The Dravidian languages are a language family spoken by more than 215 million people, mainly in Southern India and northern Sri Lanka, with pockets elsewhere in South Asia. Since the colonial era, there have been small but significant immigrant communities outside South Asia in countries such as Hong Kong, Malaysia, Singapore, Indonesia, Mauritius and the Philippines.
Epigraphically the Dravidian languages are first attested in the 2nd century BCE as Tamil-Brahmi script on the cave walls discovered in the Madurai and Tirunelveli districts of Tamil Nadu. The Dravidian languages with the most speakers are (in descending order of number of speakers) Telugu, Tamil, Kannada and Malayalam, all of which have long literary traditions. Smaller literary languages are Tulu and Kodava. There are also small groups of Dravidian-speaking scheduled tribes, who live outside Dravidian-speaking areas, such as the Kurukh in Eastern India and Gondi in Central India.
Only two Dravidian languages are spoken exclusively outside the post-1947 state of India: Brahui in the Balochistan region of Pakistan and Afghanistan; and Dhangar, a dialect of Kurukh, in parts of Nepal and Bhutan. Dravidian place names along the Arabian Sea coasts and Dravidian grammatical influence such as clusivity in the Indo-Aryan languages, namely Marathi, Konkani, Gujarati, Marwari, and Sindhi, suggest that Dravidian languages were once spoken more widely across the Indian subcontinent.
Though some scholars have argued that the Dravidian languages may have been brought to India by migrations in the fourth or third millennium BCE or even earlier, the Dravidian languages cannot easily be connected to any other language family and they could well be indigenous to India.[note 1]
The origin of the Sanskrit word drāviḍa is the word tamiẓ (Tamil). Kamil Zvelebil cites the forms such as dramila (in Daṇḍin's Sanskrit work Avanisundarīkathā) damiḷa (found in the Sri Lankan (Ceylonese) chronicle Mahavamsa) and then goes on to say, "The forms damiḷa/damila almost certainly provide a connection of dr(a/ā)viḍa " and "... tamiḷ < tamiẓ ...whereby the further development might have been *tamiẓ > *damiḷ > damiḷa- / damila- and further, with the intrusive, 'hypercorrect' (or perhaps analogical) -r-, into dr(a/ā)viḍa. The -m-/-v- alternation is a common enough phenomenon in Dravidian phonology"
Joseph (1989: IJDL 18.2:134-42) gives extensive references to the use of the term draviḍa, dramila first as the name of a people, then of a country. Sinhala BCE inscriptions cite dameḍa-, damela- denoting Tamil merchants. Early Buddhist and Jaina sources used damiḷa- to refer to a people of south India (presumably Tamil); damilaraṭṭha- was a southern non-Aryan country; dramiḷa-, dramiḍa, and draviḍa- were used as variants to designate a country in the south (Bṛhatsamhita-, Kādambarī, Daśakumāracarita-, fourth to seventh centuries CE) (1989: 134–138). It appears that damiḷa- was older than draviḍa- which could be its Sanskritization.
Based on what Krishnamurti states (referring to a scholarly paper published in the International Journal of Dravidian Linguistics), the Sanskrit word draviḍa itself is later than damiḷa since the dates for the forms with -r- are centuries later than the dates for the forms without -r- (damiḷa, dameḍa-, damela- etc.).
The 14th century Sanskrit text Lilatilakam, which is a grammar of Manipravalam, states that the spoken languages of present-day Kerala and Tamil Nadu were similar, terming them as "Dramiḍa". The author doesn't consider the "Karṇṇāṭa" (Kannada) and the "Andhra" (Telugu) languages as "Dramiḍa", because they were very different from the language of the "Tamil Veda" (Tiruvaymoli), but states that some people would include them in the "Dramiḍa" category.
In 1816, Alexander D. Campbell suggested the existence of a Dravidian language family in his Grammar of the Teloogoo Language, in which he and Francis W. Ellis argued that Tamil and Telugu descended from a common, non-Indo-European ancestor. In 1856 Robert Caldwell published his , which considerably expanded the Dravidian umbrella and established Dravidian as one of the major language groups of the world. Caldwell coined the term "Dravidian" for this family of languages, based on the usage of the Sanskrit word द्रविदा (Dravidā) in the work Tantravārttika by Kumārila Bhaṭṭa. In his own words, Caldwell says,Comparative Grammar of the Dravidian or South-Indian Family of Languages
The word I have chosen is 'Dravidian', from Drāviḍa, the adjectival form of Draviḍa. This term, it is true, has sometimes been used, and is still sometimes used, in almost as restricted a sense as that of Tamil itself, so that though on the whole it is the best term I can find, I admit it is not perfectly free from ambiguity. It is a term which has already been used more or less distinctively by Sanskrit philologists, as a generic appellation for the South Indian people and their languages, and it is the only single term they ever seem to have used in this manner. I have, therefore, no doubt of the propriety of adopting it.
The Dravidian languages form a close-knit family. Most scholars agree on four groups: South (or South Dravidian I), South-Central (or South Dravidian II), Central, and North Dravidian, but there are different proposals regarding the relationship between these groups. Earlier classifications grouped Central and South-Central Dravidian in a single branch. Krishnamurti groups South-Central and South Dravidian. Languages recognized as official languages of India appear here in boldface.
Some authors deny that North Dravidian forms a valid subgroup, splitting it into Northeast (Kurukh–Malto) and Northwest (Brahui). Their affiliation has been proposed based primarily on a small number of common phonetic developments, including:
McAlpin (2003) notes that no exact conditioning can be established for the first two changes, and proposes that distinct Proto-Dravidian *q and *kʲ should be reconstructed behind these correspondences, and that Brahui, Kurukh-Malto, and the rest of Dravidian may be three coordinate branches, possibly with Brahui being the earliest language to split off. A few morphological parallels between Brahui and Kurukh-Malto are also known, but according to McAlpin they are analyzable as shared archaisms rather than shared innovations.
In addition, Ethnologue lists several unclassified Dravidian languages: Allar, Bazigar, Bharia, Malankuravan (possibly a dialect of Malayalam), and Vishavan. Ethnologue also lists several unclassified Southern Dravidian languages: Mala Malasar, Malasar, Thachanadan, Ullatan, Kalanadi, Kumbaran, Kunduvadi, Kurichiya, Attapady Kurumba, Muduga, Pathiya, and Wayanad Chetti. Pattapu may also be Southern.
A computational phylogenetic study of the Dravidian language family was undertaken by Kolipakam, et al. (2018). Kolipakam, et al. (2018) supports the internal coherence of the four Dravidian branches South (or South Dravidian I), South-Central (or South Dravidian II), Central, and North, but is uncertain about the precise relationships of these four branches to each other. The date of Dravidian is estimated to be 4,500 years old.
Since 1981, the Census of India has reported only languages with more than 10,000 speakers, including 17 Dravidian languages. In 1981, these accounted for approximately 24% of India's population.
In the 2001 census, they included 214 million people, about 21% of India's total population of 1.02 billion. In addition, the largest Dravidian-speaking group outside India, Tamil speakers in Sri Lanka, number around 4.7 million. The total number of speakers of Dravidian languages is around 227 million people, around 13% of the population of the Indian subcontinent.
Telugu is the most spoken Dravidian language, with over 74 million native speakers. The total number of speakers of Telugu, including those whose first language is not Telugu, is around 84 million people, which is around 6% of India's total population.
The smallest branch of the Dravidian languages is the Central branch, which has only around 200,000 speakers. These languages are mostly tribal, and spoken in central India.
The next-largest is the South-Central branch, which has 78 million native speakers, the vast majority of whom speak Telugu. This branch also includes the tribal language Gondi spoken in central India.
The largest group is South Dravidian, with almost 150 million speakers. Tamil, Malayalam, and Kannada make up around 98% of the speakers, with Tamil being by far the most spoken language, with almost half of all South Dravidian speakers speaking it.
The Dravidian family has defied all of the attempts to show a connection with other languages, including Indo-European, Hurrian, Basque, Sumerian, Korean and Japanese. Comparisons have been made not just with the other language families of the Indian subcontinent (Indo-European, Austroasiatic, Sino-Tibetan, and Nihali), but with all typologically similar language families of the Old World. Nonetheless, although there are no readily detectable genealogical connections, Dravidian shares strong areal features with the Indo-Aryan languages, which have been attributed to a substratum influence from Dravidian.
Dravidian languages display typological similarities with the Uralic language group, suggesting to some a prolonged period of contact in the past. This idea is popular amongst Dravidian linguists and has been supported by a number of scholars, including Robert Caldwell, Thomas Burrow, Kamil Zvelebil, and Mikhail Andronov. This hypothesis has, however, been rejected by some specialists in Uralic languages, and has in recent times also been criticised by other Dravidian linguists such as Bhadriraju Krishnamurti.
In the early 1970s, the linguist David McAlpin produced a detailed proposal of a genetic relationship between Dravidian and the extinct Elamite language of ancient Elam (present-day southwestern Iran). The Elamo-Dravidian hypothesis was supported in the late 1980s by the archaeologist Colin Renfrew and the geneticist Luigi Luca Cavalli-Sforza, who suggested that Proto-Dravidian was brought to India by farmers from the Iranian part of the Fertile Crescent. (In his 2000 book, Cavalli-Sforza suggested western India, northern India and northern Iran as alternative starting points.) However, linguists have found McAlpin's cognates unconvincing and criticized his proposed phonological rules as ad hoc. Elamite is generally believed by scholars to be a language isolate, and the theory has had no effect on studies of the language.
Dravidian is one of the primary language families in the Nostratic proposal, which would link most languages in North Africa, Europe and Western Asia into a family with its origins in the Fertile Crescent sometime between the last Ice Age and the emergence of Proto-Indo-European 4,000–6,000 BCE. However, the general consensus is that such deep connections are not, or not yet, demonstrable.
The origins of the Dravidian languages, as well as their subsequent development and the period of their differentiation are unclear, partially due to the lack of comparative linguistic research into the Dravidian languages. Though some scholars have argued that the Dravidian languages may have been brought to India by migrations in the fourth or third millennium BCE or even earlier, the Dravidian languages cannot easily be connected to any other language, and they could well be indigenous to India.[note 1] Proto-Dravidian was spoken in the 4th or 3rd millennium BCE, and it is thought that the Dravidian languages were the most widespread indigenous languages in the Indian subcontinent before the advance of the Indo-Aryan languages.
As a proto-language, the Proto-Dravidian language is not itself attested in the historical record. Its modern conception is based solely on reconstruction. It was suggested in the 1980s that the language was spoken in the 4th millennium BCE, and started disintegrating into various branches around 3rd millennium BCE. According to Krishnamurti, Proto-Dravidian may have been spoken in the Indus civilization, suggesting a "tentative date of Proto-Dravidian around the early part of the third millennium." Krishnamurti further states that South Dravidian I (including pre-Tamil) and South Dravidian II (including Pre-Telugu) split around the eleventh century BCE, with the other major branches splitting off at around the same time. Kolipakam et al. (2018) estimate the Dravidian language family to be approximately 4,500 years old.
The Indus Valley civilisation (3,300–1,900 BCE), located in Northwestern Indian subcontinent, is often understood to have been Dravidian. Already in 1924, when announcing the discovery of the IVC, John Marshall stated that (one of) the language(s) may have been Dravidic. Cultural and linguistic similarities have been cited by researchers Henry Heras, Kamil Zvelebil, Asko Parpola and Iravatham Mahadevan as being strong evidence for a proto-Dravidian origin of the ancient Indus Valley civilisation. The discovery in Tamil Nadu of a late Neolithic (early 2nd millennium BCE, i.e. post-dating Harappan decline) stone celt allegedly marked with Indus signs has been considered by some to be significant for the Dravidian identification.
Yuri Knorozov surmised that the symbols represent a logosyllabic script and suggested, based on computer analysis, an underlying agglutinative Dravidian language as the most likely candidate for the underlying language. Knorozov's suggestion was preceded by the work of Henry Heras, who suggested several readings of signs based on a proto-Dravidian assumption.
Linguist Asko Parpola writes that the Indus script and Harappan language are "most likely to have belonged to the Dravidian family". Parpola led a Finnish team in investigating the inscriptions using computer analysis. Based on a proto-Dravidian assumption, they proposed readings of many signs, some agreeing with the suggested readings of Heras and Knorozov (such as equating the "fish" sign with the Dravidian word for fish, "min") but disagreeing on several other readings. A comprehensive description of Parpola's work until 1994 is given in his book Deciphering the Indus Script.
Although in modern times speakers of the various Dravidian languages have mainly occupied the southern portion of India, in earlier times they probably were spoken in a larger area. After the Indo-Aryan migrations into north-western India, starting ca. 1500 BCE, and the establishment of the Kuru kingdom ca. 1100 BCE, a process of Sanskritisation started, which resulted in a language shift in northern India. Southern India has remained majority Dravidian, but pockets of Dravidian can be found in central India, Pakistan, Bangladesh and Nepal.
The Kurukh and Malto are pockets of Dravidian languages in central India, spoken by people who may have migrated from south India. They do have myths about external origins. The Kurukh have traditionally claimed to be from the Deccan Peninsula, more specifically Karnataka. The same tradition has existed of the Brahui, who call themselves immigrants. Holding this same view of the Brahui are many scholars  such as L. H. Horace Perera and M. Ratnasabapathy.
The Brahui population of Pakistan's Balochistan province has been taken by some as the linguistic equivalent of a relict population, perhaps indicating that Dravidian languages were formerly much more widespread and were supplanted by the incoming Indo-Aryan languages. However, it has been argued that the absence of any Old Iranian (Avestan) loanwords in Brahui suggests that the Brahui migrated to Balochistan from central India less than 1,000 years ago. The main Iranian contributor to Brahui vocabulary, Balochi, is a western Iranian language like Kurdish, and arrived in the area from the west only around 1,000 AD. Sound changes shared with Kurukh and Malto also suggest that Brahui was originally spoken near them in central India.
Dravidian languages show extensive lexical (vocabulary) borrowing, but only a few traits of structural (either phonological or grammatical) borrowing from Indo-Aryan, whereas Indo-Aryan shows more structural than lexical borrowings from the Dravidian languages. Many of these features are already present in the oldest known Indo-Aryan language, the language of the Rigveda (c. 1500 BCE), which also includes over a dozen words borrowed from Dravidian.
Vedic Sanskrit has retroflex consonants (ṭ/ḍ, ṇ) with about 88 words in the Rigveda having unconditioned retroflexes. Some sample words are Iṭanta, Kaṇva, śakaṭī, kevaṭa, puṇya and maṇḍūka. Since other Indo-European languages, including other Indo-Iranian languages, lack retroflex consonants, their presence in Indo-Aryan is often cited as evidence of substrate influence from close contact of the Vedic speakers with speakers of a foreign language family rich in retroflex consonants. The Dravidian family is a serious candidate since it is rich in retroflex phonemes reconstructible back to the Proto-Dravidian stage.
In addition, a number of grammatical features of Vedic Sanskrit not found in its sister Avestan language appear to have been borrowed from Dravidian languages. These include the gerund, which has the same function as in Dravidian. Some linguists explain this asymmetrical borrowing by arguing that Middle Indo-Aryan languages were built on a Dravidian substratum. These scholars argue that the most plausible explanation for the presence of Dravidian structural features in Indic is language shift, that is, native Dravidian speakers learning and adopting Indic languages. Although each of the innovative traits in Indic could be accounted for by internal explanations, early Dravidian influence is the only explanation that can account for all of the innovations at once; moreover, it accounts for several of the innovative traits in Indic better than any internal explanation that has been proposed.
The most characteristic grammatical features of Dravidian languages are:
Dravidian languages are noted for the lack of distinction between aspirated and unaspirated stops. While some Dravidian languages have accepted large numbers of loan words from Sanskrit and other Indo-Iranian languages in addition to their already vast vocabulary, in which the orthography shows distinctions in voice and aspiration, the words are pronounced in Dravidian according to different rules of phonology and phonotactics: aspiration of plosives is generally absent, regardless of the spelling of the word. This is not a universal phenomenon and is generally avoided in formal or careful speech, especially when reciting. For instance, Tamil does not distinguish between voiced and voiceless stops. In fact, the Tamil alphabet lacks symbols for voiced and aspirated stops. Dravidian languages are also characterized by a three-way distinction between dental, alveolar, and retroflex places of articulation as well as large numbers of liquids.
Proto-Dravidian had five short and long vowels: *a, *ā, *i, *ī, *u, *ū, *e, *ē, *o, *ō. There were no diphthongs; ai and au are treated as *ay and *av (or *aw). The five-vowel system is largely preserved in the descendent subgroups.
The earliest known Dravidian inscriptions are 76 Old Tamil inscriptions on cave walls in Madurai and Tirunelveli districts in Tamil Nadu, dating from the 2nd century BCE. These inscriptions are written in a variant of the Brahmi script called Tamil Brahmi. The earliest long text in Old Tamil is the Tolkāppiyam, an early work on Tamil grammar and poetics, whose oldest layers could date from the 1st century BCE.