Marathi is written in the Devanagari script. It consists of 11 vowels, 40 consonants, and 2 sound modifiers.
The Devanagari syllabary is organised according to the way sounds are produced using the voice box and the mouth.
First are the vowels, which, with one exception, come in pairs:
अ आ इ ई उ ऊ ऋ ए ऐ ओ औ
Using the International Alphabet of Sanskrit Transliteration (IAST), these vowels would be represented thus:
a ā i ī u ū ṛ e ai o au
- अ (a) is pronounced as a default vowel, similar to a schwa.
- आ (ā) is like the vowel a in father.
- इ (i) is like the vowel in tin.
- ई (ī) is like the vowel in see.
- ऋ (rī) is like the vowel in ceen.
- उ (u) is like the vowel in book.
- ऊ (ū) is like the vowel in tool.
- ए (e) is like the vowel in May except there is no diphthong.
- ऐ (ai) is one of the only two diphthong vowels in Marathi, and pronounced differently depending on regional accent. Often it is pronounced like the vowel in hat.
- ओ (o) is like the vowel in hole.
- औ (au) is the other diphthong vowel, also with varying pronunciation. It is often pronounced like the vowel in saw or awe.
All vowels in Marathi have two forms: Their standalone form and their mātrā form. The mātrā form modifies consonants. Here are the eleven vowels paired with the syllable क (k):
- Note that there is no mātrā form for the first vowel, अ --a. This is because all Marathi consonants, unless part of a conjunct (see below), or they appear at the end of a word, automatically contain this vowel. So, the letter क is pronounced ka.
There are 25 regular consonants (consonants that stop air from moving out of the mouth) in Marathi, and they are organized into groups (vargas) of five. The vargas are ordered according to where the tongue is in the mouth. Each successive varga refers to a successively forward position of the tongue. The vargas are ordered and named thus (with an example of a corresponding consonant):
- Velar (k)
- Palatal (j)
- Retroflex (English t)
- Dental (Spanish t)
- Labial (p)
The five consonants in each group are ordered thus:
- unaspirated, unvoiced
- aspirated, unvoiced
- unaspirated, voiced
- aspirated, voiced
The difference between voiced and unvoiced consonants is as simple for an English speaker to understand as understanding the difference between the consonants "k" and "g" or "j" and "ch." However, aspiration is a distinction foreign to most English speakers. To understand the difference, do this simple exercise:
Hold your palm about an inch in front of your mouth. Say the word "pit." Notice that your hand will likely feel a short burst of air coming from your mouth. Now say the word "spit" and notice that there is little or no burst of air. Now attempt to say "pit" again, this time without letting the burst of air happen.
Here are the five vargas of regular consonants, followed by their corresponding IAST characters:
क ख ग घ ङ
च छ ज झ ञ
ट ठ ड ढ ण
त थ द ध न
प फ ब भ म
k kh g gh ṅ
c ch j jh ñ
ṭ ṭh ḍ ḍh ṇ
t th d dh n
p ph b bh m
- The nasal consonants in the first two vargas are almost never found written alone; they are usually conjuncted (more on conjuncts later) with another consonant in their corresponding vargas, as the other nasal consonants often are.
- The consonants in the first two vargas, as well as the last, are pronounced exactly as they are in English. In other words, a Marathi k sounds like an English "k" (without aspiration).
- The consonant written 'c' can be confusing; it is pronounced like the first sound in the word chair but without aspiration. The consonant ch is virtually indistinguishable from c to the untrained English speaker, but is indeed a different letter.
- The English consonants "t" and "d" fall somewhere in between the Marathi consonants t and d, and ṭ and ḍ. They are closer to the latter (retroflex), so when Marathi speakers say or speak English words containing t or d, they almost always use the retroflex version. To achieve these sounds, curl the tongue back and touch the tip to the roof of the mouth.
- The dental consonants are pronounced exactly like the "t" and "d" in Spanish, with the tongue touching the back of the upper teeth.
- Voiced, aspirated consonants (gh, bh, etc.) are by far the hardest sounds for the English speaker to learn how to make; however, with some practice they are not overly difficult.
There are four semivowels in Marathi:
य र ल व
y r l w (v)
- y is very similar to its corresponding English sound.
- r, like most non-English "r" sounds, is flipped, but not rolled.
- l is the same as the English "l".
- w (v) has a somewhat flexible pronunciation, depending on context and regional dialects. Sometimes it is pronounced like the English "w" (as in the word swami), and other times it is closer to the English "v". It is often pronounced somewhere in between the two sounds.
There are three sibilants:
श ष स
ś ṣ s
- ś is the same as "sh" in English.
- ṣ is also similar to "sh" in English, and though it is technically pronounced farther back in the mouth, functionally there is very little or no difference between the two in spoken Marathi.
- s is the same as "s" in English.
There is one fricative consonant:
- h is pronounced the same as "h" in English.
There are many sounds found in Marathi that do not directly correspond to any Devanagari letter:
- क़ (q) is pronounced farther back in the throat than क. For example, the Arabic word Qu'ran begins with this letter.
- ग़ (ġ) is pronounced in the back of the throat, similar to the French "r."
- ख़ (kh) is pronounced like "ch" in German words like Bach or Reich.
- ज़ (z) is the same as the English "z."
- ड़ (ṛ) is difficult to describe, except by example. The Marathi word गाड़ी (gaṛī) sounds very much like "gardee." Notice that the character used to denote this character is the same as denotes ऋ. The only way to tell the difference is context; if ṛ is followed by a consonant, it corresponds to ऋ (example: ऋषि -- ṛṣi or "Rishi"). If it is followed by a vowel or falls at the end of a word, is is most likely ड़.
- ढ़ (ṛh) is the aspirated version of ड़.
- फ़ is the same as the English sound "f."
The anusvara is notated with a small dot above the corresponding letter. In IAST, it is notated ṃ. It can have two different effects:
- In mid-word, when it appears before a consonant, it has the the same effect as if that consonant's corresponding nasal consonant were placed right before it. For example, if placed before the "k" sound in the English word "buck," it would make the word change to "bunk."
- Placed at the end of a word (after a vowel with no further consonants), it simply nasalizes the vowel, often sounding as if the English "ng" (as in "song") had been placed at the end of the word, though lightly.
The candrabindu (meaning "moon dot") is very similar to the anusvara and is notated with a dot and a small crescent (see below). When placed above a vowel, the vowel becomes nasalized.
The visarga is notated with what looks like an English colon. It is transliterated with ḥ. It is only found at the end of words, and only found on Sanskrit loan words. It has the effect of adding a h as well as a lighter version of the preceding vowel. For example, नमः is pronounced namaha, with the last a having as little emphasis as possible.
The halant is a small diagonal line which indicates that the default vowel अ -- a is not to be pronounced. It only appears at the end of words. In Marathi, however, this symbol is almost never seen, as, unlike Sanskrit, the default vowel is almost never pronounced on the final consonants of words. Example: नाम् -- nām (name).
Very often, two or more consonants are combined.