Devanāgarī

Devanāgarī ( —, pronounced , and in English pronounced ) is an abugida writing system used to write, either along with other scripts, or exclusively, several North Indian languages, including Sanskrit, Hindi, Marathi, Sindhi, Bihari, Bhili, Konkani, Bhojpuri, Nepali from Nepal and sometimes Kashmiri and Romani. It is written and read from left to right.

The transliteration used in this article follows the popular IAST conventions. The ITRANS ^* is a lossless transliteration scheme of Devanāgarī into ASCII that is widely used on Usenet. In ITRANS, the word Devanāgarī is written as "devanaagarii".

Origins

Devanāgarī emerged around 1200 AD out of the Siddham script, gradually replacing the earlier, closely related Sharada script (which remained in parallel use in Kashmir). Both are immediate descendants of the Gupta script, ultimately deriving from the Brāhmī script attested from the 3rd century BC; Nāgarī appeared in approx. the 8th century as an eastern variant of the Gupta script, contemporary to Sharada, its western variant. The descendants of Brahmi form the Brahmic family, including the alphabets employed for many other South and South-East Asian languages.

Etymology

Sanskrit nāgarī is the feminine of nāgara "urban(e)", an adjectival vrddhi derivative from nagara "city"; the feminine form is used because of its original application to qualify the feminine noun lipi "script" ("urban(e) script", i.e. the script of the cultured). There were several varieties in use, one of which was distinguished by affixing deva "deity" to form a tatpurusha compound meaning the "urban(e) of the deities (= gods)", i.e. "divine urban(e) [script". However, the widespread use of "Devanāgarī" is a relatively recent phenomenon; well into the twentieth century, and even today, simply "Nāgari" was also in use for this same script. The rapid spread of the usage of "Devanāgarī" seems also to be connected with the almost exclusive use of this script in colonial times to publish works in Sanskrit, even though traditionally nearly all indigenous scripts have actually been employed for this language. This has led to the establishment of such a close connection between the script and Sanskrit that it is, erroneously, widely regarded as "the Sanskrit script" today.

Interpreted by popular etymology to refer to a "City of the Gods", the name in certain Yogic traditions was taken to refer to the body of the individual. The philosophy behind this is that when one meditates on the specific sounds of the Devanagari alphabet, the written forms appear spontaneously in the mind.

Principles

Devanāgarī has 12 svara (pure sounds, or vowels) and 34 vyanjana (ornamented sounds, consonants). An akshara is formed by the combination of zero or one vyanjana and one or more svar, and represents a phonetic unit of the shabda (utterance). The akshara is written by applying standard diacritical modifiers to the vyanjana corresponding to the svara. An akshara is usually more basic and predictable than the syllable in English. For example, the English 'cat' (considered to have just one syllable) is written as two aksharas, the 'k-a' and the 'ta'.

The svara and vyanjana are ordered and grouped logically for studying or reciting. Thus the pure sounds, 'a', 'i', 'u' and their lengthened versions ('aa', 'ii', 'uu') are followed by the combined ('e', 'ai', 'o', 'au'), nasal ('.m') and aspirated ('.h') forms. The vyanjana themselves are grouped into 6 groups (rows) of 5 (columns). The first five rows progress as velar, palatal, retroflex, dental and labial, corresponding to utilizing or touching the tongue to progressively outer parts of the mouth when making the sound. Additional vyanjana are technically sonorants, sibilants or widely used conjunct forms. For each row or group, the columns logically progress to softer sounds, paired with aspirated forms, ending in the nasal form for that group.

Devanāgarī is written from left to right. In Sanskrit, words were written together without spaces, so that the top bar is unbroken, although there were some exceptions to this rule. The break of the top line primarily marks breath groups. In modern languages, word breaks are used. Languages written with Devanāgarī require no case distinction.

The Devanāgarī writing system can be called an abugida, as each consonant has an inherent vowel (a), that can be changed with the different vowel signs. Most consonants can be joined to one or two other consonants so that the inherent vowel is suppressed. The resulting conjunct form is called a ligature. Many ligatures appear simply as two individual consonants joined together, and so are a form of ligature. Some ligatures are more elaborately formed and not as easily recognized as containing the individual consonants.

When reading Sanskrit written in Devanāgarī, the pronunciation is completely unambiguous. Similarly, any word in Sanskrit is considered to be written only in one manner (discounting modern typesetting variations in depicting conjunct forms). However, for modern languages, certain conventions have been made (e.g. truncating the vowel form of the last consonant while speaking, even as it continues to be written in full form). There are also some modern conventions for writing English words in Devanagari.

Certain Sanskrit texts and mantras are typically written with additional diacritical marks above and below the akshara to denote pitch and tempo, to ensure completely accurate reproduction of the sound.

Symbols of Devanāgarī

All the vowels in Devanāgarī are attached to the top or bottom of the consonant or to an <aa> vowel sign attached to the right of the consonant, with the exception of the <i> vowel sign, which is attached on the left. In the Devanāgarī vowel table below, the "Letter" column contains the symbol used when a vowel occurs without a consonant, the "Vowel sign with <p>" column contains the symbol used when a vowel is attached to a consonant, shown with the <p> letter as an example, the "Unicode name" column contains the name given in the Unicode specification for the vowel, and the "IPA" column contains the International Phonetic Alphabet character(s) corresponding to the Hindi pronunciation of the Devanāgarī character.

Vowels

The vowels of the Devanāgarī script with their word-initial Devanāgarī symbol, diacritical mark with the consonant प् (/p/), pronunciation (of the vowel alone and of /p/+vowel) in International Phonetic Alphabet (IPA), equivalent in International Alphabet of Sanskrit Transliteration (IAST) and Indian languages transliteration (ITRANS) and (approximate) equivalents in Standard English are listed below:

Independent Vowel	Diacritical mark with “प्”	Pronunciation	Pronunciation with /p/	IAST equiv.	ITRANS equiv.	English eqivalent
अ	प	or	or	a	a	short Schwa: as the a in above or ago
आ	पा			ā	A	long open central unrounded vowel: as the a in father
इ	पि			i	i	short close front unrounded vowel: as i in bit
ई	पी			ī	I	long close front unrounded vowel: as i in machine
उ	पु			u	u	short close back rounded vowel: as u in put
ऊ	पू			ū	U	long close back rounded vowel: as oo in school
ए	पे			e	e	long close-mid front unrounded vowel: as a in game (not a diphthong), or é in café
ऐ	पै			ai	ai	a long diphthong: approx. as ei in height
ओ	पो			o	o	long close-mid back rounded vowel: as o in tone (not a diphthong)
औ	पौ			au	au	a long diphthong: approx. as ou in house
ऋ	पृ				R	short syllabic vowel-like Alveolar trill :
ॠ	पॄ				RR	long syllabic vowel-like Alveolar trill: a longer version of
ऌ	पॢ				LR	short syllabic vowel-like Alveolar lateral approximant: approx. as handle
ॡ	पॣ				LRR	long syllabic vowel-like Alveolar lateral approximant: a longer version of

Additional points:

The vowel in Sanskrit is more central and less back than the closest English equivalent, . The schwa () is always short in Sanskrit.
All vowels in Hindi, short or long, can be nasalized. All vowels can have acute, grave or circumflex pitch accent (in Vedic Sanskrit).
In Hindi, ऋ is pronounced as . The last three vowels in the table above do not occur in Hindi at all.
Note that the ancient Sanskrit grammarians have classified the vowel system as velars, retroflexes, palatals and plosives rather than as back, central and mid vowels. Hence and are classified respectively as palato-velar (a+i) labio-velar (a+u) vowels respectively. But the grammarians have classified them as diphthongs and in prosody, each is given two mātrās. This does not necessarily mean that they are proper diphthongs, but neither excludes the possibility that they could have been proper diphthongs at a very ancient stage. These vowels are pronounced as long and respectively by most learned Sanskrit Brahmins and priests of today. Other than the "four" diphthongs, Sanskrit usually disallows any other diphthongs—vowels in succession, if occur, are converted to semivowels according to predetermined rules.
In Sanskrit and in some other dialects of Hindi (as well as in a few words in Standard Hindi), the vowel is pronounced as a diphthong or rather than . Similarly, the vowel is pronounced in some words as the diphthong or rather than . Other than these, Hindi does not have true diphthongs—two vowels might occur sequentially but then they are pronounced as two syllables (a glide might come in between while speaking). Otherwise in Standard Hindi, (ai) is long near-open front unrounded vowel: as a in cat; (au) is long open-mid back rounded vowel: as au in caught.
The short open-mid front unrounded vowel (: as e in get), does not have any symbol or diacritic in devanagari script. It occurs only as an allophonic variant of schwa (in place word-middle a, determined only by convention) in certain words in the Standard khariboli dialect of Hindi. E.g., the orthography dictates that must be pronounced as , but it is actually pronounced as . It also occurs in loanwords from English, where it might be accorded a new vowel symbol of (chandra: ). The short open-mid back rounded vowel (), does not exist in Hindi at all, other than for English loanwords. In orthography, a new symbol has been invented for it: .
Unicode transliteration scheme differs for some characters from IAST scheme. The differences are: ā→aa, ī→ii, ū→uu, →rr, →ll.

Consonants

The table below shows the traditional listing of the Sanskrit consonants with the (nearest) equivalents in English/Spanish. The parentheses give the corresponding transliteration (of the consonant alone) in IAST scheme—the most popular one. Each consonant shown below is deemed to be followed by the neutral vowel schwa (), and is named in the table as such.

Plosives
	Unaspirated Voiceless	Aspirated Voiceless	Unaspirated Voiced	Aspirated Voiced	Nasal
Velar	क (k) ; English: skip	ख (kh) ; English: cat	ग (g) ; English: game	घ (gh) ; Aspirated	ङ (ṅ) ; English: ring
Palatal	च (c) ; ≈English: chat	छ (ch) ; Aspirated	ज (j) ; ≈English: jam	झ (jh) ; Aspirated	ञ (ñ) ; English: finch
Retroflex	ट (ṭ) ; American Eng: hurting	ठ (ṭh) ; Aspirated	ड (ḍ) ; American Eng: murder	ढ (ḍh) ; Aspirated	ण (ṇ) ; American Eng: hunter
Apico-Dental	त (t) ; Spanish: tomate	थ (th) ; Aspirated	द (d) ; Spanish: donde	ध (dh) ; Aspirated	न (n) ; English: name
Labial	प (p) ; English: spin	फ (ph) ; English: pit	ब (b) ; English: bone	भ (bh) ; Aspirated	म (m) ; English: mine

Non-Plosives/Sonorants
	Palatal	Retroflex	Dental/ Alveolar	Velar/ Glottal
Approximant	य (y) ; English: you	र (r) ; American Eng: tearing	ल (l) ; English: love	व (v) ; English: vase
Sibilant/ Fricative	श (ś) ; English: ship	ष (ṣ) ; Retroflex form of	स (s) ; English: same	ह (h) ; ≈English home

At the end of the traditional table of alphabets, three cosonantal clusters are also added: क्ष (in Hindi), त्र and ज्ञ (in Hindi). Other than these, sounds borrowed from the other languages like Persian and Arabic are written with a dot (bindu or nukta) beneath the nearest approximate letter. They are not included in the traditional listing. Many native Hindi speakers, especially those who come from rural backgrounds and do not speak really good khariboli or Urdu, confused these sounds (except ) and pronounce them as the nearest equivalents in Sanskritized Hindi (listed in column 4). These are:

**Extra sounds**
Symbol	IPA Pronunciation and name	English (etc.) equiv.
क़	voiceless uvular plosive	Arabic: Qur'an
फ़	voiceless labiodental fricative	English: fun
ख़	voiceless velar fricative	German: doch
ग़	voiced velar fricative	Persian: Mughal
ज़	voiced alveolar fricative	English: zoo
ड़	unaspirated retroflex flap
ढ़	aspirated retroflex flap

Additional points:

The "r" of Sanskrit is as in Standard American English. In modern Sanskrit pronunciation, the vowel "ṛ" is sometimes realised as or , although many people (especially if their native language retains the original Sanskrit sounds, as is the case in Malayalam) do make the ̩ In Hindi, is as pronounced in Spanish perro.
There is no retroflex flap in Sanskrit. In modern Hindi and other Indo-Aryan languages, they have sprung up as the allophonic flap variants of Sanskrit's simple voiced retroflex plosives. The (ṇ or ण) in Sanskrit is not a flap but a simple nasal stop, although it is pronounced by modern pundits while chanting as a nasal variant of the voiced retroflex flap. Hindi has two proper retroflex flaps.
Aspiration is actually a puff of breath that may follow a plosive consonant. English speakers could try pronouncing the words "kite", "take", "chip" and "pat" with a greater-than-usual puff of breath after the first consonant. The corresponding unaspirated plosives must be pronounced with no significant puff of breath at all.
For practicing the voiced aspirates, one could try: "drag him", "said him", "enrage him", "grab him". The voiced aspirated plosives (also called as murmur stops) are extremely important and frequent in Sanskrit. Sanskrit (and its daughters) is the only language that has faithfully preserved these original Proto-Indo-European stops.
The dental consonants in Sanskrit are as in Spanish or French. They can be pronounced by pronouncing and (of English) by pressing the tip of the tongue against the back of the teeth rather than against the back of the alveolar ridge as done by English speakers. The normal "t" and "d" in IAST transliteration are the dental stops; and they occur much, much more frequently than the retroflex stops.
The retroflex consonants are the most difficult to pronounce. They are pronounced by curling the tongue such that its tip touches the roof of the mouth, like how the Americans pronounce their "r". The retroflex flaps are pronounced in a similar way, by bringing the tongue's tip to the roof of the mouth and giving it a sharp flap downwards. However, bringing the tip of the tongue a bit above the normal alveolar ridge would also work fine. The normal alveolar plosives of English and do not exist as such in Sanskrit/Hindi.
The palatal plosives of Sanskrit/Hindi do not have a sharp frictional sound following them, as what happened in English chips and jam. These are more of pure plosives than affricates.
Sanskrit/Hindi has no . Its nearest equivalent is , which is very close to , but does not a friction or buzzing sound associated with it. But in consonant clusters, this may allophonically change to .
The palatal sibilant of Sanskrit (IAST: ś) is very close to like the English sh in ship (although the Sanskrit phoneme is the voiceless alveolo-palatal fricative ) while the English phoneme is the voiceless postalveolar fricative with lip rounding). Today, speakers of Sanskrit vary the palatal fricative from to . In Hindi, it is always pronounced as in ship.
The retroflex sibilant is pronounced like , but with the tongue curled upwards towards the roof of the mouth. In Mādhyandini branch of Yajurveda, this phoneme is allowed to be pronounced at certain places as . In Hindi, this is pronounced as the English sh in ship.
The Sanskrit is a voiced allophone of the normal h. In Hindi, it is pronounced as in home.

Another consonant is ळ is not used in Hindi. It is retroflex, and used in Vedic Sanskrit, and Marathi.

Ligatures

Consonant clusters of two or more phonemes are realized by combining the aksharas into ligatures. Typically, the preceding akshara loses its vertical stroke and is put in direct contact with the succeeding one. In cases of aksharas that do not have vertical strokes in their independent form, the following aksharas are usually placed underneath the preceding one. In some cases, the ligatures take forms not readily recognizable as composed of the individual aksharas (e.g. <jñ>). Consonant clusters involving <r> are treated as a special case: preceding <r-> is realized as a right-facing hook above the following akshara, and following <-r> appears as a slanted stroke attached to the vertical stroke of the preceding akshara. Similarly for a cluster /XYZa/, both X and Y would be "halved". There are many variants for this consonant cluster writing in Devanāgarī script. The most common system is shown below for the traditional table. Here the second vowel is taken to be /n/, followed by the schwa.

ka-group	क्न	ख्न	ग्न	घ्न	ङ्न
cha-group	च्न	छ्न	ज्न	झ्न	ञ्न
Ta-group	ट्न	ठ्न	ड्न	ढ्न	ण्न
ta-group	त्न	थ्न	द्न	ध्न	न्न
pa-group	प्न	फ्न	ब्न	भ्न	म्न
ya-group	य्न	र्न	ल्न	व्न
va-group	श्न	ष्न	स्न	ह्न

Diacritics

अं (called anusvāra), pronounced as (IAST: ) is used for nasalizing the vowel in the syllable, the word-final allophone of /m/ and /n/. The diacritic (called chandrabindu/anunāsika) is used in certain shakhas instead of the anusvāra in certain phonetic contexts.

अः (called visarga), pronounced as (IAST: ) is the word-final allophone of and .

If a lonely consonant needs to be written without any following vowel, it is given a halanta/virāma diacritic below (प्).

avagraha ऽ is used in western editions to mark elision of a word-initial in sandhi.

Accent marks

The pitch accent of Vedic Sanskrit is written with various symbols depending on shakha. In the Rigveda, anudatta is written with a bar below the line (॒), udatta with a stroke above the line (॑) while svarita is unmarked.

Numerals

**Devanāgarī numerals**
०	१	२	३	४	५	६	७	८	९
0	1	2	3	4	5	6	7	8	9

Encodings

ISCII

ISCII is a fixed-length 8-bit encoding. The lower 128 codepoints are plain ASCII, the upper 128 codepoints are ISCII-specific.

It has been designed for representing not only Devanāgarī, but also various other Indic scripts as well as a Latin-based script with diacritic marks used for transliteration of the Indic scripts.

ISCII has largely been obsoleted by Unicode, which has however attempted to preserve the ISCII layout for its Indic language blocks.

See:

ISCII – IS13194:1991 ^*

Devanāgarī in Unicode

The Unicode range for Devanāgarī is U+0900 .. U+097F. Grey blocks indicate characters that are undefined.

U+090x

ँ

ं

ः

ऄ

अ

आ

इ

ई

उ

ऊ

ऋ

ऌ

ऍ

ऎ

ए

U+091x

ऐ

ऑ

ऒ

ओ

औ

क

ख

ग

घ

ङ

च

छ

ज

झ

ञ

ट

U+092x

ठ

ड

ढ

ण

त

थ

द

ध

न

ऩ

प

फ

ब

भ

म

य

U+093x

र

ऱ

ल

ळ

ऴ

व

श

ष

स

ह

़

ऽ

ा

ि

U+094x

ी

ु

ू

ृ

ॄ

ॅ

ॆ

े

ै

ॉ

ॊ

ो

ौ

्

U+095x

ॐ

॑

॒

॓

॔

क़

ख़

ग़

ज़

ड़

ढ़

फ़

य़

U+096x

ॠ

ॡ

ॢ

ॣ

।

॥

०

१

२

३

४

५

६

७

८

९

U+097x

॰

Devanāgarī Keyboard Layouts

Devanāgarī and Devanāgarī-QWERTY keyboard layouts for Mac OS X

The Mac OS X operating system supports convenient editing for the Devanāgarī script by insertion of appropriate Unicode characters with two different keyboard layouts available for use. To input Devanāgarī text, one goes to System Preferences → International → Input Menu and enables the keyboard layout that is wished to be to uses. One then views the keyboard layout at the Indian Language Kit Manual at Apple Docs.

|| devanagari-qwerty-layout-option.jpg

INSCRIPT

Typewriter

Phonetic

See home page

Software

Apple Type Services for Unicode Imaging - Macintosh
Graphite - open source (SIL)
Pango - open source (GNOME)
Uniscribe - Windows
WorldScript - Macintosh, replaced by the Apple Type Services for Unicode Imaging, mentioned above
Baraha - Devanāgarī Input using English Keyboard

External links

Electronic resources

Abugida writing systems | Alphabetic writing systems

Gourt Home::Gourt :: Devanāgarī