The Indus Script and Brahmi: Ancient Indus Valley Script Iravatham Mahadevan Interview

4. The Indus Script and Brahmi

Q: What do you think the relationship is between the Indus script and the Brahmi script, since you know both of them? 

A: Several scholars have said that there is a relationship between the two, that the Indus script survived and slowly became linear and ultimately lead to the Brahmi script. I do not at all believe in this theory. The Indus script was in existence not later than at the most about 1500 B.C. The earliest undisputed examples of the Brahmi script are only from the days of Ashoka, around 300 B.C. One might take the origin of the Brahmi script still farther, to the beginnings of the Indo-Gangetic, Iron Age civilization, in the middle of the first millennia B.C., since Ashoka does not claim to have invented the Brahmi script, it is not unlikely that the Brahmi script was known before his times, and perhaps used by the merchants commercially as the Allchins have suggested in their recent book. The absence of such inscriptions by Chandragupta, the illustrious grandfather of Ashoka could be explained by saying that stone inscriptions were not in the Indian tradition and they came to us along with the Persian tradition. This is not unlikely, but even so there is at least a gap of 1,000 years before the introduction of the Brahmi script and the complete collapse of the Indus script. 

There is a parallel in the history of writing elsewhere in the history of the world. Mycenian Linear B script was written in syllabic script in about the 14th or 15th century B.C. and that has nothing to do with the later script of the Greeks, which was taken over from the Phoenicians. There again was a gap of one thousand years. Personally, I believe that the Indus script was too closely tied up with the Indus language, whatever it was, and when that language ceased to be spoken and became dead, the incoming Aryans could not use that script. Now that again has a parallel in the Egyptian script which was too closely tied up in the Egyptian language and could not survive it. The cuneiform scripts were much more adaptable to a wide variety of languages. So perhaps the logographic Indus script had a one-to-one relation with the words of the Indus language and could not be used in another language. What has survived of the Indus script may be symbols of various kinds, totem signs, royal signs and insignia on punch marked coins and flags and traditions in our mythology of gods, attributes, weapons and so on but not as a writing system. 

Another reason to say that the Brahmi script is not related to the Indus script is that the connection between Brahmi and some form of Semitic script is too strong. Buehler pointed out the relation between Alif and A, B and Bay, Gameen and Ga, and so on. At least I can see about 10 of the 22 Semitic characters very closely resemble Brahmi both in form and sound. Statistically, such a resemblance is impossible except when there is genetic relationship. I do not say that Brahmi script itself came from the Semitic script, but some elements of the Semitic were taken over, others were locally added, improvements were made, the order of the sounds were changed, the diacritical marks were locally invented, the aspirates were invented, the additional vowels were joined, so that Brahmi is a much developed and transformed script. But the idea of the Brahmi script comes from the alphabetic, Semitic script and I believe in Buehler's theory. For these two reasons I do not agree with the scholars, most of them Indian, who believe that the Brahmi script is a remote descendant of the Indus script.

5. One or Many Languages ?

Q: Do you believe that the Indus script represented a single language or could it have represented a multiple number of languages? 

A: This is a very interesting question. I will give you the parallel of the Chinese language. The Chinese have only one script but they have many mutually unintelligible dialects. Such a situation might have existed in the far flung areas of Harappa, but I believe basically that the Harappan script was intelligible to all the people of Mohenjo-daro and Harappa. If there were differences, it would have been more diacritical than linguistic. 

The very strong reason for this was the one found out by Hunter long ago. Hunter pointed out that the frequency and combination of signs at Harappa and Mohenjo-daro are the same, showing that the language used there must have been the same - subject to what I said earlier that there might have been diacritical differences. If you take English and French, both are written in a form of the Roman script but the frequency and combination of characters in the two languages is different. Therefore, all over the greater Indus Valley the language was the same and it was the same throughout the civilization. There is one major exception to this and that is very interesting, again pointed out by Hunter, and recently restated by Parpola and his colleagues with more evidence. 

In some of the Indus Valley seals found in the Middle East, particularly the round seals which must have been locally manufactured, the order of the signs and their combination are totally dissimilar to what we find in Harappa and Mohenjo-daro., This may have been an attempt to use the Harappan script by the natives of the Indus Valley who went over to the Middle East for trade purposes, to adapt the Indus script to a local Middle Eastern language. But for this exception, within the Indus Valley itself, in all its areas and throughout the time, the language was the same as proved by the frequency of the Indus signs.

6. Tissue of Indus Civilization ?

Q: To have the same language over such a wide area and time, what does this imply about the political or social organization of the culture? 

A: To me one conclusion is irresistible. It is not a migrant civilization, it is not that a handful of settlers came and settled on the sea coast. This is a large, native, indigenous civilization. It is surprising that people hardly realize the extent of the Harappan civilization. It was more than a million square kilometers in area, much larger than modern Pakistan, much larger than all the other ancient civilizations, excepting China of course, put together. The Sumerian, the Akkadian, the Egyptian, Hittite and so on. Over such a large and fairly populous area, judging from the number of villages and cities. Several estimates of the population of Harappa and Mohenjo-daro have been made and they seem to have been very large cities by ancient standards. This only goes to confirm our supposition that you must look for a local language as a candidate for the Harappan script. 

One might ask, could not those people have totally vanished? 

It has happened in history before, the Sumerians became totally extinct. But for the accident of their script having been taken over by the Akkadians, the world might never know of the existence of Sumerians. But here again, the scale and the magnitude of the Harappan civilization speaks against its total extinction. As all scholars who have studied the problem agree, the incoming Aryans were relatively a very small minority and they were able to dominate only culturally and ultimately, in the assimilated Indo-Aryan or north Indian people, the indigenous racial element must have slowly surfaced. That is why we have no such thing as early Aryan pottery, because the pottery continued to be made by the local people. As someone has said jokingly, archaeology knows of no Aryans, only linguistics knows of Aryans. This is true. The answer to this is that the incoming Aryans were small in number. In this respect there was no cultural discontinuity. The real discontinuity was in language, principally, and in religion and ritual in the earliest levels, but in later levels, modern Hinduism as we know it is a composite of both pre-Aryan, native, animistic and tribal religions and the incoming Aryan religion. Perhaps when the Indus script is deciphered, I would not be surprised to find that the greater part of modern Hinduism has a Harappan lineage. 

Q: You have such an extensive civilization that does not seem to been very militaristic, it does not seem to have had supreme authoritarian rulers, the fortifications seem to have been as much against floods or from cultural accretions as for defensive purposes. What would have been the tissue that held together such a large group of people for so a long period of time? 

A: It is not quite true that they were not militaristic at all, except in the sense that they didn't seem to have been aggressive. But they had defense fortifications, they had weapons, spears, swords and the large terra-cotta stone slingballs which they stored on their fortifications. But it appears that this very large civilization was peaceful and for a long time did not fear any enemies from outside. What held them together could have been a dominant priestly elite, binding the people together with the help of religion, religious practices as has happened in later Hinduism where millions of Hindus follow the dictates of religion not by force but by persuasion. This could have been so. Here again, I am treading on fairly dangerous ground, for there are scholars who say that there is no proof that the Indus Valley was ruled by priests. This is true only in a very narrow sense, in the sense that the script has not been deciphered. But if you look at the seals, you can clearly see priestly rituals, naked priests offering worship to naked, horned gods. At least one sealing shows a woman being beheaded by a sickle, it could be an example of human sacrifice. You have examples of worship of the peepal tree. 

Scholars have said there is no evidence for kings. Well, we have no large graves, we have no kings buried with their queens and their chariots in position, we have no large collection of crowns and jewels. This may be a reflection of the types of rulers they had. 

The native tradition of India was largely one of janapadas, before the Mauryan Empire, before even the 6th century B.C. Before that time, north India was divided into a large number of republics, janapadas; both Mahavira and Buddha belonged to republican clans. It is quite possible that the Harappan states were earlier examples of janapadas. I know I am speculating, but one has to explain the absence of royal graves and royal paraphernalia. It could be they had elders, they had ganas, they had republican, a crude republican form of government, they had priestly oligarchies and the chief went by rotation. All these things are not impossible, and for the keys to these one must look at the surviving institutions. The republican janapada tradition could be an important element in judging the absence of royal paraphernalia in the Harappan polity. 

Q: There is no royal paraphernalia in those traditions? 

A: None at all. There are no great palaces, no royal graves, no evidence of a large standing army. In this respect the Harappan civilization was very unique. Now these are problems crying for answer and unfortunately the Indus script remains unread. 

But one thing is clear. Even if the script is read, it is not likely to throw much greater light on what we already know archaeologically. The linguistic question would be solved, and that is very important. But in the absence of chronicles, legends, epics, connected accounts of the type you have the Akkadian, Sumerian and other Near Eastern civilizations, we have no accounting tablets, we have no long connected works. We know that Sumerians had advanced mathematics, they could calculate eclipses and they had epics like that of Gilgamesh. These are all not available in the Indus Valley. To that extent perhaps, this civilization is distinctly mute. 

7. Writing on What ?

Q: Don't you think they may have written on other objects, like palm leaves or cloth? 

A: This has been suggested by Parpola and others, that they probably wrote on cloth or on leaves like birch leaves. I do not know about the palm leaves, palm trees were there of course, but if they had written on palm leaves and they had used a bronze stylus, these must have survived. Nor was there any writing on clay except seal impressions. But perhaps on large barks or on prepared cloth these could have written and this would have perished in the humid and warm climate of the Indus Valley. 

Q: But is it likely? For a civilization that lasted so long, they would have seen that these writings on cloth or something like that perished, probably rather quickly. Wouldn't they then have tried to write more important things on stone or clay? 

A: The same thing has happened later. Why did not the great and victorious Emperor Chandrgupta who stopped the Greek armies from invading India - why did he not like King Darius record his conquests in stone? The answer seems to be one of culture, it was not in the Indian tradition to make inscriptions on stone and the earliest such inscriptions are those of Ashoka. They clearly show Persian influence. This could be one answer.

The other thing was that in the river valley civilization, in modern Sindh proper, there is not much stone, it is available only in the highlands. In Harappa and Mohenjo-daro brick was the dominant material for building, so stone inscriptions would not have been a natural choice. But they could have used clay inscriptions of course, and they knew that clay was being used in the contemporary Akkadian civilization and they did not do that. They could have used metal for example, since they did write on metal and we have a large number of copper tablets from Mohenjo-daro. Well, this is all some of the might-have-beens of our history. Even the recent Dholavira [Gujarat, India] find - I was disappointed that the very large, nearly 10 feet wide, wooden board contained really nothing but a magnified version of a seal. In fact, I have identified all the 10 characters in that famous board as occurring on seals already in the same sequences. So even the opportunity of a large writing surface was missed. How I wish it contained a very long narrative of the conquest of that city by a warrior! 

8. The Fish Sign

A: Lets come to the specific signs. What do you believe may be some of the best interpretations offered of certain signs? 

A: Like all Dravidian scholars, I too began with Father Heras. Father Heras was a Spanish Jesuit priest who worked in India and was a celebrated Professor of History in Bombay. It was his brilliant idea that the fish sign in the Indus script represented the word for fish in all the Dravidian languages, which is "meen," and he pointed out that the word "min" also represented a star or planet in all the Dravidian languages. He said that perhaps the Harappans used the fish sign to represent a star or a planet. This is really the starting point for decipherment for all the Dravidian scholars who followed him, the Russians, the Finnish and myself. Only Fairservis broke away from the tradition, but his identification of the fish sign as a loop or a knot in rope is very unconvincing. I have seen far too many seals and sealings with realistic, life-like fish symbols, there is no doubt at all that the sign represents the fish. 

But another and more valid objection is, why wouldn't they pictorialize the star as a star? Draw five or six lines and add an asterisk mark - that's how the Sumerians, the Akkadians and the Chinese represented a star. The theory behind pictorial writing is that you use pictures to represent the sound of objects that are difficult to draw. In an example given by Parpola himself, "can" in the noun form is a container, in the verb, I "can" do it - that cannot be written as a picture. But in the case of a star it is much easier and it occupies much less space to draw the picture of a star than a fish. Parpola has given a reply to this, not perhaps wholly convincing, but I still think that the fish-meen-star homophony is a good one, although I readily admit that it has not been proved. That could only come if the word "meen" was written elsewhere syllabically or if you have a bilingual. 

For example we have proved the direction of the Indus script. It is no longer open to debate. Those who read the Indus script from the left, their work is condemned to failure right at the beginning. The fish hypothesis is not that conclusively proved, but it still is a very attractive one. 

There are some corroborative details. The numbers three, six and seven before the fish correspond to the well known asterisms, three-fish in the warrior constellation, six-fish for Pleades, seven-fish the Great Bear and so on, but then when you come to the diacritical marks over the fish symbol which Parpola reads as the names of several planets, it is much more open to question. Diacritical marks are very tiny little tick marks and they are not inherently pictorial so any hypothesis about them is only arbitrary. 

9. The Terminal Sign

As regards the other signs, the position is even weaker. There is the famous terminal sign, the most frequent sign, which occupies ten to twelve percent of all Harappan writing. This sign, popularly called the jar sign, is as popular as the letter "e" is in the English language. 

Q: What are your thoughts about the terminal sign of the Indus script? 

A: It is one of the most important signs. The problem is there is a dispute about what the object itself is a picture of. Hunter said it was a vessel, with long lips or handles. B.B. Lal, the famous Indian archaeologist has a very good paper on it, where he proves that some of the variants of this sign very closely resemble jar forms found at Kalibangan. My personal view is that particularly the representation of this sign freely drawn on pottery and copper tablets leaves little room for doubt that it is a container. 

Parpola now thinks that it represents the front view of a bull or a cow with curved horns or a face. There is some parallel, it is not a very far-fetched idea, but the symbolism of the jar is much nearer. I am influenced by the fact that in the Indian tradition the ruling classes, the princes and the priests, always claim to have come from a jar. The jar-born elite is a very famous old Indian symbolism starting all the way from the RgVeda, where Agastiya and Vasishta are supposed to have been born out of jars. The major rishis were jar-born. So I have tried on an ideographic, symbolic level to connect the two, but I have sufficiently emphasized in my writing that this is not decipherment. When the script died, the most important user of that symbol retained it in his mythology. Who was he? Invariably a priest or a ruler. Therefore through the pathway of Indian mythology I arrive at the conclusion that the jar sign represented the ruling elite of the Indus Valley civilization. 

But within the Indus script itself it might have performed the function of a grammatical suffix. It could have been a nominal suffix, used only by the elite. Very early in my work I had tried to find the phonetic value of that sign, but for many years now I have given up working on phonetic signs because there is insufficient evidence to arrive at the phonetic value of any sign so far. I occasionally still try to do so, however.

Q: Parpola's argument that it is a representation of a cow has some appeal to me on the level of looking at cows straight on. It also seems to me that there must be some deep connection between the cow and Indus culture. Even in Harappa today cattle thieving, cattle chasing remain popular. It is changing to the buffalo, but in those days it was cattle and they must have been represented somehow. 

A: Well, you have plenty of that in the animal symbolism of the seals. You have the so-called unicorn which may be a side view of an animal with two horns, one behind the other. You have the magnificent bull with the hump, or a short-horned bull always in an aggressive mood. There is no doubt that the Indus Valley people thought along these lines and could have worshipped animals, or the animals could very well have been their totem signs, one group having the unicorn, most probably the ruling group, another the elephant, another the rhinoceros and so on.

Whether the jar sign is a front view of a cow I have my reservations, but I may be influenced by my own connecting this with the mythological tradition, so I cannot claim to be a disinterested witness. Nevertheless, looking at the pictorial parallels provided by Parpola himself in his book I am not convinced - in which case at least occasionally some seal or another should show a greater fidelity to a bull or a cow's face. The two parallel lines on top of the jar are parallel, they are not horizontal or slanted. If they are slanted they are slanted downwards, they do not look like horns at all to me. 

Similarly many years earlier the Russians said it represented the front view of the hull of a ship. Well, they have not been pursuing this. Those who are not dealing with the decipherment of the Indus script like B.B. Lal all say it looks like a jar. It could be an agricultural [storage] jar, or a ritualistic jar as I believe. It does not appear to be the same as the perforated jar, but I think that it is some kind of a jar which is related to the ruling elite. But in the Indus Valley itself it would have primarily a phonetic value. 

There is also a famous sealing where you find this particular jar with a lid on it shown. About half a dozen examples are available of this jar sign which I think Parpola has overlooked. But the argument is still unsettled. Here again I do not think how the question will be settled without a bilingual or a syllabic version of the Indus script being spelt out and a jar being replaced by two symbols which could be read. But that is still in the future.

10. The Arrow Sign

Similarly the twin sign of this jar sign is what is called the arrow sign, or the lance sign. It is a twin functional in the sense that both these signs occur at the end, almost always after other signs which may represent names, so therefore it is another type of grammatical suffix. But one view is, like the one presently held by Parpola, that one is genitive and one is dative. I think this is unlikely because in which case you must occasionally have this symbol following the same names. But the use of the jar sign and the arrow sign are generally mutually exclusive; where the one occurs, the other never does, showing they are integrally connected, semantically related to the names which precede those symbols, which would rule out an explanation like a case ending. However, the position is that this is all still in the realm of speculation. No one has any hard evidence. I have suggested that the jar sign represented the elite who later developed the myth of the jar-born sages and jar-born rulers. I am not suggesting that the myth existed, even in the Indus civilization, but it developed into a myth later. 

Let me give you a pictorial example. Let us assume that English is a pictorial language and all English gentlemen who called themselves "Squire" wrote a square sign after their names. Then, centuries later, the myth would develop about English gentlemen and their connection with squares, that their houses were squares, or their temples were square. In the long development of the Indian civilization, the jar sign acquired a myth, so when I deal with it I am looking not at the jar sign as it was understood in the Indus Valley but as it was understood centuries later. The survival would be among the same groups. In other words, if the Vedic rishis claimed special affinity to the jar then the Harappan priests had something to do with the jar, broadly speaking. It is not a linguistic argument, it is a cultural argument. Similarly the arrow sign may have to do with warriors, and a sign with shwoing a plain simple man may be a servant. 

11. Signs and Castes

Q: What about the signs where you have some very convincing thoughts, the trader and so forth. The logical basis that traders would have to be represented in some way on seals which were meant for trade makes sense. [See "Dictionary" for Mahadevan's speculations]

A: It is a moot question whether the Harappans had castes like we have in later India. But they certainly must have had occupational groups, priests, scribes, traders, warriors, why not? Perhaps they were not very rigid divisions. In any case, even if these terminal ideograms represent different groups, we must remember that the ideograms themselves are combined - for example the symbol of a bearer is combined with that of a jar, and that is followed by a harrow sometimes. So they could not have referred to exclusive caste groups like we understand the term. The possibility is that originally these were all mere phonetic symbols, but later the groups whose names used these symbols most, they only had surviving symbols with the loss of the language, so the symbols became mythical symbols or representations of their own cult objects. It is a jungle really, very difficult to trace out the environment in India which is not only multi-cultural, multi-racial and multi-linguistic. 

The Dravidians and the Indo-Aryans interacted with each other and we have the Indo-Aryan languages with Dravidian features and we have Dravidian languages with Indo-Aryan features. Hinduism has both Aryan and Dravidian elements, not to speak of modern Hinduism also being influenced by Islam and its own native offshoots like Buddhism and Jainism. There are no clear cut parallels. 

What is more, the same symbol in the Indus valley could have taken multiple forms later. The jar sign is one example. Kunda is both jar and the fire pit. In later India, there are communities which claim to have arisen from the jar, and other from the fire pit. Now both have the same word and both could be two very different responses to the same common tradition.

12. Parpola's Work

Q: What are your thoughts on Parpola's work, its importance and where you disagree with him. 

A: Parpola's work is without doubt the most valuable contribution to date in the field of the Indus script. As I have mentioned in my latest paper, his work transcends linguistic boundaries. His contribution in publishing the first concordance, the first computer studies, the UNESCO volumes of seals are very great. One of his biggest achievements now is a standardized sign list which is so comprehensive that it replaces all other sign lists, including the one prepared by me more than twenty years ago. He has also meticulously recorded every little variant of every sign, running into thousands of variants. He has in his book, while it deals with the Dravidian hypothesis, also given a formal analysis of the Indus script, its functional character, frequency distribution analysis, syntactical analysis, how an Indus sentence could be broken up into slots and so on. Therefore, while I still consider that Parpola's latest decipherment to be not wholly successful, therein might lie the seeds of future decipherments. There are many, many things he says which make sound sense, like his emphasis on contacts with the Near East, his proof as to why a Dravidian language is involved. He has marshaled all the arguments in favor of a Dravidian hypothesis which I find convincing. I would say he has laid the groundwork for a successful decipherment which may happen any day now or maybe later. The world of scholarship recognizes that Parpola's is the most serious work done to date on the Indus script.

Q: One of your charges against him is that he emphasizes too much the religious nature of these various signs. First I thought that was a good point. But Mark Kenoyer makes the point that in Islam, in Christianity, Hinduism, and so on, names and religion are closely intertwined. Most names have a religious origin. Doesn't that make your charge against Parpola somewhat weaker? 

A: It is not a charge but an observation. I agree that the Harappan names could be religious and in that sense the contents of the seals could be religious, as were the Near Eastern seals. But religion is not the only thing. For example, when Parpola says the fish sign is not only a phonetic symbol, it is not only a representation of a star or a planet, but that it is a god in itself - he calls the fish sign a god and it has values far transcending its phonetic values - well, that seems to me to be overstating the case. Further, when you look at the sum total of his decipherments, it's all gods - you have Murugan, you have planets, you have stars. Okay, the Harappans had gods and they could very well be these. I also believe that the fish-star parallel may have something to it. But then where are the Harappan people in his decipherment? Where is the scribe? Where is the ordinary petty government official, the tax collectors, the warriors, the sailors? Take Near Eastern seals: X, son of y, gave this. Where is son, daughter, wife, husband, father? Where are those little cementing particles of language without which you cannot write? I deal with Indian inscriptions. Even the briefest of the Brahmi inscriptions, both in Prakit and Tamil, cannot do without a modicum of grammar. They cannot just be a string of nouns. You will have to have "and," conjunctions, "of," genetive case, "to," dative case. You have to have verbal participles. A language which consists of nothing but a string of names and the names all being gods to me seems very unlikely.

We know that the Harappan civilization was a very advanced urban civilization and it must have employed a large army of bureaucrats because it has such standardized weights, bricks, standardized versions of everything. So there must have been efficient administration of everything, tax collection and civic and sanitary affairs. Almost everyone has a seal - the seals have a boss at the back and it has a hole to be worn around the neck. So the holder must have mentioned not only his name, but also his calling, his profession. Unless we have a word for priest, king, noble, tax collector, scribe, unless we have a little more of what I call ordinary, day-to-day life of the people it looks to me like the decipherment is slanted. It may be true, but wholly one-sided. 

Q: Couldn't all those names of these specific functions also have a religious side to them? All these people, workers, scribes, tax collectors couldn't they have had a religious side to their function? 

A: Of course, we have such parallels in the city states of the Near East, where the king was also the god, the palace was also the temple and everybody had both a religious and secular identity. Nevertheless, if you look at the message of the writing in the Near East it is not all religious. I am not talking of the longer texts, even the seal texts, have things like so-and-so belonging to such and such city. For example, Harappa and Mohenjo-daro were very large cities. They must have had names. The people from there possibly used their city names. The secular, mundane day-to-day-life of the Indus valley is not sufficiently represented. In fact, those who have excavated those cities are struck by the absence of outward religious symbols. No temple has been excavated, no large deities have been excavated. I would be the last person to deny that Harappans had a religion and that it was very important and reflected in their seals, but that it would be the only message - that, I think, is unlikely. 

Q: Lets talk about the bangles decipherment by Parpola. 

A: Parpola has pointed out that the bangles are inscribed, and among the signs the sign of the interlocking circle or ovals are very common and they occur with greater frequency on these bangles. So I am fairly convinced that perhaps the interlocking circles do pictorially represent a pair of bangles. It is very likely. Now very large quantities of stoneware bangles have been discovered from Mohenjo-daro by [Michael Jansen's] German team. But when you try to give a phonetic value for it, it becomes very difficult. Parpola has chosen a word which means twisted wire bangle, or twisted wire amulet or a twisted wire earring or nose ring, where the operative word is twisting, the root there is murugu, which means in old Dravidian "to twist." But the stoneware, the polished vitrified stoneware bangles have no twists on them, so that is very unlikely. There are other words for bangles but he doesn't choose them because they are not homophonous with the word for Murukan that he is looking for. I personally believe that if the Indus Valley people were Dravidians, one of their gods was called Murukan - it is very likely, but he is hiding in still some other sign.

13. The Unicorn

Q: What do you think about the unicorn? 

A: We still do not know exactly what animal it is. There is no animal with a single horn like it of course. It still is very likely that it is only a pictorial representation of an animal with two horns, where the other horn is behind the one horn we see from one side. The animal looks more like an antelope than like a bull, this has been commented on. The first problem is in identifying the animal, about which there are several theories. 

Secondly, since it is invariably present with this ritualistic device which I have identified as a filtering device, it shows that the unicorn also had a religious connotation in the Indus Valley, like the Golden Calf of the Near East. Whether it was itself worshipped is a matter of doubt. It could be associated with God, or some fertility cult. Note that it is always the bull, not the cow. In this sense the Hindu veneration of the cow seems to be different from the Harappan veneration of the bull. I have never seen a cow being represented on any of the seals, except when it is a bull mating a cow, but never a cow by itself. Therefore it could be some kind of a fertility sign, or a sign of one of their Gods. 

Or, most likely, it could be the totem sign of one of the largest or most powerful ruling elites of the Harappan polity. With very few exceptions the filter device which you find before the unicorn is not placed before other animals. There are some exceptions, but they are rare. For the moment, the unicorn seems to be a very tantalizing symbol about which anything firm is yet to be found out. 

Q: Kenoyer and others don't believe that it represents a two-horned animal because the craft level was high enough that they could easily have shown two horns if they wanted to. What about the thought that it was some sort of composite symbol, an integrative clan symbol? 

A: Yes, I think S.R. Rao has also said it is a composite of a camel, a horse, and so on, although the camel and horse are never found in the Indus Valley. You cannot have a composite without the constituent animals being found there. But it could very well be a representation of what was already a mythological animal in the Indus polity, there being no real representation. We have other such animals, three-headed and two-headed animals. These were only mythological animals and the unicorn could very well have been one of them. 

Q: If it was a symbol of the ruling class, why did it vanish so entirely from later Indian mythology? 

A: It is difficult to say. One theory could be that it was closely associated with the rulership of Harappa and became the first casualty; when the ruling elite was destroyed the myth of unicorn went along with it. This is one possibility. The other is, taking the soma parallel further, who is most associated with soma in the RgVeda? Indra. What is most associated with the filter symbol in the unicorn seal, the unicorn. Could the unicorn be the prototype of Indra? One clue is the most common simile describing the soma drink in the RgVeda is the bull. It is very difficult to understand why a drink would remind you of a bull. A drink is a liquid, a bull is a very strong animal. Could this again be a reflection of the bull-filter device combination on the Indus seals? 

Another possibility is that Indra himself displaced Varuna. The original soma drinker was Varuna, Indra is a late-comer or Varuna was a pre-Aryan god. Could the Indrasoma combination of the RgVeda be a combination of proto-Varuna, proto-soma combination in the Indus Valley. These are speculations. 

14. The Indus and Dravidian Cultural Relationship

Q: How do you conceive of the relationship between the Indus culture that existed five thousand years ago and contemporary Dravidian culture here in South India? Prof. Dani, for example, says that doesn't believe that the Indus language was Dravidian because there is just not enough cultural continuity between what is today in South India and what was then in the Indus Valley. 

A: I think any direct relationship between the Indus Valley and the deep Dravidian south is unlikely because of the vast gap in space and time. Something like 2,000 years and 2,000 miles. But linguistically, if the Indus script is deciphered, we may hopefully find that the proto-Dravidian roots of the Harappan language and South Indian Dravidian languages are similar. This is a hypothesis. 

If you ask what similarity is likely to emerge, the first and most important similarity is linguistic. Culturally, there is a problem. The modern speakers of Dravidian languages are the result of millennia long intermixture of races. There are no Aryans in India, nor are there any Dravidians. Those who talk about Dravidians in the political sense, I do not agree with them at all. There are no Dravidian people or Aryan people - just like both Pakistanis and Indians are racially very similar. We are both the product of a very long period of intermarriage, there have been migrations. You cannot now racially segregate any element of the Indian population. Thus there is no sense in saying that the people in Tamil Nadu are the inheritors of the Indus Valley culture. You could very well say that people living in Harappa or Mohenjo-daro today are even more likely to be the inheritors of that civilization. 

In fact, I plow a somewhat lonely furrow in this. I often say that if the key to the Indus script linguistically is Dravidian, then culturally the key to the Indus script is Vedic. What I mean is that the cultural traits of the Indus Valley civilization are likely to have been absorbed by the successor Indo-Aryan civilization in Punjab and Sindh, and that the civilization in the far south would have changed out of recognition. In any case, the present South Indian civilization is already the product of both Indo-Aryan and Dravidian cultures, and the language itself is completely mixed up with both elements. Tamil alone retains most of the earlier Dravidian linguistic structure. Malayalam, Telugu and Kannada have become Indo-Aryanized much more, and culturally, the Hindu religion is a complete combination of all these elements. Therefore while it is legitimate to look for survivals, those survivals are as likely to be found in the RgVeda as in Purananuru, a Tamil work, as likely to be found in Punjab and Sindh as in India and Sri Lanka. So we have to separate our approach of a linguistic connection where it is permissible to construct proto-languages and try to decipher a language, but if you are looking at the survival of cultural and social traits of Harappan civilization they are likely to be all over the subcontinent, overlaid with centuries of transformation in culture and of language. Some of the myths may survive but may become unrecognizable. It is not a very easy or straightforward relationship that you can trace, it is a tangle. 

Q: What about the man and bull festival we were discussing . . . 

A: One of the cultural traits in the Indus Valley is that they had the bull fight. Some famous sealing show a man running towards a bull, catching hold of its horns, doing a somersault over the back of the bull, and landing at the other end. Even today in the Dravidian south bull fighting and bull chasing are very common sports. Yesterday, Tamil Nadu had this year's bull festivals where young men in the villages chase bulls and get hurt in the process. This is an assertion of their manhood and they can claim the hands of the fair maiden only after they are able to get hold of the horns of the bull and prove their heroism. This is very likely to be one of the traits which connect the Dravidian south with the Indus Valley. But such traditions are also known, for example, in Spain and in Portugal and the Iberian peninsula. There may well be a pre-historic connection between these very similar cults. 

15. Tamil Paleography

Q: The last thing I'd like to ask you is about your work in early Tamil paleography, and how this is connected if at all to the Indus script work. 

A: The two are very different fields. The Indus script is a logographic script, which means that each sign stands for a whole word or a whole syllable. The Tamil script, which is an offshoot of the Brahmi script, is a quasi-alphabetical script, where each symbol stands for a vowel or a consonant or a consonant combined with vowels. The principles of studying these two languages is completely different. Plus the Indus script is undeciphered. In the case of the early Tamil script we have the example of Brahmi, which is almost identical except for change in some syllables. 

Nevertheless, it happens that there was some diffculty in understanding the cave inscriptions in Tamil Nadu written in the Tamil variant of the Brahmi script and I took up that challenge and I have been working on that for more than three decades and I am now completing my book on that. 

Q: What had you added to that field? 

A: Since the cave inscriptions represent the earliest known stratum of a Dravidian language, their phonology and their grammar are most important. I have constructed an inscriptional grammar and compared that with the oldest known grammar in Tamil and oldest Tamil literature, the Sangam poetry. The oldest stratum of Tamil language known from literature is not very different from the cave inscriptions, showing that they were not very far removed in time. At the same time, the cave inscriptions represent the very beginning of literacy, so the main contribution coming out of this is in firming up our ideas of how Tamil was at the very beginning of its literate period. Of course we have a number of other spin-off benefits. These caves were all created for Jaina monks, so we come to know about the early history of Jainism in the Tamil country. Plus the caves were donated by traders, so we know about a lot of trading, in gold, pearls, sugar cane, gems, salt and so on. We find the Kings, the Pandiyas or the Cheras who are mentioned in the oldest Sangam literature appearing in these inscriptions, giving for the first time historical veracity to what was so long known only from fables or ballads. 

Paleographically, we now know how the Tamil script originated. There were many theories, but now we know it came from the Brahmi script and slowly became rounded because in the south, people wrote on palm leaves with an iron stylus and the letters got rounded. Each stage of this transformation can now be documented. The inscriptions I am studying cover from the time of Asoka, roughly the 3rd century B.C. to about the 5th century A.D. Thereafter from the 6th or 7th century A.D. we have hundreds of stone temples in Tamil Nadu with thousands of inscriptions, so there is no mystery afterwards. 

Much of what I have said now I have already published, but now I will bring out a definitive book. In fact, it is this work I am now engaged in and the Indus script, for me, is somewhat on the back burner. 

16. The Future

Q: When you do get back to the Indus script, what are you hoping to do? 

A: I have some unpublished, rather unfinished papers. One of the things I want to get back to is whether I can fix the phonetic values of some signs. I have some leads, particularly concerning the phonetic values of the terminal signs, which are very important. If at least a few phonetic values could be established, they will have a cascading effect. Many of the symbols could yield their results. But this is still a dream.

I also still want to work on the filter theory, whether by combining Dravidian theory, Indo-Aryan or by looking at the finds in the Indus Valley itself.

But I have no illusions that I will decipher the Indus script, nor do I have any regret. I think the ultimate decipherment of any language is built on the unsuccessful attempts of so many earlier scholars, so I am content to remain one of them. I am very hopeful that further discoveries by teams like Kenoyer's and others, and the active work of Parpola and his team will all bear fruit, hopefully in my lifetime.