The romanisation schemes

Each scheme is a versioned table written from its published standard. They are for display: rendering a script for a reader in the form a catalogue, a citation, a map or a passport would use. For matching across them, use the fold.

tetrak-translit schemes prints every scheme with its current version.

Armenian

Id	Scheme	Standard	Where you meet it	Reversible
hy/ala_lc	ALA-LC	ALA-LC Romanization Tables, Armenian (2011)	Library catalogues: Koha, WorldCat, the Library of Congress	no
hy/iso_9985	ISO 9985	ISO 9985:1996	Data exchange; one letter per letter	yes
hy/bgn_pcgn	BGN/PCGN	BGN/PCGN 1981 System	Maps and place names: Yerevan, Gyumri	no
hy/hubschmann_meillet	Hübschmann-Meillet	Hübschmann (1897), Meillet (1913)	Linguistics and classical Armenian scholarship	yes
hy/western	Western Armenian pronunciation	Common practice, not a standard	Diaspora names: Hagop, Kevork, Boghosian	no

The same name five ways:

Armenian	ala_lc	iso_9985	bgn_pcgn	hubschmann_meillet	western
Երևան	Erevan	Erewan	Yerevan	Erewan	Yerevan
Երեւան (classical)	Erewan	Erewan	Yerevan	Erewan	Yerevan
Յակոբ	Yakob	Yakob	Yakob	Yakob	Hagop
Գէորգ (classical)	Gēorg	Gēorg	Georg	Gēorg	Kevork
Պօղոսեան (classical)	Pōghosean	Pòġosean	Poghosean	Pōłosean	Boghosian
Չարենց	Chʻarentsʻ	Čarenc’	Ch’arents’	Čʿarencʿ	Charents
Ոսկան	Oskan	Oskan	Voskan	Oskan	Vosgan

ALA-LC

The Library of Congress table has no positional rules: ե is e and ո is o in every position, which is why the Library of Congress writes Erevan and Oskan where BGN/PCGN writes Yerevan and Voskan. Aspiration is marked with the modifier letter turned comma ʻ (U+02BB), the character catalogue records actually carry. The ligature և is ev but the letter ւ is w, so a classical-orthography imprint spelling the city Երեւան romanises as Erewan: a real distinction in catalogue records, not an error, and one the fold equates. Lossy: ts is ծ or տ+ս, dz is ձ or դ+զ, kh is խ or կ+հ, and so on.

ISO 9985

The international standard, designed to reverse: every letter is distinct, by diacritic where the plain letters run out. Aspiration is a plain apostrophe; ճ is č with a dot below and չ is plain č; ը is ë, օ is ò, ու is ow. Any apostrophe-like character (', ’, ʻ, ʿ) is accepted on input when reversing.

BGN/PCGN

The system behind the English spellings of Armenian place names. It is the only Armenian standard here with positional rules: ե is ye and ո is vo at the start of a word, and և is yev there and ev elsewhere. ռ is rr, ը is y. Written for reformed orthography; a classical spelling goes through it letter by letter without the Western sound values. Lossy: the same digraph ambiguity as ALA-LC, and y is both ը and յ.

Hübschmann-Meillet

The transcription of Armenian studies, from Hübschmann’s Armenische Grammatik and Meillet’s Altarmenisches Elementarbuch, as the Revue des Études Arméniennes uses it: x for խ, ł for ղ, c for ծ, j for ձ, and the aspirates marked with ʿ (U+02BF). One letter per letter, so it reverses. ու is u, which would clash with ո+ւ, but reformed orthography never writes a free ւ after ո and the classical digraph is the same two letters.

Western Armenian pronunciation

Not a standard but the practice: how names were spelled by sound in Istanbul, Beirut, Aleppo, Paris and Los Angeles. Western Armenian swapped the stops, so բ is p and պ is b, գ is k and կ is g, դ is t and տ is d, and the affricates likewise. Word-initial յ is h (Hagop), ո after a vowel is vo (Kevork), and the classical digraphs եա and իւ are ia and iu (Vartanian, Harutiun). ը is written e; many real spellings omit it (Mkrtich), which no table can predict. Never meant to be reversed: it cannot say whether a p is բ or փ.

Georgian

Id	Scheme	Standard	Where you meet it	Reversible
ka/ala_lc	ALA-LC	ALA-LC Romanization Tables, Georgian (2011)	Library catalogues	no
ka/iso_9984	ISO 9984	ISO 9984:1996	Data exchange; one letter per letter	yes
ka/national	Georgian national system	Adopted 2002; BGN/PCGN 2009	Road signs, passports, maps: Tbilisi, Kutaisi	no

The same name three ways:

Georgian	ala_lc	iso_9984	national
თბილისი	tʻbilisi	t’bilisi	tbilisi
ქუთაისი	kʻutʻaisi	k’ut’aisi	kutaisi
ჭავჭავაძე	chavchavadze	čavčavaje	ch’avch’avadze
წერეთელი	tseretʻeli	ceret’eli	ts’ereteli
მცხეთა	mtsʻkhetʻa	mc’xet’a	mtskheta
ყაზბეგი	qazbegi	qazbegi	q’azbegi

Mkhedruli has no capitals, so a Georgian rendering is lowercase; a catalogue capitalises proper names by its own convention. Mtavruli, the capital forms Unicode added in 2018, renders as capitals.

The apostrophe runs the opposite way

Georgian has pairs of aspirated and ejective consonants (თ/ტ, ფ/პ, ქ/კ, ჩ/ჭ, ც/წ). ALA-LC and ISO 9984 mark the aspirate with the apostrophe (tʻbilisi, kʻutʻaisi) and leave the ejective plain. The national system marks the ejective (ts'ereteli, ch'avch'avadze) and leaves the aspirate plain. The same apostrophe on the same word therefore means opposite letters in the two systems, which is the commonest source of confusion between them, and the reason the fold strips the mark and merges each pair.

ALA-LC

Aspirates with ʻ (U+02BB), ejectives plain. Defines the five archaic letters (ჱ ē, ჲ y, ჳ w, ჴ ẖ, ჵ ō). Lossy: dz, zh, gh, sh and kh can each be one letter or two.

ISO 9984

One letter per letter, so it reverses: háčeks for the sibilants and affricates (ž š č ǰ), x for ხ, ḡ for ღ, j for ძ, c for წ, and a plain apostrophe on the aspirates. Defines the archaic letters as ALA-LC does.

The national system

The 2002 system of the Georgian State Department of Geodesy and Cartography, adopted by BGN/PCGN in 2009: no diacritics, apostrophe on the ejective. The spelling on signs, passports and maps. Does not define the archaic letters, which pass through unchanged. Lossy.

Versioning

Each scheme carries its own version, separate from the package’s. A change to a mapping bumps it, so a stored string can say which table produced it. Every table names the standard it was written from, and a pull request changing one must cite the standard too; nothing is ever copied from another implementation.