support, use ⟨ı͗⟩ (dotless i + right half-ring combining character) instead of ⟨ꞽ⟩ (glottal i).
Ge'ez	Encyclopaedia Aethiopica	[2]
Greek (to 1453)	ALA-LC, with some modifications: Include diacritical markings for the acute, circumflex, and grave. Omit macrons for eta and omega if they have a circumflex. For other diacritics, include both the macron and the given diacritic. Transliterate the iota subscript diacritic as a plain ⟨i⟩ after the given vowel. In an unaccented syllable, transliterate ⟨ᾳ⟩ as ⟨āi⟩ to distinguish from ⟨αι⟩. Include a diaeresis in the transliteration when present, except in the following sequences: ⟨αϋ⟩, ⟨εϋ⟩, ⟨οϋ⟩, and ⟨ωϋ⟩; these are already distinguished from diphthongs by using ⟨y⟩ instead of ⟨u⟩ for the upsilon. Transliterate the sequence ⟨υι⟩ as ⟨yi⟩ instead of ⟨ui⟩.	On-site
Greek (from 1453)	ELOT 743 (Type 2)
Hebrew and Judeo-Aramaic	Academic transliteration system (with spirantization) from The SBL Handbook of Style, 2nd edition, with one modification: Use half rings for ʾaleph and ʿayin (as in the first edition) instead of using single quotation marks.		[3]
Indo-Aryan and Dravidian languages	For Sanskrit, IAST For Pali, Pali Text Society romanization For Urdu, DIN 31635:2011-07 For other languages, ISO 15919 with the following options chosen: Non-uniform vowels option: Omit macrons for ⟨ē⟩ and ⟨ō⟩ if short equivalents do not exist in the given language. Simplified nasalization option: Always transliterate the anusvara and the Gurmukhi bindi as ⟨ṁ⟩, except at the end of a word in Malayalam where ⟨m⟩ should be used. Always transliterate the candrabindu as ⟨m̐⟩. Always transliterate the tippi in Gurmukhi script as ⟨ṃ⟩. Malayalam alveolar option: For Malayalam only, rather than ⟨ṟṟ⟩ and ⟨nṟ⟩, use ⟨ṯṯ⟩ and ⟨nṯ⟩, respectively. Diacritic option: Use diacritics. One-way urpha option: Transliterate the urpha ⟨र्‍⟩ in Nepali as plain ⟨r⟩ rather than ⟨r̆⟩. Slight modification: For Indo-Aryan languages with schwa deletion, do not include the silent short ⟨a⟩ (e.g. for Hindi, do śiv rather than śiva for शिव). There is no strict rule on what short ⟨a⟩ vowels are silent or pronounced, and it is left up to the editor to determine whether it is silent or not.		[4], on-site
Japanese	Modified Hepburn
Korean	Revised Romanization	[5]	[6][dead link]
Kurdish	DIN 31635:2011-07	On-site
Misc. ancient West Semitic languages	Conventional scholarly transliteration. Vocalized forms may be added after in parentheses using the standard five Latin-alphabet vowels; if the vowel is long, have a macron over it (unless it is written using a mater lectionis, in which case use a circumflex).	On-site
Mongolian and Buryat	For Mongolian script (labeled as Classical Mongolian on the wiki), a variant of the Mostaert-Vladimirtsov system For Cyrillic (Khalkha Mongolian and Buryat), THL Mongolian-Cyrillic Transliteration For the Buryat letter ⟨һ⟩ not present in Khalkha Mongolian, romanize as ⟨h⟩. If this causes ambiguities with digraphs, put a middle dot ⟨·⟩ between the letters.	On-site, [7]
Old Church Slavonic and Old East Slavic	Variant of the conventional scholarly transliteration.
Persian	For Perso-Arabic script, DIN 31635:2011-07, with modifications: Render the consonant ⟨ج⟩ as ⟨j⟩ instead of DIN 31635's ⟨ǧ⟩. Render the consonant ⟨ض⟩ as ⟨ż⟩ (like in DMG (1935)) instead of as ⟨ḍ⟩. Render some consonants and vowels differently depending on variety of Persian. For Classical and Afghan Persian, split long ⟨ī⟩ and ⟨ū⟩ into ⟨ī⟩/⟨ē⟩ and ⟨ū⟩/⟨ō⟩ based on etymology. Render the geminated semivowel ⟨y⟩ as the consonant twice regardless of which vowel it follows. The sequence ⟨iyy⟩ is ⟨ī⟩ in word-final position, as in DIN 31635:2011-07. Include the consonant ⟨y⟩ in the iżāfa suffix after a vowel, rather than deleting it. For Cyrillic script (Tajik), TBD	On-site
Russian	BGN/PCGN	[8]	[9]
Serbian	Gajica
Syriac	ALA-LC, with two modifications: Mark spirantization of the bgdkpt letters with a macron below if spirantization is marked in the original text; vocalized texts mark spirantization with a dot below the letter. Use a circumflex for matres lectionis instead of an acute.	[10]
Thai	ISO 11940		[11]
Tibetic languages	THL Extended Wylie	[12]	[13]
Ukrainian	Idiosyncratic system	On-site

Individual languages

Arabic and Persian

This section is for rules on transliterating Arabic, as well as rules on transliterating Persian written in the Perso-Arabic script.

Note that, for the purposes of this documentation, Classical Persian is considered the "default" variety of Persian. Note also that while this page outlines the general patterns for converting from Classical pronunciations to modern Iranian Persian pronunciations, some individual words have evolved in a way that does not 100% fit with the patterns described.

Letters

Several letters are named differently in Persian but, for simplicity, only their Arabic names are given. Persian-exclusive letters have their Classical Persian names given.

Consonants

Letter		ArArabic	FaPersian
ا/ء	Hamza/ʾalif	ʾ
ب	Bāʾ	b
پ	Pē	—	p
ت	Tāʾ	t
ة	Tāʾ marbūṭa	t~h	—
ث	Ṯāʾ	ṯ	ṯ~s̱
ج	Jīm	j
چ	Čē	—	č
ح	Ḥāʾ	ḥ
خ	Ḫāʾ	ḫ
د	Dāl	d
ذ	Ḏāl	ḏ	ḏ~ẕ
ر	Rāʾ	r
ز	Zāy	z
ژ	Žē	—	ž
س	Sīn	s
ش	Šīn	š
ص	Ṣād	ṣ
ض	Ḍād	ḍ	ż
ط	Ṭāʾ	ṭ
ظ	Ẓāʾ	ẓ
ع	ʿAyn	ʿ
غ	Ġayn	ġ
ف	Fāʾ	f
ق	Qāf	q
ک/ك	Kāf	k
گ	Gāf	—	g
ل	Lām	l
م	Mīm	m
ن	Nūn	n
ه/ھ	Hāʾ	h
و	Wāw	w	w~v
ی/ي	Yāʾ	y

Long vowels

Letter		ArArabic	FaPersian
ى/ا	ʾAlif	ā
و	Wāw	ū	ū~ō
ی/ي	Yāʾ	ī	ī~ē

Note that wāw is occasionally used as a short vowel in Persian.

The Arabic definite article

In the Arabic definite article, if before a "sun letter" (t, ṯ, d, ḏ, r, z, s, š, ṣ, ḍ, ṭ, ẓ, l, n), transliterate the lām instead as said consonant (e.g. ٱلرَّحْمَٰنُ ar-raḥmānu). In vocalized texts, this is represented by there being no diacritic over the lām and a šadda over the following consonant.

A long vowel becomes a short vowel if directly before the definite article, in which case there should be no macron in the transliteration.

Diacritics

Other than the hamza and madda, diacritics for pointing will not be present in most texts, and are almost never used for Persian. Note that diacritics are not universally written the exact same way. Romanize as if they were present. Diacritics here are presented as the diacritic applied to a taṭwīl.

Transliterate the following as letters after the pointed character; do not transliterate short vowels if they precede a long vowel.

Diacritic		ArArabic	FaPersian
ـَ	Fatḥa	a
ـُ	Ḍamma	u	u~o
ـِ	Kasra	i	i~e
ـٰ	ʾAlif ḫanjariyya	ā
ـً	Fatḥatān	an
ـٌ	Ḍammatān	un	un~on
ـٍ	Kasratān	in	in~en

Transliterate the hamza (⟨ـٔ⟩) as ⟨ʾ⟩ before the pointed character when it represents a glottal stop. Do not include a semivowel consonant (⟨w⟩ or ⟨y⟩) when used on such a letter; e.g. رُؤُوس is ruʾūs‎, not *ruʾwūs‎. When used in Persian for the iżāfa suffix, however, transliterate using the consonant ⟨y⟩.

Transliterate the šadda/tašdīd (⟨ـّ⟩) by writing the consonant twice. In Persian, in word-final position, instead of rendering the short vowel ⟨i⟩ followed by ⟨yy⟩ as ⟨iyy⟩, render this sequence as ⟨ī⟩, even in words of Arabic origin.

Transliterate the ʾalif with madda (⟨آ⟩) as ⟨ʾā⟩ in all positions in Arabic. For Persian, transliterate as ⟨ā⟩ in word-initial position instead.

Transliterate hamzat al-waṣl (marked in Qurʾānic vocalization as ⟨ٱ⟩ or as a bare ⟨ا⟩ in other texts) as an apostrophe (⟨'⟩) if the vowel is elided due to a preceding vowel, but as a vowel if it is pronounced (e.g. ⟨a⟩ in the definite article). If there is no ʾalif present at all in the spelling, do not include an apostrophe.

If a letter has a sukūn (⟨ـْ⟩), transliterate it only as the consonant with no following vowel.

Diphthongs

Render as a vowel followed by a semivowel. Short diphthongs differ in different varieties of Persian:

Diph.	ArArabic	FaPersian
ـَوْ	aw	aw~ow
ـَيْ/ـَیْ	ay	ay~ey

Note that, in Iranian Persian, ⟨aww⟩ is not considered to have a diphthong and should be rendered as ⟨avv⟩, but ⟨ayy⟩ should be rendered as ⟨eyy⟩.

Even in Iranian Persian, do not reduce diphthongs to a short vowel at word-final position (e.g. Ḫosrow, not *Ḫosro).

Hyphenation

Arabic

For Arabic, transliterate the definite article al, the conjunctive particles wa and fa, the interrogative particle a, and the single-letter prepositions bi, li, ta, and ka with a hyphen (⟨-⟩) between them and the word they modify. An exception is the definite article that exists as part of the word Allāh.

All other prefixes and suffixes not mentioned should not have a hyphen.

Persian

Put a hyphen before the iżāfa suffix (i.e. ⟨-i⟩ or ⟨-yi⟩) in Persian genitive constructions.

In Persian, some older texts do not put a space after the conjunction و wa/u, but it should be added, matching modern orthography.

For the Persian copula است ast, if the vowel is elided (represented orthographically by the ʾalif and space being removed), join it to the preceding word with a hyphen as -st.

When the preposition ba (be in Iranian Persian) lacks the hāʾ and is joined to the following word, separate it with a hyphen.

If elements of compound words are not connected where the letters normally would be (represented by a zero-width non-joiner in Unicode), represent this with a hyphen. The same applies for verbal prefixes and suffixes.

All other prefixes and suffixes not mentioned should not have a hyphen.

Pausal form

This section applies only for Arabic.

In the pausal form of a noun, the final short vowel and ⟨n⟩ (if applicable) of case endings will be omitted, with the following exceptions:

• The ending ⟨an⟩ in the indefinite accusative is retained.
• Short vowel case suffixes are retained before personal pronoun suffixes.
• If the nun of a case suffix is written as a full letter, do not delete it.

Tāʾ marbūṭa, instead of being transliterated as ⟨t⟩, will be transliterated as ⟨h⟩ after a long vowel and as nothing after a short vowel, unless the word is in the construct state, in which case it will still be transliterated as ⟨t⟩.

In Standard Arabic, pausal form is used at the end of a sentence. Colloquial Arabic will use pausal form in all cases. Note that in vocalized texts, pausal form is not indicated. Some sources will omit the vowel diacritic on the last letter when giving standalone nouns or noun phrases, but this is non-standard.

For this wiki, generally use the Standard Arabic rule of using pausal form at the end of a sentence or on the last word of a standalone noun phrase.

Miscellaneous

Replace punctuation and numerals with the Latin script equivalents.

For Arabic, on the second word of a compound preposition, the short vowel should be dropped if you are also transcribing every noun in pausal form instead of only at the end of a sentence. For example, use min qabli if using Standard Arabic rules, or min qabl if using colloquial rules.

For Persian, render the sequence ⟨ḫwa⟩ as ⟨ḫo⟩ for Iranian Persian and ⟨ḫu⟩ for Afghan Persian. Render the sequence ⟨ḫwā⟩ as ⟨ḫā⟩ for both Iranian and Afghan Persian.

Examples

The first, second, and third examples are nouns and noun phrases which have two romanizations given: the first includes Standard Arabic case endings for all words, and the second uses pausal form for the last word in the phrase (which is the recommendation for wiki purposes when referencing a word or name).

The fourth example is a whole Standard Arabic sentence where case endings are given for most words, but the last word's case ending is dropped as an instance of pausal form.

Arabic

Unpointed	Pointed	Romanization
العربية	ٱلْعَرَبِيَّةُ	Al-ʿarabiyyatu
		Al-ʿarabiyya
عيسى ابن مريم	عِيسَى ٱبْنُ مَرْيَمَ	ʿĪsā 'bnu maryama
		ʿĪsā 'bnu maryam
سلطنة مسقط وعمان	سَلْطَنَةُ مَسْقَطٍ وَعُمَانَ	Salṭanatu masqaṭin wa-ʿumāna
		Salṭanatu masqaṭin wa-ʿumān
لا إله إلا الله محمد رسول الله	لَا إِلَٰهَ إِلَّا ٱللَّٰهُ مُحَمَّدٌ رَسُولُ ٱللَّٰهِ	Lā ʾilāha ʾillā 'llāhu muḥammadun rasūlu 'llāh

These pointings below for Persian (reflecting Classical Persian pronunciation) are hypothetical and may be incorrect.

Persian

Unpointed	Pointed	Classical	Iranian
فارسی	فَارْسِی	Fārsī	Fārsī
خسرو انوشیروان	خُسْرَوْ اَنُوشِیرْوَان	Ḫusraw anōšērwān	Ḫosrow anūšīrvān
ممالک محروسهٔ ایران	مَمَالِکِ مَحْرُوسَهِٔ اِیراَن	Mamālik-i maḥrūsa-yi ērān	Mamālek-e maḥrūse-ye īrān
این نیز بگذرد	اِین نِیز بُگْذَرَد	Īn nīz bugzarad	Īn nīz bogzarad

Kurdish

Consonants

Letter	Translit.
ب	b
پ	p
ت	t
ج	c
چ	ç
ح	ẖ
خ	x
د	d
ر	r
ڕ	ṟ
ز	z
ژ	j
س	s
ش	ş
ع	ʿ
غ	x̱
ف	f
ڤ	v
ق	q
ک	k
گ	g
ل	l
ڵ	ḻ
م	m
ن	n
و	w
ھ	h
ى	y
ئ

Vowels

Letter	Translit.
	i
ا	a
و	u
وو	û
ۆ	o
ە	e
ى	î
ێ	ê

Greek (to 1453)

Letters

Letter		Translit.
α	Alpha	a
β	Beta	b
γ	Gamma	g~n
δ	Delta	d
ε	Epsilon	e
ζ	Zeta	z
η	Eta	ē
θ	Theta	th
λ	Lambda	l
μ	Mu	m
ν	Nu	n
ξ	Xi	x
ο	Omicron	o
π	Pi	p
ρ	Rho	r~rh
σ	Sigma	s
τ	Tau	t
υ	Upsilon	y~u
φ	Phi	ph
χ	Chi	ch
ψ	Psi	ps
ω	Omega	ō

Diacritics

Note that, for the digraphs ⟨αι⟩, ⟨ει⟩, ⟨οι⟩, ⟨υι⟩, ⟨αυ⟩, ⟨ευ⟩, ⟨ηυ⟩, ⟨ου⟩, and ⟨ωυ⟩, accent marks and breathing marks will be on the second letter.

Capital instances of where iota subscript would be used may occasionally be written as a normal iota after the capital letter. Romanize as if it were written with an iota subscript.

⟨◌⟩ in the table stands for any letter.

Diacritic		Translit.
◌́	Acute	◌́
◌̀	Grave	◌̀
◌̂	Circumflex	◌̂
◌̓	Smooth breathing
◌̔	Rough breathing	h◌~◌h
◌̈	Diaeresis	◌̈
◌ͅ	Iota subscript	◌i~◌̄i

Punctuation

Romanize the ano teleia (·) as a semicolon (;) and the coronis (᾽) as an apostrophe (').

Miscellaneous ancient West Semitic

Note that this does not exhaustively include every script.

PhnxPhoenician	UgarUgaritic	ArmiImperial Aramaic	SarbOld South Arabian	Transl.
𐤀		𐡀	𐩱	ʾ
	𐎀			a͗
𐤁	𐎁	𐡁‎	𐩨	b
𐤂	𐎂	‎𐡂	𐩴	g
	𐎃		𐩭	ḫ
𐤃	𐎄	𐡃‎	𐩵	d
𐤄	𐎅	𐡄‎	𐩠	h
𐤅	𐎆	𐡅‎	𐩥	w
𐤆	𐎇	𐡆‎	𐩸	z
𐤇	𐎈	𐡇‎	𐩢	ḥ
𐤈	𐎉	𐡈‎	𐩷	ṭ
𐤉	𐎊	𐡉‎	𐩺	y
𐤊	𐎋	𐡊‎	𐩫	k
𐤋	𐎍	𐡋‎	𐩡	l
𐤌	𐎎	𐡌‎	𐩣	m
	𐎏		𐩹	ḏ
𐤍	𐎐	𐡍‎	𐩬	n
	𐎑		𐩼	ẓ
𐤎	𐎒	𐡎‎	𐩪	s
			𐩯	ś
𐤏	𐎓	𐡏‎	𐩲	ʿ
𐤐	𐎔	𐡐‎	𐩰	p
𐤑	𐎕	𐡑‎	𐩮	ṣ
𐤒	𐎖	𐡒‎	𐩤	q
𐤓	𐎗	𐡓‎	𐩧	r
𐩦	𐎌	𐡔‎	𐩦	š
	𐎘		𐩻	ṯ
	𐎙		𐩶	ġ
𐤕	𐎚	𐡕‎	𐩩	t
	𐎛			ı͗
	𐎜			u͗
	𐎝			s̀

Mongolian

This section is for rules on transliterating Mongolian written in the Mongolian script (which is to be labeled "Classical Mongolian" on this wiki). For details on Cyrillic script (which is to be labeled "Khalkha Mongolian" on this wiki), please look at THL's website.

Note that while several vowels look identical to one another, they are encoded separately in Unicode and they can be inferred based on vowel harmony.

Vowels

Mongolian	V-M
ᠠ	a
ᠡ	e
ᠢ	i
ᠣ	o
ᠤ	u
ᠥ	ö
ᠦ	ü
ᠧ	ē

Some consonants look identical in at least some contexts (and may be distinguished depending on position in a word), but are encoded separately in Unicode. Some other consonants exist but are rare and not listed here.

Consonants

Mongolian	V-M
ᠨ	n
ᠩ	ng
ᠪ	b
ᠫ	p
ᠬ	q~k
ᠭ	ɣ~g
ᠯ	l
ᠮ	m
ᠰ	s
ᠱ	š
ᠲ	t
ᠳ	d
ᠴ	č
ᠵ	ǰ
ᠶ	y
ᠷ	r
ᠸ	w
ᠹ	f
ᠺ	g
ᠻ	k
ᠼ	c
ᠽ	j
ᠾ	h
ᢊ	ñ
ᠿ	ž
ᡀ	lh
ᡁ‍	zh
ᡂ‍	ch

The following digraphs exist. Do not include the medial ⟨y⟩ for any other combination of vowels.

Consonants

Mongolian	V-M
ᠠᠢ	ayi
ᠡᠢ	eyi
ᠣᠢ	oyi
ᠤᠢ	uyi

Old Church Slavonic and Old East Slavic

Cyrillic	Translit.
а	a
б	b
в	v
г	g
д	d
е~є	e
ж	ž
ѕ~ꙃ	dz
з~ꙁ	z
и~і~ї	i
к	k
л	l
м	m
н	n
о~ѡ	o
п	p
р	r
с	s
т	t
оу~ꙋ~oѵ~у	u
ф~ѳ	f
х	h? x?
ѿ	otŭ
ц	c
ч	č
ш	š
ъ	ŭ
ꙑ	y
ь	ĭ
ѣ	ě
ꙗ	ja
ѥ~є	je
ю	ju
ѫ	ǫ
ѧ	ę
ѭ	jǫ
ѩ	ję
ѯ	ks
ѱ	ps
ѵ	i~v

Ukrainian

Cyrillic	Translit.
а	a
б	b
в	v
г	h
ґ	g
д	d
е	e
є	ye
ж	zh
з	z
и	y
і	i
ї	yi~ï
й	ĭ~y
к	k
л	l
м	m
н	n
о	o
п	p
р	r
с	s
т	t
у	u
ф	f
х	kh
ц	ts
ч	ch
ш	sh
щ	shch
ь	’
ю	yu
я	ya
’	”

Put a middle dot ⟨·⟩ between two letters if necessary to distinguish from digraphs.

Examples

Ukrainian	Romanization
Український	Ukraïns’kyĭ
Тарас Григорович Шевченко	Taras Hryhorovych Shevchenko
Галицько-Волинське князівство	Halyts’ko-Volyns’ke knyazivstvo
За Карпати відіб’ється, згомонить степами, України слава стане поміж народами.	Za Karpaty vidib”yet’sya, z·homonit’ stepamy, Ukraïni slava stane nomizh narodamy.

Urdu

Consonants

Letter	Romanization
ا/ء	ʾ
ب	b
پ	p
ت	t
ٹ	ṭ
ث	s̱
ج	j
چ	c
ح	ḥ
خ	ẖ
د	d
ڈ	ḍ
ذ	ẕ
ر	r
ڑ	ṛ
ز	z
ژ	ž
س	s
ش	š
ص	ṣ
ض	z̤
ط	t̤
ظ	ẓ
ع	ʿ
غ	ġ
ف	f
ق	q
ک	k
گ	g
ل	l
م	m
ن	n
ں	ṉ
و	v
ہ/ھ	h
ى	y

Vowels

Short ⟨a⟩, ⟨i⟩, and ⟨u⟩ are unwritten, but should be included in the romanization regardless. The same letters can stand for multiple different vowels (and و and ی can also be used for consonants).

Letter	Romanization
و	ū, o, au
ی	ī, ai
ے	e