Bāla’y-balan

Benjamin Trigona-Harany

March 9, 2018

This work is licensed under a Creative Commons Attribution-NonCommercial-

ShareAlike 3.0 Unported License. ⃝BY: ⃝c
$ ⃝
\
1 Introduction
Bāla’y-balan is the name of the language created by the Ottoman mystic Mehmed
ibn Fethullah ibn Ebü Tâlib in the mid-16th century and described in his Kitâb-ı
Bâleybelen [Book of Bāla’y-balan].¹ The author, who is more commonly known
as “Muhyî-i Gülşenî”, was born in the Balkan city of Edirne to a family that
traced its origins to Shiraz in Persia. Most of Muhyî’s life, however, was spent
in Egypt, where he was an active member of the Gülşenî sufi order.² Muhyî’s
mysticism was the motivation for the creation of Bāla’y-balan, which he en-
visioned as a language of devotion rather than a regional lingua franca or a
replacement for the languages spoken in the Ottoman Empire. Like his native
Ottoman, Muhyî drew on elements of Turkish, Arabic and Persian when creat-
ing Bāla’y-balan and elements of each are recognisable in the lexicography and
morphology of his language.
Presently, the only monograph on Bāla’y-balan is Bâleybelen: İlk Yapma Dil
by Mustafa Koç (Istanbul: 2005). Besides providing a transcription of all the
Bāla’y-balan texts, Koç includes orignal research on Muhyî’s life, the Gülşenî
order, the creation of Bāla’y-balan and references to it in contemporary texts.
Koç also provides his own summary of the language’s grammar and has com-
piled the glosses from the orignal manuscripts into a 10,000-word Bāla’y-balan–
Turkish lexicon and its Turkish–Bāla’y-balan inverse.
The name Bāla’y-balan is formed from the noun bāl ‘language’ and the in-
finitive balam ‘to give life; to revive’. The linking suffix -a is attached to the
governing noun bāl, giving bāla. The definite article y- precedes the goverened
noun balan, which is formed from the verb stem bal- plus the active participle
suffix -an, creating the word ‘the one who gives life; the reviver’. As a whole
¹Bāla’y-balan transcribed into the Latin alphabet will appear in bold face. Arabic, Persian and
Turkish will all appear in italics using the standard transcription system for Ottoman. Square
brackets indicate IPA notation. English equivalents for these words will appear in single quotes
when it is necessary to distinguish them from the rest of the text: ‫ ظفم‬ẓafam [zafam] yazmak
‘to write’. Unattested and hypothetical examples of Bāla’y-balan will be prefixed with the
symbol §.
²For more on the Gülşenî movement, see John Curry’s Transforming Muslim mystical thought
in the Ooman Empire.

1
Bāla’y-balan translates as ‘the language of the one who gives life’ or simply ‘the
language of the reviver’. The Arabic word muhyî (deriving from the Arabic root
‫حي‬
ّ ) also translates as balan, so an alternate reading of Bāla’y-balan could be
simply ‘the language of Muhyî’.

2 Phonology and Orthography

Bāla’y-balan is written using the modified Arabic script also employed for Ot-
toman Turkish and Persian. Because Muhyî did not describe how his language
was to be pronounced in Kitâb-ı Bâleybelen, we must rely uniquely on orthog-
raphy, and it is, therefore, not possible to reconstruct the ponetics of Bāla’y-
balan with complete certainty. However, I have done my best to determine
how he probably intended Bāla’y-balan to be pronouced by evaluating Muhyî’s
decisions with respect to his language’s orthography. The following section
describes how I came up with my proposed phonology of Bāla’y-balan; readers
not interested in the rationale for my decision can skip directly to the section
on the Bāla’y-balan alphabet.

2.1 Ottoman Turkish Orthography

Orthographically, Bāla’y-balan is heavily influenced by Ottoman Turkish, and
the modified Arabic alphabet used to write it.³ While Arabic orthography dis-
tinguishes between three short and three long vowels, Turkish has four back
vowels (a, ı, o, u) and four front vowels (e, i, ö, ü). To accommodate these extra
sounds, Ottoman Turkish makes use of the fact that Turkish phonology does
not distinguish between certain consonant pairs found in Arabic; the letters ‫س‬
and ‫ص‬, for example, are both pronounced [s] in Turkish. When appearing in
Turkish words, however, the letters show whether surrounding vowels should
be pronounced in their front or back form. Thus, the conditional suffixes -sa
and -se were written using the Arabic consonants ‫ ص‬and ‫ س‬respectively, where
³The same changes to the standard Arabic script are also present in Persian.

2
only these letters indicate how the following vowel should be pronounced.⁴ The
consonant pairs which serve this function in Ottoman Turkish were ‫د‬-‫ط‬, ‫ت‬-‫ط‬
and ‫س‬-‫ص‬. Two more, ‫ك‬-‫ ق‬and ‫گ‬-‫غ‬, indicate a pronunciation difference in the
consonant in addition to the vowel: for example, ‫ ك‬is [c] while ‫ ق‬is [k]. The
words ‫ اولمك‬and ‫ اولمق‬differ only in their final consonant but are pronounced
ölmek and olmak respectively.
There are additionally numerous Arabic letters which are never used to
write Turkish words and whose pronunciation is identical to a native Turk-
ish consonant. For example, the letters ‫ح‬, ‫ خ‬and ‫ ه‬are all pronounced [h], while
only the latter appears in words of Turkish origin. Similarly, the letters ‫ز‬, ‫ذ‬, ‫ض‬
and ‫ ظ‬are all realised [z] in Turkish. A number of these consonants also influ-
ence the pronunciation of neighbouring vowels, but since these only appear in
loan words these sound changes do not have any systematic semantic meaning.
Muhyî appears to have explicitly avoided using consonants which are not
native to the Turkish language in function words and affixes (which are used
extensively in Bāla’y-balan for syntactic purposes). Consonants which are used
in Ottoman Turkish orthography only to mark vowel pronunciation are used,
but only in lexical words (nouns, verbs, adjectives and adverbs). Certain con-
sonants, namely ‫ذ‬, ‫ ض‬and ‫ظ‬, are rarely encountered at all in the language.⁵
The purpose of eliminating these letters from grammatical suffixes an function
words appears to be to either mimic Turkish usage or to eliminate confusion
between phonemes due to their identical pronunciation. If these letters were
given their full Arabic values, there would be no reason to restrict their usage
in Bāla’y-balan.
Further evidence comes in the letter names Muhyî created for Bāla’y-balan
since the standard Arabic names of the letters would have been ambiguous:
‫حاء‬, ‫ خاء‬and ‫ هاء‬are all hâ in Turkish. Assuming that all three letters were pro-
nounced identically in Bāla’y-balan, each would need a distinct name, which is
exactly what Muhyî did: [ħaː], [xav] and [hir] respectively. The same is done
for all other consonant groups ([d], [s], [t] and [z]) where confusion would
⁴An additional Arabic consonant, ‫ث‬, is also pronounced [s] in Turkish but appears only in
Arabic loan words and does not influence the pronunciation of surrounding vowels.
⁵These letters are also some of the rarest in Arabic.

3
have occurred. Note that since Turkish distinguishes between [c] and [k], the
letters ‫ ك‬and ‫ ق‬do not require this distinction and are known as qin and kin
respectively.⁶
It is also reasonable to assume that Muhyî would select consonants which
are more readibly producable by speakers of Persian, Turkish and other lan-
guages which do not share the full range of Arabic consonantal phonology.
Many of the sounds which I argue have been excluded from Bāla’y-balan would
not even have been pronounced identically in the various dialects of Arabic
spoken in the Ottoman Empire.
Since Muhyî has included consonants which, in Ottoman Turkish, influence
the pronunciation of vowels in the same word, the second question is whether
Bāla’y-balan has all eight vowels of Ottoman, the same three present in Arabic
or some combination thereof. In my transcription, I have opted to only use the
short [a], [u] and [i] and long [aː], [uː] and [iː] that can be definitively elicited
from Muhyî’s writing. I have done this with the caveat that he may have in-
tended the short vowels in Bāla’y-balan to be influenced by neighbouring con-
sonants. Under normal circumstances, the Arabic short vowels [a], [u] and [i]
should be read [e], [i] and [y] in Ottoman; these usually change to [a], [ɯ] and
[u] under the influence of the following consonants: ‫ح‬, ‫خ‬, ‫ص‬, ‫ض‬, ‫ط‬, ‫ظ‬, ‫ع‬, ‫غ‬
and ‫ق‬. The sounds [o] and [ø] are usually only present in the initial syllable of
Turkish words, and I have not used them in my Bāla’y-balan phonology.
Words in Bāla’y-balan are not required to have vowel harmony, a distin-
guishing feature of the Turkish language.⁷ Words such as qaydak kapı ‘door,
gate’, which has both a back q and a front k, would not be possible in a pure
Turkish word. This point has further significance when we consider Bāla’y-
balan’s suffixes, which will not harmonise with the vowels in the root to which
they are attached.
⁶This distinction is not reflected in the spelling system of modern Turkish, in which both are k.
Azeri Turkish, and other Turkic languages of the former Soviet Union that have adopted the
Latin alphabet, do make use of q to distinguish the two sounds in writing.
⁷The definite article y- does in fact harmonise with the following word, but this is unlike the
harmony found in Turkish. See section ⁇.

4
e Alphabet
Because Bāla’y-balan’s phonetics are uniquely tied to its orthography, the latter
will be presented first. There are 34 letters proper used in the language, or 35 if
we consider amam a letter in its own right. One letter, anā is effectively used
as a vowel marker only, and two more, var and yam, can be used as either
consonants or vowels.
In the transcription, a macron over a vowel indicates lengthening. Other
letters with diacritics may or may not represent a pronunciation difference be-
tween them and their base form; refer to the IPA column for guidance on pro-
nunciation.
Each letter in the Arabic alphabet may have up to four different forms de-
pending on its position in the word (initial, medial, final or isolated). Only the
isolated form is presented in the table below. Letters in Bāla’y-balan’s alphabet
also have numerical values which are used to write numbers, as explained in
section ⁇.
Transcription Name Arabic IPA
ā anā ‫ا‬ [aː]
ˀ amam ‫ء‬ [ʔ]
b bar ‫ب‬ [b]
p pī ‫پ‬ [p]
t tir ‫ت‬ [t]
ŝ sam ‫ث‬ [s]
c cā ‫ج‬ [dʒ]
ç çī ‫چ‬ [tʃ]
ḥ ḥā ‫ح‬ [h]
x xav ‫خ‬ [h]
d dav ‫د‬ [d]
ẕ ẕav ‫ذ‬ [z]
r ran ‫ر‬ [ɾ]
z zan ‫ز‬ [z]
j jī ‫ژ‬ [ʒ]

5
 s sir ‫س‬ [s]
ş şir ‫ش‬ [ʃ]
ṣ ṣad ‫ص‬ [s]
ż żad ‫ض‬ [d]
ṭ ṭī ‫ط‬ [t]
ẓ ẓī ‫ظ‬ [z]
ˁ ˁī ‫ع‬ [ʔ]
ġ ġī ‫غ‬ [ɡ]
f fin ‫ف‬ []
q qin ‫ق‬ [k]
k kin ‫ك‬ [c]
g gī ‫گ‬ [ɟ]
ñ gim ‫ڭ‬ [ɲ]
l lā ‫ل‬ [l] [ɫ]
m mā ‫م‬ [m]
n nā ‫ن‬ [n]
ū, v var ‫و‬ [uː] [ʋ]
ä, h hir ‫ه‬ [a] [h]
ī, y yam ‫ی‬ [iː] [j]
Table 1: Bāla’y-balan alphabet

There are several other additional orthographic devices used in Bāla’y-balan,

including diacritic marks used to show the presence of a short vowel following
a consonant. In table ⁇, diacritics are shown in conjunction with the letter ‫ف‬,
except qam, which may only appear above the letter ‫ا‬.

Transcription Name Arabic IPA

ā qam ‫ٓا‬ [aː]

- qaqnam ‫ف‬ّ -
- ˁazam ‫ْف‬ -
a kavam ‫َف‬ [a]

6
 ‫ُف‬
‫ِف‬
u cumam [u]
i ṣiyam [i]
Table 2: Bāla’y-balan diacritics

Vowels
1. In word-initial position, the letter ‫ ا‬does not itself have a vowel sound; in-
stead it is a glottal stop that must be accompanied by a long or short vowel
marker as explained in the following points. In practice, the short vowel
diacritics are often implied rather than written in Arabic, Persian and
Turkish, although Muhyî was diligent about providing diacritics when
writing Bāla’y-balan.

2. In word-initial position, [aː] is represented by ‫ ا‬with the horizontal qam

diacritic: ٓ‫ا‬. In medial and word-final positions, ā is written with a plain ‫ا‬
and a kavam diacritic above the preceding consonant.

3. In word-initial position, [uː] is represented by the digraph ‫اُو‬. In medial

and word-final positions, ū is written with a plain ‫ و‬and a cumam diacritic
above the preceding consonant.

4. In word-initial position, [iː] is represented by the digraph ‫اِی‬. In medial

and word-final positions, ī is written with a plain ‫ ی‬and a ṣiyam diacritic
below the preceding consonant.

5. Bāla’y-balan’s short vowels, [a], [u] and [i], are indicated only through
the use of the three diacritics, kavam, cumam and ṣiyam. In word-initial
position, these diacritics are placed above or below the letter ‫ا‬. In medial
position, the diacritics are placed above or below the consonant that pre-
cedes the vowel. In figure ⁇, kavam is used to indicate long and short
َ ‫ات َی ْب‬
vowels.⁸ ‫كشَ ا‬ َ ‫َات َیوكَشَ ا و َح‬
َ ‫ ذ‬ẕāta’ya-vkaşā ve ḥāta’ya-bkaşā.⁹ Word-
⁸Note that in Kitâb-ı Bâleybelen, Muhyî often wrote Bāla’y-balan in red ink and Ottoman Turk-
ish in black. In this example, however, diacritic marks for Bāla’y-balan are, nevertheless, writ-
ten in black.
⁹Koç transcribes these as ẕâte’y-vekeşâ ve ḥâte’y-bekeşâ, which shows a different reading of the

7
 final [a] is indicated by a ‫ ه‬and will be transcribed here by ä. Short [u]
and [i] do not occur word-finally in Bāla’y-balan. A suffix which consists
of a single short vowel is marked by the addition of a diacritic as in the
case of the -a suffix on ẕāt and ḥāt in the example above.

Figure 1: ẕāta’ya-vkaşā ve ḥāta’ya-bkaşā

6. The ˁazam diacritic is used to indicate that a consonant has no following

vowel.

Consonants
1. The qaqnam diacritic indicates that a consonant should be doubled; ex-
cept in the case of affixation, Bāla’y-balan does not allow two identical
consonants to be written side-by-side.

Syllables
1. Bāla’y-balan allows the following syllable: V, VC, CV, CVC, CVCC.¹⁰
Open CV syllables are sometimes transformed into VC syllables by the
addition of the definite article y- as described in section ⁇. Short or long
vowel length is permissible in all syllable types. Syllables with two final
consonants are rare but possible: maqd ‘sausage’; madr ‘biol. sage’.¹¹
definite article and the first syllable in each of the following nouns. The text, especially in the
second of the two compounds, would seem to support my reading. See the discussion of ˁazam
in point ⁇ and section ⁇.
¹⁰Initial vowels are preceded by a glottal stop ‫ ا‬in writing.
¹¹Note that the qaqnam diacritic indicates a doubled consonant which is split between two
syllables.

8
 2. Two vowels may not appear in sequence, and in the case that affixation
should cause two vowels to occur side-by-side, Bāla’y-balan will either:
a) eliminate one of the vowels; b) insert a buffer consonant (often ˀ or y).

3 Morphology
Bāla’y-balan is an agglutinating language which makes heavy use of prefixes,
suffixes, circumfixes, vowel lengthening and compounding to create words from
noun or verb roots. A number of suffixes in Bāla’y-balan are characterised by
being equally valid as prefixes as they are as suffixes, although it is still only a
minority that function in this manner.
Nouns and pronouns in Bāla’y-balan are marked with the following cases:
nominative, accusative, ablative, dative, locative and instrumental; each of these
may in turn be combined with a possessed case. Bāla’y-balan cases follow the
meanings generally used in classical grammar: nominative (subject of verb); ac-
cusative (object of verb); ablative (direction from); dative (direction towards);
locative (position at); instrumental (performed with). Nouns and pronouns may
also be inflected for number by the addition of a plural suffix. Nouns and pro-
nouns are not marked for gender.

3.1 Pronouns
Bāla’y-balan has six pronouns, which in turn have a complex declension for
case. In table ⁇, the two root columns indicate the roots to which suffixes in
table ⁇ may be applied. There are short and long roots for each pronoun. The
enclitic may be attached to a verb in lieu of a pronominal object in the accusative
or dataive case. The possessed enclitics, which are attached to nouns to indicate
possession, are discussed below (see table ⁇).
Table ⁇ shows pronoun declension in Bāla’y-balan. The symbols N, L and
S indicate the root to which the suffix may be attached: (N)ominative; (L)ong
root; and (S)hort root. The accusative suffix -ī requires the buffer -b- after roots

9
 Root
Person Nominative Short Long Enclitic
1st sg. nav -na- -nav- -na
2nd sg. gav -ga- -gav- -ka
3rd sg. ra/rā -ra- -rā- -ra/-ū
1st pl. navā -nā- -navā- -nā
2nd pl. gavā -gā- -gavā- -kā
3rd pl. rayā/rāyā -rayā- -rāyā- -rayā
Table 3: Pronouns

ending in a vowel. There are a number of suffixes which have multiple mean-
ings; the prefix ba-, for example, may indicate the dative, locative or instru-
mental cases.

Case Form Example

Nominative see table ⁇ nav ‘I’
L/S-rā narā ‘me’
Accusative
L-(b)ī rābī ‘him’
fa-N fanav ‘from me’
Ablative
L-fa navfa ‘from me’
L/S-mī namī ‘to me’
L/S-rā rayārā ‘to them’
Dative
ba-N banavā ‘to us’
bayā-N bayāgavā ‘to you (pl)’
ba-N bagav ‘on you (sg)’
da-N darā ‘on her’
Locative
N-ba gavāda ‘on us’
N-da rāyāda ‘on them’
N-na gavna ‘with you (sg)’
Instrumental N-mī navmī ‘with me’
ba-N bagavā ‘with us’
Table 4: Pronominal case suffixes

10
3.2 Nouns
Plural
The Bāla’y-balan plural is created by the addition of the suffix -ā: ‫ نِو‬niv ‘flower’
and ‫ نِ َوا‬nivā ‘flowers’. When the base noun already ends with a -ā, then -y- is
added as a buffer between the two vowels. When the base noun ends in the
vowel ‫[ ه‬a], then the long ā is added, and the ‫ ه‬is vocalised with its consonontal
value [h]: ‫ ظفه‬ẓafa ‘book’, ‫ ظفها‬ẓafahā.

Accusative
A noun may be placed in accusative case by the addition of the suffix -rā. This
may be added to both singular and plural nouns: ‫ شَ ْم َسا نِ َوا َرا‬Şamsā nivārā ‘They
smelled the flowers’.

3.3 Possession
Possessive relationship in Bāla’y-balan are indicated in two very different ways
depending on whether the possessor is a pronoun or a noun. Enclitic pro-
nouns are used in Bāla’y-balan to show that a noun is possessed by a pronoun.
These enclitic suffixes are declined for person and for certain cases (that pos-
sessed predicates have a special form different from both the nominative and
accusative). Possessed nouns in the accusative, instrumental, ablative, dative
and locative cases are declined identically. For possession by a noun, the ap-
propriate 3rd person, either singular or plural, is employed.

3.4 Genitive Constructs

The Bāla’y-balan genitive construct is widely used to show a relationship be-
tween two nouns. The construct has two base forms, definite and indefinite, as
well as a combined form. In all cases, the governing (possessed) noun appears

11
first while the modifying (possessor) noun follows. Note that the genitive con-
struct has uses beyond showing a simple possessive relationship between two
nouns, and is often found to describe an action upon a noun by using a gerun as
the governing noun: navama’y-zunā ‘an updating of the books’ from navam
‘to renew, update’ and zun ‘register; account’. It can also be used to attach a
modifying adjective to a noun: yi-qsama’y-hinam ‘the fifth part’ from qisam
‘number’ and hinam ‘fifth’.

Indefinite
When the modifying noun is indefinite, the governing noun takes the suffix
-a: av’a gīr ‘dog of hunting > hunting dog’. The indefinite genetive suffix is
normally written with the kavam diacritic. However, if the noun ends with the
letter ‫ه‬, an amam is attached instead. Likewise, if the noun ends in a long vowel,
the buffer consonant -y- is used to separate the two vowels: subāya dal ‘water
of a sea > sea water’.
dara zar ‘mother of a woman > mother-in-law’.

Definite
In a definite genitive construct, the definite article y- is additionally attached to
the noun that it marks. The definite article can also be attached to the governing
noun to make the entire construct definite, as in yi-qsama’y-hinam above. See
section ⁇ for a full treatment of the definite article.

Person Nominative Pred. Other

1st sg. -al/-ab/-ad/-nav/-an -al -al
2nd sg. -ag/-nuv -ga -ag
3rd sg. -ī/-av/-rā/-ra¹² -ī -an
1st pl. -lā -alā -alā
2nd pl. -agā/-ūyā -gā -agā

¹²After plurals, the third-person singular suffixes ī and -av require the buffer consonant -b-: -bī
and -bav.

12
 3rd pl. -iyā/-avā/-rāyā¹³ -yā -anā
Table 5: Suffixes of possession

3.5 Demonstratives

Noun Adjective
This rā rā
That gī vā
Table 6: Bāla’y-balan demonstratives

3.6 Passive
p.169 kafam = demek, kāfam = denilmek

3.7 Definite and Indefinite Nouns

p.99 -ā indefinite ya- definite
¹³After plurals, the suffix -avā loses the initial vowel: -vā.

13
3.8 Predicates
present predicate suffix (after possessed nouns only⁇) -ar -> see p. 98

4 Syntax

4.1 Definite Article

1. In a definite genitive construct, the definite article y is placed before the
noun that it marks. In Bāla’y-balan the article is attached directly to the
noun, while in transcription, a dash will be used to separate the two.

2. There is disagreement on how the article should be formed in certain in-

stances. According to Koç, the article is y when following a word ending
in a vowel or preceding a noun beginning in a vowel: y-anā. Elsewhere,
the article requires a buffer vowel between it and its noun: ya-bar.
The original manuscript, however, suggests that the rules are somewhat
more complex. The main difference is that the article absorbs the first
vowel from nouns with an open first syllable. The steps for selecting the
appropriate form for an article in Bāla’y-balan can be summarised as:

a) if the noun starts with a vowel, no buffer is necessary

b) if there is not a vowel immediately preceding the article, use a buffer
vowel¹⁴
c) if the noun consists of a single syllable, no buffer is necessary
d) if the noun starts with a closed syllable, no buffer is necessary
e) use a buffer vowel

If the buffer vowel is necessary, the first syllable of the noun loses a short
vowel, which will have been represented by a diacritic, and that same
¹⁴The presence of a preceeding vowel does not mean that there will not be a buffer vowel.

14
 diacritic is placed over the y-, giving it the form ya-, yu- or yi-.¹⁵ In the
case of a long vowel, the diacritic is reduplicated in both the article, which
will have a short vowel, and the first syllable of the noun, which will retain
its long vowel.
The following examples from a single passage in Kitâb-ı Bâleybelen demon-
strate the application of these rules.¹⁶ Koç does not follow this pattern,
preferring y- in all cases except ya’ḫşenâ, but no explanation is given for
this deviation.¹⁷

a) ‫َر ِگ ْو َزا َو ُی ْن َشا‬

ra-givzāva’yu-nşā

b) ‫َف ِمی َم ْی َم ْف َنا‬

ā-mīmā’y-mafnā

c) ‫اَ َی ْن َف ْم‬
a ya-nfam

d) ‫َجا َم ْی َن ْن َش َنا اَ َیخْ َش َنا‬

cāma’y-nanşanā a ya’xşanā

e) ‫َر ْی َك ْر َف َنا َرا َیا َی ْع َش َنا‬

rā’y-karfanā rāyā ya-ˁşanā

5 Numerals

5.1 Cardinals
Numbers in Bāla’y-balan are based on the decimal system as in Arabic, Persian
and Turkish.
¹⁵Note that the first syllable in balan is closed, so *bāla’ya-blan is not the correct form of Bāla’y-
balan.
¹⁶All diacritics are from the original: Kitâb-ı Bâleybelen.
¹⁷Page 188.

15
 1 ad 10 yā
2 baz 20 kar
3 cil 30 liş
4 dum 40 mud
5 hin 50 nih
6 vus 60 sav
7 zā 70 ˁaz
8 ḥī 80 fī
9 hid 90 ṭay
0 baḥlā 100 qā 1000 ġan

Table 7: Bāla’y-balan numbers

A few of the two-digit numbers have alternate forms: 40 mud or mut; 60

sav, siv, suv, sax or sux; 90 ṭay or ṭī.
The 10 digits can be combined to create larger numbers although they are
not always arrange in the order of most to least significance. For numbers be-
tween 11 and 101, the numbers 1 through 9 precede the numbers 10 through 100:
hidyā ‘nine-ten’; bazkar ‘two-twenty’. Compare these to: qā adyā ‘hundred-
one-ten’; qā dumnih ‘hundred-four-fifty’. For the number 101, both orderings
are possible: adqā = qaˀad ‘one-hundred’. The reversed form may be used when
using multiples of 10 plus 100: qabaz = bazqā ‘twenty-hundred’.¹⁸
Most digits have shortened forms used in compounds, fractions and ordinals
that are created by dropping the final consonant. Shortened forms are possible
for the numbers 11 through 17: adyā = ayā 11; bazyā = bayā 12; cilyā = ciyā
13; dumyā = duyā 14; hinyā = hiyā 15; vusyā = vuyā 17; zāyā = zayā 1; ḥīyā
18; hidyā 19. This pattern repeats itself for 21 through 27, 31 through 37 and
so on. In the rare case, of inversion and multiples of 10, reduced forms are also
possible: yaqā 110; kaqā 120.
Hundreds are formed by adding the multiplier before qā 100: cilqā ‘three-
hundred’; vusqā zāliş ‘six-hundred-seven-thirty’. Again, shortened forms are
possible for the hundreds, although in this case from 200 to 800: bazqā = baqā
¹⁸Note that qā can lose its long vowel in compounds.

16
200; cilqā = ciqā 300; dumqā = duqā 400; hinqā = hiqā 500; vusqā = vuqā 600;
zāqā = zaqā 700; ḥīqā = ḥaqā 800; hidqā 900.¹⁹
Thousands are formed in the same manner as the hundreds, and shortened
forms : cilġan = ciġan ‘three-thousand’; vusġan ḥīqā hifī ‘six-thousand-eight-
hundred-five-eighty’. Thousands and hundreds may be combined for larger
numbers: dukar ġan qā ciṭay ‘four-twenty-thousand-hundred-three-ninety’
24,193. A shortened form, ġa, may be used in the place of ġan: ġanad = ġaˀad
1,001; ġanbaz = ġabaz 1,002.
Other numeral orders are possible when they do not produce ambiguity:
ciyā qaġan ‘three-ten-hundred-one thousand’ 1,113. Contrast this with qā ciyā
ġan ‘hundred-three-ten-thousand’ 113,000.

5.2 Fractions
Fractions are formed in several different manners. The standard means is to
separate the numerator from the denominator with the prefixes da or fa: ad
facil ‘1⁄3’. It is also possible to omit the prefix in thes constructs and use simple
juxtaposition: adcil. The shortened digits may also be used in the numerator:
acil. Finally, it is also possible to reverse the order of the numerator and denom-
inator: cilad. This would imply that in Bāla’y-balan, it is not possible to express
improper fractions, such as 3⁄2, since the denominator is alway assumed the be
larger of the two numbers in a fraction statement. Therefore dumcil would be
interpreted as ‘3⁄4’ rather than ‘4⁄3’, which would be stated as ad ad facil ‘1 1⁄3’.

5.3 Ordinals
Ordinals can be regularly formed with the -(y)am suffix attached to the last
element in the number: adam ‘first’; yāyam ‘tenth’; qāyam ‘hundredth’. With
certain single digit ordinals, the sufix -am can optionally be fixed to the first
¹⁹The number hid has no reduced form in Bāla’y-balan.

17
syllable of the number; others have irregular forms: ad + am > am²⁰; baz + am
> bam; cil + am > cam; dum + am > dam; hin + am > ham; vus + am > vam; zā
+ am > zam; hī + am > ham; hid + am > hidam.

5.4 Distributives
Distributive nouns can be formed in Bāla’y-balan by replacing the vowel of the
last syllable in a number with ā: ād ‘one each’; ġān ‘a thousand each’.
Distributive adjecties are formed in the same manner with an additional
suffix -i.²¹ A y is inserted between a word-final vowel and the suffix -i: ‫ای اَ َو ْر‬
َ ‫َی‬
yāyi avar ‘a hundred horses each’.
Using reduplication, adverbial constructs such as bazābaz ‘two-by-two’ can
also be formed. Note that when the numeral ends in a long vowel, the first loses
its length and a buffer v is used: qā + v + qā > qavāqā ‘a hundred at a time’.

5.5 Abjad
Each base letter²² of the Bāla’y-balan alphabet has a numerical value which
is used to write numbers. The system, in Arabic abjad, completely replaces
the standard Indic digits which are regularly used in languages which use the
Arabic alphabet.²³
In Kitâb-ı Bâleybelen, Muhyî also presented variant characters to replace
the standard letters for writing Bāla’y-balan’s numbers. There is a one-to-one
relationship between the new characters and the standard Arabic letters, so it is
possible that these symbols could have additionally been used to write Bāla’y-
²⁰Note that vowel-initial words are considered to start with a glottal stop represented by the
letter anā, so combining ‫ اَد‬with the suffix ‫ َـم‬produces ‫اَم‬.
²¹The suffix is written with a ṣiyam diacritic.
²²The base letters are only those found in the Arabic alphabet and exclude the additional forms
used for Persian and Ottoman Turkish. Persian letters take the value of the original letter.
²³The Indic numbers are: ٥, ١, ٢, ٣, ٤, ٥, ٦, ٧, ٨, ٩.

18
balan.²⁴ Bāla’y-balan’s numerical symbols are in fact partially copies of Arabic’s
Indic digits (seven standard forms and two variants used in Persian: ۴ and ۵)
and one letter (‫) ظ‬. The remainder are Muhyî’s own invention. There is some
precedence for alphabet creation at the time; Koç mentions that a contemporary
sufi is said to have written a “strange tongue” in an “unusual alphabet”.²⁵ The
work in question, however, is lost, and the new characters in Kitāb-ı Bâleybelen
are only used numerically.

‫ا‬ ‫ب‬ ‫ج‬ ‫د‬ ‫ه‬ ‫و‬ ‫ز‬ ‫ح‬ ‫ط‬

١ ٢ ٣ ٥ ٦ ٩
1 2 3 4 5 6 7 8 9
‫ی‬ ‫ك‬ ‫ل‬ ‫م‬ ‫ن‬ ‫س‬ ‫ع‬ ‫ص ف‬
٧ ٨
10 20 30 40 50 60 70 80 90
‫ق‬ ‫ر‬ ‫ش‬ ‫ت‬ ‫ث‬ ‫خ‬ ‫ذ‬ ‫ض‬ ‫ظ‬ ‫غ‬
۴ ۵ ‫ط‬
100 200 300 400 500 600 700 800 900 1000

Table 8: Abjad values in Bāla’y-balan

Numbers 1999 and below are formed from the Abjad symbols by arranging
them in a descending order: ‫( هشغ‬1000-300-5) = 1305.²⁶ It should be noted that
despite the Arabic alphabet being read right-to-left, the highest order digits in
Bāla’y-balan numbers still appear to the left of lower order numbers.
For numbers in the thousands, Bāla’y-balan suffers from the fact that ‫غ‬
always equals 1000 and is not used as a thousand marker: ‫( جنظغسر‬200-60-
1000-900-50-3) = 261,953. Muhyî was aware of this difficulty and announced
that “200,000 is written 200 and 60,000 is written 60”, with context differenti-
²⁴Table ⁇ is currently incomplete as many glyphs are not available. Consequently, I will use
the Arabic alphabet in all examples.
²⁵See Koç, Bâleybelen, 53-54.
²⁶Under normal circumstances the Arabic letters should be joined as normal; however the sym-
bols Muhyî devised for his numbering system do not connect, so they will appear as such in
this document.

19
ating between the number which are and are not in the thousands. He does
not, however, provide the means of writing numbers such as 2,000; the most
reasonable assumption would be ‫( غب‬2-1000), but this is not stated.

6 Glossary

20