---
authority_id: bgnpcgn
id: 2007
language: iso-639-2:fas
source_script: Arab
destination_script: Latn
name: TRANSLITERATION SYSTEM FOR PERSIAN BGN/PCGN 1958 System
url: https://github.com/interscript/interscript-ruby/files/5180821/BGN_Romanization_Guide_1964_persian_1958.pdf
creation_date: 1958
confirmation_date: 2017-11
description: |
The BGN/PCGN 1958 System for the Persian alphabet was
adopted by the BGN in 1946 and by the PCGN In 1958. It Is
used for the romanization of standard geographic names in
Iran.
The Persian alphabet is the same as the Arabic alphabet
except for the addition of four consonant symbols for
sounds which do not occur in Arabic. Therefore, the same
conversion values are used for the Persian alphabet as for
the Arabic alphabet wherever phonetically valid. Where
letters common to the two alphabets have different
pronunciation in the two languages, conversion values
reflecting the Persian pronunciation are provided in the
Persian system.
It may be noted that neither the system of pointing nor the
conversion values for vowels and diphthongs are the same
for Persian as for Arabic.
Persian presents the same problems of vowel pointing as
Arabic in that the short vowel symbols necessary for the
proper rendering of short vowels, long vowels, and
diphthongs are almost always omitted from written Persian.
It follows that, in order to produce transliter- tions in
terms of standard written Persian, the transliterator must
be able to identify the words used in names and must know
their standard written Persian spelling, their proper vowel
pointing, and how to eliminate pecularities due to
dialectical variation.
The notes and special rules explain details of
transliteration not stated in the tables.
notes:
- |
1)The Persian writing system, in its generally used form,
employs only the symbols found in Table I, and the alef maddeh
(آ) of Table II. The symbols in column 1 of Table II are
traditional auxiliary symbols used by Persians only
sporadically to indicate the exact reading of a given word.
In the case of i, ū, ey and ow, a choice of symbology
exists, one system adhering to the Arabic tradition, the
other being a modification of it. Likewise, there are two
ways of writing alef maq§ureh. The BGN/PCGN transliteration
system presupposes forms fully pointed with the Persian
symbols represented in Table II, though these are omitted
from the examples given below.
- 2) Alef ( أ ) occurs with the following uses
a. Initially, it indicates that the word begins with a vowel or
diphthong, and the alef itself is not transliterated, e.g.,
انجيرة Anjīreh, ايواني Eyvānī
b. Initially with maddeh ( آ ),it indicates ā, e.g., آبادان Ābādān.
c. Medially and finally. it represents ā, e.g., e.g., آبادان Ābādān,
توكا Tūkā.
d. With hamzeh. See Note 5.
- 3) Where special considerations are paramount, the sub-dot
may be used in place of the cedilla
- 4) The two-letter symbols kh, zh, sh, and gh are convenient
for use and easily comprehensible. The infrequent sequences
of kāf, ze, sīn, and gāf followed by he can be unequivocally
transliterated for bibliographical purposes by the use of a
slant line (“shilling,” “solidus") to separate true h from
a preceding k, z, s, or g; that is, kh represents khe, but
k/h stands for kāf followed by he.
- 5) Hamzeh (ء) is not regarded as a letter of the Persian
alphabet, but as a diacritic mark, and as such is not
always expressed in actual writing. In fully pointed words,
however, it appears in several graphic forms, standing
alone or written in conjunction with alef, vav, or ye (the
ye in this instance being undotted). All these forms are
transliterated simply with the apostrophe in the form
resembling a small raised figure 9 (’) ,e.g., جزء Joz’
صفراء şafrā’, مأخذ ma’khaz̄, مؤتمن mu’taman, پائين pā’īn Occasionally a word may
appear pointed according to the strict Arabic tradition so
as to indicate an initial hamzeh. This initial hamzeh is
never transliterated, but only the vowel which it carries,
e.g., •Jfi , whether pointed or 0 . is "transliterated
simply Anjireh. ' Hamzeh is also sometimes written over
final he or ye to represent the -ye form of the eg&feh, cf.
Note 8.
- 6) Alef maddeh is used initially to indicate ā, e.g., آبادان Ābādān
Noninitially it indicates ’ā in words of Arabic origin,
and must be so transliterated, e.g., قرآن Qor’ān, مآب ma’āb
- 7) Tashdid ( ّ) indicates doubling of the consonant over which
it is written and is represented in transliteration by writing
the consonant twice, e.g محمّد Mohammad The sequence kasreh ye ye
is transliterated īy.In names of Arabic origin containing the
definite article, the vowel of the article is transliterated according
to its Persian pronunciation and the lam, when followed by a “sun letter”
(t, s̄, d, z̄, r, s, sh, ẕ, ţ, z̧, ţ, z̧, l, n) is assimilated. The article
is written in lower case and separate from the preceding and following word,
e.g., Zeyn od Din, but *£4 . Zeyn ed Dīn. Cf. Special Rule 1.
- 8) The relational (ezafeh) suffix is transliterated -e after final consonants
(except for silent h). After silent h and after vowels, it is transliterated
-ye. It is usually not expressed in Persian writing after a consonant
(kasreh being understood), e.g., كوهِ مَرغٌوب Kuh-e Marghūb. After final alef or vav
it is written with ye (ي), e.g., بايِ آب Pa-ye Ab, جويِ آس Jū-ye Ās. After ye (ي)
and silent he it is written as hamzeh over the letter, e.g., دَهَانهٴ مَمبَر Dahāneh-ye Mambar,
سَلَيٴ بُذٌرگ Salasi-ye Bozorrg.
- 1. In Persian names of Arabic origin, the word division
used for transliterating Arabic will obtain, except that
names ending in the word Allah will be written solid and
the a of Allah will be replaced by o, e.g., The Arabic
definite article when not subject to assimilation will be
written al in name-initial position but usually, ol
elsewhere, e.g., Al JJal but 2u ol Faqar. Cf. Note 7 above.
- 2. In Persian names in general, in sequences of name
elements pronounced as a single unit, division into
separate words will be made after Persian letters written
in the word final form. (But see the Rule immediately
below.)
- 3. The word abad which occurs very frequently at
the end of place names will be written solid with the
preceding word; e.g., Allahabad, not Allah Abad.
- 4. Persian derivational endings such as vand and Turkish and other
endings such as lar, li and lu will be written solid with
the preceding word, whether so written in Persian script or
not.
- 5. The letter he, when used in Turkish fashion for
word internal e, will be transliterated eh as usual but
will be written solid with the remainder of the word.
- 6. Kheplus vdv followed by long vowel or khe plus vdv followed
by gammeh will be rendered khv although the v is not
pronounced.
tests:
- source: اَنجِيرة
expected: Anjīreh
- source: اِيْوَانِي
expected: Eyvānī
- source: آبادان
expected: Ābādān
- source: تُوكا
expected: Tūkā
- source: آبادان
expected: Ābādān
- source: قُرآن
expected: Qor’ān
- source: مَآب
expected: Ma’āb
- source: مُحَمَّد
expected: Moḩammad
- source: كُوهِ مَرغُوب
expected: Kūh-e Marghūb
- source: پَايِ آب
expected: Pā-ye Āb
- source: جُويِ آس
expected: Jū-ye Ās
- source: دَهَانِهٴ مَمبَر
expected: Dahāneh-ye Mambar
- source: سَلَسِيٴ بُذُرگ
expected: Salasī-ye Boz̄org
- source: عَبداللَّه
expected: ‘Abdollāh
- source: ذُو الفَقَار
expected: Z̄ū ol Faqār
- source: اللَّه آبَاد
expected: Allāhābād
- source: اِيران
expected: Īrān
map:
postrules:
- pattern: (?<=\b)(?<!\b[‘|’|'|-])[\u0061-\uFFFF]
result: "upcase"
- pattern: " Ol"
result: " ol"
characters:
'\s\u0622\u0628\u064E\u0627\u062F' : 'ābād' # Special Rule 3
'\u064e(?=\u0629)' : '' # َ fatha followed by ta' marboota
'\u064e(?=a[h|t])' : '' # َ fatha followed by ta' marboota, handling different order of conversion
# pointing
'\u064e' : 'a' # َ fatha
'\u0650' : 'e' # ِ kasra
'\u064f' : 'o' # ُ damma
'\u064e\u0627' : 'ā' # ـَا fatha followed by ا
'\u0649\u0670' : 'á' # ىٰ
'\u0622' : '’ā' # آ
'\b\u0622' : 'ā' # آ
'\u0650[\u064a|\u06cc]' : 'ī' # ـِي kasra followed by ي
# '\u0650\u06cc' : 'ī' # ـِي kasra followed by ي
'\u064f\u0648' : 'ū' # ـُو damma followed by و
'\u064e[\u064a|\u06cc]\u0652' : 'ey' # ـَيْ
'\u0650[\u064a|\u06cc]\u0652' : 'ey' # ـِيْ
'\u064e\u0648' : 'ow' # ـَو
'\u0652' : '' # ْ sokoon
'\u0650\u064a\u0651' : 'īy' # ـِيّ
'[\u0654|\u0674]' : '-e' # ٴ ezafeh
'(?<=[\u064a|\u0647])[\u0654|\u0674]' : '-ye' # ٴ ezafeh
'\u0650\b' : '-e' # ِ kasra
'[\u064a|\u06cc]\u0650\b' : '-ye' # ِ kasra
# shadda
'\u0628\u0651' : 'bb' # ب
'\u062a\u0651' : 'tt' # ت
'\u062b\u0651' : 's̄s̄' # ث
'\u062c\u0651' : 'jj' # ج
'\u062d\u0651' : 'ḩḩ' # ح
'\u062e\u0651' : 'kh' # خ
'\u062f\u0651' : 'dd' # د
'\u0630\u0651' : 'z̄z̄' # ذ
'\u0631\u0651' : 'rr' # ر
'\u0632\u0651' : 'zz' # ز
'\u0633\u0651' : 'ss' # س
'\u0634\u0651' : 'sh' # ش
'\u0635\u0651' : 'şş' # ص
'\u0636\u0651' : 'ẕẕ' # ض
'\u0637\u0651' : 'ţţ' # ط
'\u0638\u0651' : 'z̧z̧' # ظ
'\u063a\u0651' : 'gh' # غ
'\u0641\u0651' : 'ff' # ف
'\u0642\u0651' : 'qq' # ق
'\u0643\u0651' : 'kk' # ك
'\u0644\u0651' : 'll' # ل
'\u0645\u0651' : 'mm' # م
'\u0646\u0651' : 'nn' # ن
'\u0647\u0651' : 'hh' # ه
'\u0648\u0651' : 'vv' # و
'\u064a\u0651' : 'yy' # ي
'\u0659': 'ê'
# ta' marboota
'\u0629' : 'eh'
'\b\u0627\u0644' : 'al ' # ال
'\s\b\u0627\u0644' : ' ol ' # ال #special Rule 1
'\ufdf2': 'Allāh' # See note 5
'\u0627\u0644\u0644\u0651\u064e\u0647' : "Allāh"
'(?<!\b)\u0627\u0644\u0644\u0651\u064e\u0647' : "ollāh" # Special Rule 1
# '\uFE8E' : '' # ﺎ
# Sun letters
'\b\u0627\u0644\u062a' : 'at t' # الت
'\b\u0627\u0644\u062b' : 'as̄ s̄' # الث
'\b\u0627\u0644\u062f' : 'ad d' # الد
'\b\u0627\u0644\u0630' : 'az̄ z̄' # الذ
'\b\u0627\u0644\u0631' : 'ar r' # الر
'\b\u0627\u0644\u0632' : 'az z' # الز
'\b\u0627\u0644\u0633' : 'as s' # الس
'\b\u0627\u0644\u0634' : 'ash sh' # الش
'\b\u0627\u0644\u0635' : 'aş ş' # الص
'\b\u0627\u0644\u0636' : 'aẕ ẕ' # الض
'\b\u0627\u0644\u0637' : 'aţ ţ' # الط
'\b\u0627\u0644\u0638' : 'az̧ z̧' # الظ
'\b\u0627\u0644\u0644' : 'al l' # الل
'\b\u0627\u0644\u0646' : 'an n' # الن
'\u0621': '’' # ء
'\u0626': '’' # ئ
'\u0623' : '' # أ
'\u0625' : '' # إ
'\u0627' : 'ā' # ا
'\b\u0627' : '' # ا
# consonant characters
'\u0628' : 'b' # ب
'\u067E' : 'p' # پ
'\u062a' : 't' # ت
'\u062B' : 's̄' # ث
'\u062c' : 'j' # ج
'\u0686' : 'ch' # چ
'\u062d' : 'ḩ' # ح
'\u062e' : 'kh' # خ
'\u062f' : 'd' # د
'\u0630' : 'z̄' # ذ
'\u0631' : 'r' # ر
'\u0632' : 'z' # ز
'\u0698' : 'zh' # ژ
'\u0633' : 's' # س
'\u0634' : 'sh' # ش
'\u0635' : 'ş' # ص
'\u0636' : 'ẕ' # ض
'\u0637' : 'ţ' # ط
'\u0638' : 'z̧' # ظ
'\u0639' : '‘' # ع
'\u063a' : 'gh' # غ
'\u0641' : 'f' # ف
'\u0642' : 'q' # ق
'\u0643' : 'k' # ك
'\u06A9' : 'k' # ک
'\u06AF' : 'g' # گ
'\u0644' : 'l' # ل
'\u0645' : 'm' # م
'\u0646' : 'n' # ن
'\u0647' : 'h' # ه
'\u0648' : 'v' # و
'\u064a' : 'y' # ي
'\u0649' : 'y' # ي
'\u06D0' : 'ē' # ې
'\u06CD' : 'êy' # ۍ
This was never merged but some work was done on it. We need to check this against the implemented 1956 system.
The specification is provided by the BGN Romanization Guide published in 1964, attached here. BGN_Romanization_Guide_1964_persian_1958.pdf