Numbers in Hindi: 0 to 100, lakh and crore, Devanagari digits, IPA

Why every number from 1 to 99 has its own word

In English, twenty-one is twenty plus one. In Hindi it is इक्कीस (ikkīs), a single word in which the one and the twenty have melted together. Wikipedia's article on Hindustani numerals states it plainly: the system is contracted so far that nearly every number from 1 to 99 is irregular and has to be memorised separately. The forms grew directly out of Sanskrit through centuries of speech. Wiktionary traces ikkīs through Prakrit ekkavīsī to Sanskrit ékaviṃśati, one plus twenty.

Patterns are there for anyone who reads the grid below by columns. The ones column starts with ik-: ikkīs 21, ikattīs 31, iktālīs 41, ikyāvan 51. The twos start with b-, the threes with t-, the fives with p-. Each decade has its own ending: -īs for the twenties and thirties, -ālīs for the forties, -van for the fifties, -saṭh for the sixties, -hattar for the seventies, -āsī for the eighties and -nve for the nineties.

The most useful pattern is in the column of nines. Nineteen is उन्नीस (unnīs), twenty-nine उनतीस (untīs), thirty-nine उनतालीस (untālīs): each starts with un- and is built on the next ten, as one short of it. Wiktionary records the phrase उन्नीस बीस, nineteen-twenty, for two things that are almost the same. The pattern stops at 89 and 99, which are नवासी (navāsī) and निन्यानवे (ninyānve).

Hundreds, thousands, lakh and crore

Above a hundred the numbers in Hindi turn regular. Two hundred is दो सौ (do sau), five thousand पाँच हज़ार (pā̃c hazār), and the parts line up as in English: तीन सौ बीस is 320. Hazār is a Persian loan; Wiktionary notes that it pushed aside the older Sanskrit-derived word सहस.

The big difference is the next step. Instead of counting in thousands of thousands, Hindi names every second power of ten: लाख (lākh) is a hundred thousand, करोड़ (karoṛ) is ten million, अरब (arab) a hundred crore, or one billion. Ten lakh is a million. Indian English uses the same words, and newspapers write prices in lakh and crore.

The Indian way of writing large figures follows the words. After the first three digits, commas come every two places: 1,00,000 for one lakh, 1,50,000 for 1.5 lakh, 3,00,00,000 for three crore. The Unicode CLDR number pattern for Hindi, #,##,##0, encodes exactly this grouping, so software set to Hindi prints large figures this way.

Devanagari digits and zero

Written numbers in Hindi can use their own digits: ० १ २ ३ ४ ५ ६ ७ ८ ९. They work exactly like 0 to 9, place by place, so २०२६ is 2026. Wikipedia's article notes that written Hindi usually uses them, while CLDR sets the international digits 0 to 9 as the default for the Hindi locale and lists Devanagari as the native system. The digits in 300+ languages table shows the Devanagari row next to Bengali, Gujarati and the other Indian scripts.

Zero has two names. शून्य (śūnya) is the learned Sanskrit word preferred in Modern Standard Hindi; सिफ़र (sifar), from Arabic ṣifr, is the word Urdu prefers, and Wiktionary lists both under the Hindi entry for 0.

Ordinal numbers and fractions in Hindi

First to tenth are पहला, दूसरा, तीसरा, चौथा, पाँचवाँ, छठा, सातवाँ, आठवाँ, नवाँ, दसवाँ. From fifth on they add -vā̃ to the cardinal; the first four and sixth are irregular. Ordinals are adjectives ending in -ā, so they change with gender and number: पहला दिन the first day, पहली बार the first time, पहले दिनों में in the first days. Sixth also appears as छठवाँ.

Hindi has single words for fractions that English spells out. डेढ़ (ḍeṛh) is one and a half, ढाई (ḍhāī) two and a half; सवा (savā) adds a quarter and पौने (paune) takes one away, while साढ़े (sāṛhe) adds a half, mostly from three upwards. So डेढ़ सौ is 150, ढाई लाख is 2,50,000 and पौने दो किलो is 1.75 kilos. The same words tell the time: डेढ़ बजे is half past one, सवा एक a quarter past one, पौने दस a quarter to ten.