The Tamil script, also called tamiz ezuttu, is used for writing the Tamil language, a Dravidian language spoken by over 65,500,000 people in India, Sri Lanka, Singapore, Malaysia and Mauritius. Tamil is an official language in the south Indian state of Tamil Nadu as well as in Sri Lanka and Malaysia. The script is derived from Brahmi, so is related to many of the scripts used for writing Indian Indo-Aryan languages, to which the Tamil language itself is unrelated.
Tamil is written from left to right using an abugida containing eighteen consonants, or meyyeḻuttu "body letters", and twelve vowels, or uyireḻuttu "soul letters". The consonant inventory is much smaller than that of many other Brahmic scripts; Devanagari uses thirty-two consonant letters and Kannada thirty-four, for example. This is because voiced and unvoiced counterparts of a sound are represented using a single letter, with the pronunciation dictated by context. So க represents [kʌ] normally, but [gʌ] when it follows a preceding CV syllable. For example, in the words pronounced [kʌl] 'a stone' and [pʌgʌl] 'daytime'. Neither the script nor the spoken language distinguish between aspirated and unaspirated consonants, as many South Asian languages do.
There are five consonant letters which are not included in the eighteen accounted for above. These represent sounds which are not part of the Tamil sound inventory, but which are used for writing loan words from Sanskrit and other languages. This set of letters is called grantha and contains the letters ja, ṣa, sa, ha, and kṣa. Additional borrowed sounds, particularly fricatives such as [f] or [z], are represented by means of a sign called āytam which is written to the left of a stop to fricate it. For example, the grantha letter ஜ represents [jʌ] but ஃஜ represents [zʌ]. Sometimes non-native sounds are also represented by writing a sub- or superscript latin number next to the closest equivalent, so ஜ2 represents [jha].
A basic written symbol represents a CV syllable, called uyirmei a "living letter". In their unmarked form, consonant letters represent a consonant with an inherent [ʌ] vowel. Syllables containing other vowels are represented by writing a vowel diacritic above, below, to the left/right of, or flanking either side of a consonant. So ப represents [pʌ] but பி (ப+ி) represents [pi], and பெ (ப+ெ) represents [pe]. Note that the symbol for [e] is written to the left of the consonant it follows. One vowel, representing [ai] can be written with one of two symbols (called allographs), depending on the preceding consonant. The government of Tamil Nadu is trying to standardize the orthography so that only one allograph is consistently used. A consonant can also be marked with a diacritic called puḷḷi, a dot written centrally over the letter, which silences the vowel. So ப் represents [p] with no following vowel.
Vowels which are not preceded by a consonant, for example the [i] at the start of the word inta 'this', are written using one of the twelve independent "soul letters". In ancient texts, these were used at the beginning of metrical groups (that is, a sequence of syllables), but in the modern orthography they are used at the beginning of words. They can also be used word-internally to represent 'overlong' vowels.
The Tamil script is unusual amongst Indian writing systems in that it does not generally use conjuncts. Rather, the puḷḷi is used to place consecutive consonants in a linear string. A notable exception to this is kṣa, which tends to be written using a ligature.
Tamil writing employs a script-specific set of numerals 0-9, 100 and 1,000. It also uses a limited set of logograms, including symbols for 'day', 'month', 'year', 'debit', 'credit', 'as above', and 'rupee'.