by Tony Joseph
Since we know that the earliest evidence for goat domestication comes from Ganj Dareh in the Zagros mountains from around 7900 BCE, this becomes the upper limit for the separation. Since we also know that cuneiform writing was invented sometime before 3000 BCE, this becomes the lower limit for the separation.
Considering that these linguistic studies appeared years before the genetic evidence of a Zagrosian link to the Harappans was discovered, the close correspondence between the two disciplines on the possible chronology of migrations is striking. As we saw in chapter 2, ancient DNA study puts the latest date for admixture between the Zagros agriculturists and the Harappan Civilization at 3000 BCE – exactly the same as in the linguistic study. The genetics study does not suggest an upper limit, but archaeology has something to say on this. We know the earliest evidence for animal domestication and plant cultivation in south Asia comes from Mehrgarh and is dated to sometime around 7000 BCE – which is almost an exact fit with the chronology that linguistics suggests – especially when you take into consideration the fact that after the separation in the Zagros region estimated after 7900 BCE, it would have taken some time for the Zagrosians to reach south Asia.
What is also remarkable is how old the idea of a connection between Elamite and Dravidian languages is. It was suggested by several linguists long before anyone ever knew that such a thing as the Harappan Civilization existed.
The Harappan Civilization was discovered in the 1920s. But as early as in 1853, Edwin Norris, one of the scholars who helped decipher the Behistun inscription, was pointing out that the Elamite script did not indicate contrastive voicing and worked on the same principles as the Tamil script. Three years later, when Robert Caldwell wrote the first Comparative Grammar of the Dravidian or South Indian Family of Languages, he used Norris’s work to put forward Elamite as the first suggested affiliation for Tamil, and discussed the connection between the two languages at length. Since then, many others have mentioned or expanded on the relationship between Elamite and Dravidian languages, including Alfredo Trombetti and I.M. Diakonoff.
But the first substantive and rigorous look at the subject was a lengthy paper written by McAlpin in 1981, titled ‘Proto-Elamo-Dravidian: The Evidence and Its Implications’. This was a tour de force, though many linguists at the time were cautious in accepting it. This was not surprising since Elamite had been extinct for over two millennia and the last written Elamite records were available only from around 450 BCE. There aren’t too many word roots that can be reconstructed for Proto-Elamite based on the limited supply of such old writings. Moreover, Proto-Zagrosian and Proto-Dravidian may have separated as early as 7000 BCE to 6000 BCE, going on to evolve along their own separate pathways since then. So to establish a relationship between the two languages based on reconstructed Proto-Elamite or Proto-Zagrosian was a dauntingly difficult task to begin with.
But as McAlpin says in his magisterial work, it is this very difficulty that makes the extent of similarities surprising. Out of the 250 word roots that McAlpin was able to identify for Elamite, over 40 per cent are cognate with, or similar to, Dravidian. With such a high rate of common root words which is ‘emphatically above chance levels’, says McAlpin, ‘the basic case becomes an established prima facie one’.
Since the paper was published in 1981, the evidence has only accumulated further, and the 2013 paper by Southworth and McAlpin is evidence of that. With the new ancient DNA discovery and the existing archaeological evidence connecting the Zagros region to the agricultural beginnings in south Asia, the conclusion of McAlpin’s 1981 paper now is not just compelling, but unavoidable.
Here’s a sample of the kind of lexical connections that McAlpin uncovered. Needless to say, these would seem remarkable to anyone familiar with the Dravidian languages. These are ten of the eighty-one words that McAlpin has listed in his paper.
Proto-Elamo-Dravidian: *hit.- (to herd goats, goat); Achaemenid Elamite: hidu (adult female goat, goats in general); Proto-Dravidian: *it.- (to herd goats); Tamil: it.ai (herdsman caste); Malayalam: it.ayan (a caste of shepherds and cowherds).
Proto-Elamo-Dravidian: *pot (young animal); Achaemenid Elamite: putu (lamb); Proto-Dravidian: *pot (young animal or plant); Tamil: pottu (sapling).
Proto-Elamo-Dravidian: *vari- (to fix, tie, hold); Middle Elamite: mari- (to seize, grasp, capture); Proto-Dravidian: *vari- (to bind, tie, fasten); Tamil: vari (to bind, tie, fasten).
Proto-Elamo-Dravidian: *um- (to process grain); Achaemenid Elamite: umi- (to grind grain); Proto-Dravidian: *um (husk, chaff); Tamil: umi (husk, to become chaff); Telugu: umaka (husk, chaff).
Proto-Elamo-Dravidian: *ni (you); Old Elamite: ni (you); Proto-Dravidian: *ni (you). This is a basic term attested in all Dravidian languages and means exactly the same – you.
Proto-Elamo-Dravidian: *nal. (day); Middle Elamite: na, nana (days); Proto-Dravidian: *nal. (day).
Proto-Elamo-Dravidian: *tol.- (to perforate, bore); Middle Elamite: tullin (breach, cut, slice); Proto-Dravidian: *tol.- (to perforate, bore); Telugu: tolucu (to bore, perforate).
Proto-Elamo-Dravidian: *cah- (to die); Middle Elamite: sa- (his life should be cut off); Proto-Dravidian: *caH- (to die); Tamil ca(k/v)/ce (to die, be blighted); Malayalam: cakuka/ca- (to die).
Proto-Elamo-Dravidian: *ul (inside, interior, mind, heart, to think); Middle Elamite: ulhi (dwelling place, residence, sanctuary); Tamil: ul.l.am (mind, thought)
Proto-Elamo-Dravidian: *kat. (bed, throne); Royal Achaemenid Elamite: kat (place, throne); Proto-Dravidian: *kat.t.il (cot, bedstead, throne of distinction); Tamil: kat.t.il (cot, bedstead).
Correspondences between the two languages in terms of word roots is only one part of McAlpin’s work; the other and equally substantial part of his work looks at similarities in terms of morphology and other elements that go into the making of any language.
There is yet another striking piece of evidence that suggests that the Harappan language was Proto-Dravidian, but which McAlpin doesn’t mention because it deals not with Elamite, but with Akkadian, one of the dominant Semitic languages of the Mesopotamian Civilization. In chapter 2 (p. 103), we saw there was one plant that was imported from the Harappan Civilization to Mesopotamia: sesame. Remarkably, the name of this oilseed is the same in Akkadian and in many Dravidian languages of south India today: ellu. Writes Professor Southworth in Linguistic Archaeology of South Asia, ‘the sharing of a word for sesame between Mesopotamia (Akkadian ellu) and South India (South Dravidian 1: ellu) reinforces the hypothesis of the presence of Dravidian speakers in the prehistoric Indus Valley’.
What about Indo-European languages that today dominate the region that once belonged to the Harappan Civilization? Could not these languages have been spoken there during the Harappan period? We will come to that in chapter 4, but briefly, the generally accepted chronology for the spread of Indo-European languages around the world puts their arrival in south Asia only after 2000 BCE, when the Harappan Civilization was already in decline. This has been strongly supported by recent genetic studies based on ancient DNA which suggest that Steppe pastoralists, who took Indo-European languages to Europe, reached India only in the Late Harappan period, bringing with them an early version of Sanskrit and related cultural concepts and practices such as ritual sacrifices. Many of these practices and beliefs are reflected in the Vedas, especially the earliest one, the Rigveda. The newly arrived Indo-European-language speakers called themselves ‘Aryans’. What all of this means is that the Harappan Civilization had nothing to do with the ‘Aryans’ or Sanskrit or the Vedas, and was pre-‘Aryan’ or pre-Vedic. We will discuss this further in the next chapter.
Deciphering the Harappan script
It is no surprise then that most of the attempts at deciphering the Harappan script have assumed that the language underlying it was Dravidian. Even though the script has not been deciphered after nearly a century of concerted attempts, the efforts of two independent experts in particular have been noteworthy: Dr Iravatham Mahadevan, th
e epigraphist who helped decipher the Tamil Brahmi script in which the earliest records of Dravidian are kept, and Professor Asko Parpola. Each of them brought new perspectives on how to start reading the script.
Until the script is deciphered, it cannot be used to either buttress or weaken the common conclusions being reached by archaeology, genetics and linguistics. However, it is worth quoting Mahadevan’s convocation address at the Dravidian University in Kuppam, Andhra Pradesh, in 2015, for two reasons. One, his strong rebuttal of the belief that the Harappans spoke an Indo-European language and that they were Vedic Aryans. And two, his description of the direction in which the deciphering efforts are tending:
The results I have obtained so far confirm that the language of the Indus script is an early form of Dravidian. I do not claim to have deciphered the Indus Script completely. But I sincerely believe that I have discovered important clues for interpreting many of the frequent Indus signs and sequences proving conclusively the Dravidian character of the language and the survival of the Indus elements in the twin streams of later Dravidian and Indo-Aryan traditions . . .
There is substantial evidence that the Indus Civilization was pre-Aryan.
The Indus Civilization was mainly urban, while the early Vedic society was rural and pastoral. There were no cities in the Vedic period.
The Indus seals depict many animals but not the horse. The horse and the chariot with spoked wheels were the defining features of the Aryan-speaking societies. The bronze chariot found at Daimabad in Western Deccan, the Southernmost Indus settlement, has solid wheels and is drawn by a pair of humped bulls, not horses.
The tiger is often featured on Indus seals and sealings, but the animal is not mentioned in the Rigveda.
Mahadevan goes on to enumerate the substantial archaeological and linguistic evidence that supports the Dravidian nature of the Harappan Civilization. The evidence includes pictorial depictions on seals and sealings that suggest the worship of a buffalo-horned male god, mother goddesses, the peepul tree, the serpent and, possibly, the phallic symbol, all of which have been derived not from the earliest Vedas, but from the pre-Aryan population. Many of these went on to become part of the Indian cultural tradition as we know it today, and this is a crucial point Mahadevan makes. ‘The Indus heritage is shared by Dravidian as well as Indo-Aryan speakers. The Dravidian heritage is linguistic. The Indo-Aryan heritage is cultural, preserved through loanwords [words taken from another language], loan translations [phrases taken from another language], and myths . . . As I read it, the message of the Indus Script is: unity in diversity.’
In other words, after Indo-European-language speakers reached south Asia, the language of the Harappans became limited to south India, while the culture and myths of the Harappans melded with those of the new Indo-Aryan-language-speaking migrants to create a unique, syncretic tradition that is today seen as an essential part of Indian culture. Therefore, there is a disconnect between the earliest Vedas and the culture and practices of the Harappan Civilization, but a connect between the later Vedic corpus and the Harappan Civilization because these by then incorporate some of the ideas and themes of the Harappans. It is thus possible to see the heritage of Harappa in the language/s of the Dravidians, and in the myths, phrases and words borrowed by the Indo-Aryans from the Harappan tradition. (See section titled ‘Harappans and the Vedas: Disconnect and connect’ in chapter 4 for details.)
A figurine of a woman from the Harappan Civilization. Some see the figurine as a ‘Mother Goddess’.
Mahadevan identifies grammatical signs in the Harappan script that stand for the masculine singular suffix and also the non-masculine singular suffix. He identifies the arrow sign that frequently appears in the script as ‘ampu’ – the Dravidian word that means ‘arrow’, but also stands for the non-masculine singular suffix in the earliest Kannada, Tamil and Telugu inscriptions, which means it appears at the end of female names.6 Similarly, he identifies the frequently occurring jar sign in the Harappan script as the masculine singular suffix ‘anru’ that is often appended at the end of male names.
Mahadevan’s most striking observation about the Harappan script is that many of its signs refer to the meticulously planned Harappan cities with their grid-like streets oriented towards the four cardinal directions and fortified citadels built on a high artificial terrace. For example, he identifies the square sign with a smaller square inside it as referring to an official title, ‘The Lord of the House’, or ‘akatti’ in Dravidian, drawn from the root word ‘akam’, meaning inside the house/palace/fort/mind. The square itself, in this interpretation, would refer to the citadel or fort, and the small square inside it to its main resident. In Old Tamil, the word ‘akatton’ means ‘The Lord of the Fort’. Mahadevan suggests the name Agastya comes from ‘akatti’ and that the legend of Agastya in Old Tamil literature – where he is thought to have led eighteen kings and eighteen families of the Velir clan from north India to south India – could refer to the rulers of the Harappan Civilization who migrated to the south after the civilization declined.
By linking many signs in the script to Harappan cities themselves, he also shows how the Cheras, Cholas and Pandiyas (three Tamil dynasties with a long history spanning centuries before and after the Common Era) may be identified by the roles they played within the Harappan urban civilization. He says: ‘The Chola were counsellors “surrounding” and “advising” the Ruler. The Chera or Cheral were high officials residing in exclusive quarters (keri) with restricted entry. The Panti were the commoners who resided in the streets, “pati” of the (lower) city, “pali”. These results corroborate the folklore that the Chera, Chola and Panti were brothers who lived together in one place in ancient times.’
This is a fascinating interpretation, but until more of the script is deciphered, we will not know for sure whether it is accurate – even though in Egyptian hieroglyphics there are very similar signs, with rectangles rather than squares, that carry similar meanings: house and fortified house. Also, the title of the ancient Egyptian rulers, Pharaoh, literally means ‘Great House’.
If Mahadevan used the Harappan city structure to decipher meanings in some of its inscriptions, Parpola turned to astronomy to find meanings in other parts. His argument is that astronomy was important and prevalent during the Harappan Civilization, just as in the Mesopotamian Civilization, especially considering the way the cities were built along cardinal directions.
Both Mahadevan and Parpola begin with the assumption that the seals are likely to bear proper names of persons, with or without their official titles, just as they did in Mesopotamian seals. About thirty Harappan seals have been recovered from the Gulf and Mesopotamia, probably left there by seafaring Indian merchants, thus proving the importance the seal played as a way of guaranteeing the true provenance of merchandise. Both Mahadevan and Parpola also think the culture that followed the decline of the Harappan Civilization, after the arrival of Indo-European-language speakers, assimilated a number of Harappan concepts and cultural practices. Therefore, they use later traditions in Dravidian languages such as Tamil and Indo-European languages such as Sanskrit to substantiate their readings of the script.
The most striking of Parpola’s decipherments relates to the frequently appearing sign of the fish in the Harappan script. In Parpola’s reading, the fish sign in the Harappan script refers to ‘miin’, the Dravidian word for both fish and star. Parpola argues that the frequency of the sign in the Harappan script – about one in ten signs is either fish or a modified fish – suggests that its most common reference is to star rather than fish.
Therefore, a sign with six vertical strokes followed by a fish could mean ‘aru’ plus ‘miin’ – with ‘aru’ meaning six in Dravidian, and ‘miin’ meaning star or fish. As it happens, there is indeed a constellation that is called ‘aru-miin’ – the Pleiades. There are also Harappan inscriptions that show seven vertical strokes followed by a fish – which would read as ‘elu-miin’ in Dravidian, meaning seven star
s, which refers to the constellation Ursa Major.
According to Parpola, both the Pleiades and Ursa Major play an important role not only in early Indian mythology but also in the early history of Indian calendrical astronomy, which is probably of Harappan origin. The suggestion that the fish sign in the Harappan script could mean star predates Parpola, but he has built on the idea so extensively that he has made it impossible for anyone to ignore it.
A seal depicting a unicorn. The inscription shows seven vertical strokes followed by a fish which could refer to the constellation of Ursa Major.
The route to the south
If the dominant language of the Harappans was Proto-Dravidian, when did it reach south India from north-western India? The most common assumption is that this happened after the civilization declined and new migrants speaking Indo-European languages – or Aryans – arrived in north-western India sometime after 2000 BCE. This is the event that the Agastya or Akatti legend perhaps refers to. But Agastya and his Velir clans need not have been the first to bring Dravidian languages to south India, though they may have brought a more urbane form of it.
In fact, there is reason to think Dravidian languages may have reached south India perhaps as early as around 2800 BCE, even before the Mature Harappan period began in 2600 BCE. This is around the time that we see the first evidence of pastoralism in south India, in northern Karnataka. Those who brought Dravidian languages to south India first may not have been the urbanites of the Harappan Civilization but pastoralists, some of whom may remain so even today, like the Brahuis of Balochistan. As we saw earlier, those who arrived in south Asia from the Zagros region were probably herders rather than full-fledged farmers, and while some of them settled down as farmers (perhaps after mixing with indigenous south Asians who were already experimenting with farming), others may have remained pastoralists. There is a possibility that some of these pastoralists may have brought cattle and goat herding to south India and they could have been the first to bring a Dravidian language here too.