In Conversation with Dr Rajat Sanyal: Indian Palaeography

In Conversation with Dr Rajat Sanyal: Indian Palaeography

in Interview
Published on: 20 February 2020

Dev Kumar Jhanjh

Dev Kumar Jhanjh is a Doctoral Candidate at the Centre for Historical Studies, Jawaharlal Nehru University, New Delhi. His area of specialisation is Epigraphy and Numismatics. He is presently looking at the Political Processes in the Central Himalayan Region (From pre-State to State) (c. first century BCE-twelfth century CE).

Dev Kumar Jhanjh in conversation with Dr Rajat Sanyal on Indian Palaeography

Dev Kumar Jhanjh (DKJ): Sir, palaeography as we know focuses on the letter form and the evolution of the same and is intimately associated with deciphering and studying the inscriptions. Can you elucidate?

Rajat Sanyal (RS): Actually ‘palaeos’ means old and ‘graph’ is writing, letter writing, the study of old writing. Rather old writing is palaeograph and study of old writings is palaeography. As is epigraphy; ‘epis’ means surface, ‘graph’ is writing, so study of writing on surface is epigraphy. And these two terms in the context of any geographical or chronological frame are immensely interrelated in the sense that if you want to decipher the inscription, you will have to know the script and after you know the script, you have to know the language which is part of the inscription, which is the actual basis for the inscription without which you cannot decipher. 

So palaeography and epigraphy are always closely related. Palaeography in the Indian context has a long history. Now we have discovered from the Harappa civilisation that Indians started writing in the proto-historic period, from the mature Harappan phase. Earlier it was believed that writing evolved in India only with the appearance of Ashokan polity. Now we know that it starts from the mature Harappan phase. But as far as historical writing in India is concerned, it, of course, starts from the third century BCE, from the Ashokan inscriptions as we all know. 

And if you see, from here onwards, the Ashokan inscriptions and the post-Ashokan inscriptions till the medieval inscriptions, these are written in a wide variety of scripts as we will now discuss, and a wide variety of languages as well. So apart from the script, you have to know the compositional element of the inscription in order to decipher those. And only compositional element, understanding of the compositional element will not help unless you know the script. That is where the inter-relationship of palaeography and epigraphy happens. 

To illustrate it, exemplify it, with Ashokan inscriptions, these are written in different varieties of Prakrit as we all know. It was initially thought that it, the language of the Ashoka inscriptions, had four regional varieties. Richard Solomon had questioned that this is not four actually, there are three regional varieties of Prakrit in which Ashokan inscriptions were written, apart from these extra scripts and languages and also Kharosthi. 

If you talk of Indian language, these are three varieties of Prakrit, north-western, western and eastern variety of Prakrit. And based on these, it is generally believed that the Ashokan inscriptions were written in a Mongolian script . . . it is still believed but if we consider the wide variety of the language used in Ashokan inscriptions, we will see that there might be minor variations in the script as I will illustrate. 

So in understanding inscriptions in any context, any geographical or chronological context, understanding of the script is quintessential.

DKJ: Sir, there are many inscriptions which are undated and we often hear that palaeography helps us at least roughly to date the inscriptions. Is it so? And if yes, then how do we do this?

RS: Actually most of the inscriptions, if we look at the entire South Asian corpus of inscriptions, most of them are undated inscriptions. You do not have precise dates in calendrical years on inscriptions themselves. Either these are completely undated or they are dated in terms which we are not certain of. Say for example, an inscription is dated in the ruling year of a king as it happens mostly in eastern India or in other parts of the country. You have to know two things. One is the historical context within which the inscription was issued and the second one is the script in which the inscription is written. Only palaeography is not always helpful in understanding the date or the range of dates of the inscriptions. But if you have an idea of the historical context within which the concerned inscriptions are written, you can, of course, date them with the help of palaeography.

Why is palaeography not individually helpful in this? Because palaeography, as we know, even if you take the case of modern Indic scripts, scripts do not change overnight. It is a tendency not to move the pen from the surface that finally makes the script change. We have a tendency of writing continuously which we call cursive writing. This is the reason that script changes precisely, primarily. There are of course other reasons we will see but this is the primary reason that we come to know how the scripts change and this change takes place only over a different span of time. I would suggest a couple of centuries. So within a couple of centuries, the influences of the changes in script are difficult to understand. This can be only helpful when you have a general historical context of the inscription. If we know that, then this palaeography, of course, acts as a very useful supplementary tool in deciphering of inscriptions.

DKJ: Sir, you were talking about the scripts. So what is the methodology of deciphering these scripts?

RS: The methodology demands explication. For us, the people, the scholars of the modern world who already have in their disposition a large number of inscriptions deciphered for them. The methodology is basically to go from known to unknown. You have to study the scripts that are known and then you have to proceed towards the unknown. But if you consider the pioneers who deciphered the scripts, how did they do it? Was it the same? It is usually believed, some people believe, that it was not so. It was actually so. They also went from known to unknown.

DKJ: How?

RS: Say for example, the famous case of the decipherment of Brahmi and Kharosthi by James Prinsep.

DKJ: And these two are the earliest scripts.

RS: Yes, these two are the earliest scripts. And James Prinsep, as we all know, as the Assistant Assay-Master of the Calcutta Mint, he had deciphered this. He worked continuously for a decade and ultimately deciphered the script in 1837. 

But how did Prinsep do it? Did he decipher the Ashokan script from the Ashokan script itself? No. If you see the history of the development of Indian palaeographic studies, which is not generally studied separately from Indian epigraphic studies, Charles Wilkins was the first person who published the inscriptions in the late eighteenth century in the Asiatic Society journal and then a number of inscriptions were published in the Asiatic Society journal until James Prinsep started his research. But these inscriptions were mostly later inscriptions as they happen to surface from various parts of the country. But Prinsep, what he did, he started a new methodology. Since he was the Assay-Master of the Mint, he had at his disposal a number of early coins with bi-scriptual legends, coins with legends on both sides. And one of the sides had legends in the Greek alphabet which was known to Prinsep. The other side which contained either Kharosthi or Brahmi, these were not known to Prinsep. But he started comparing the Greek part with the corresponding Brahmi or Kharosthi part. And based on the decipherment of the Greek letters, he tried to read the corresponding Brahmi or Kharosthi letters, again the prerequisite is knowledge of language. Because Prinsep knew the language in which these inscriptions were written. He knew that these are written in either Prakrit or Sanskrit, otherwise, if you do not know Kshatrapas means king and Maharaja in Sanskrit means king, how do you compare the letterforms of Kshatrapas with Maharaja. So the basic methodology is to go from known to unknown. Then decipher in stages from late to the early as he, Prinsep, did. 

This is primarily the methodology. But it also has its ramifications. Not only looking at the script itself and corresponding known letters and you have to have knowledge of the language, but you also have to know the actual contexts in which the inscriptions are written. Palaeography is meaningless unless you know the context, historical context, of the inscriptions.

DKJ: Sir, can you show us some examples of how you decipher the scripts?

RS: Actually the decipherment has already been done by all pioneering predecessors, of course in fact without examples we cannot explain how the palaeograph is deciphered. I would rather like to illustrate this with examples. 

If you look at the Indian alphabet, structure of the Indian alphabet, actually keeping aside the debates on whether the Indian alphabet really represents a true alphabet that is a debatable issue. If you look at the general structure of the Indian alphabet, it has four different elements. We have the full vowels, and then we have the medium vowels. Then in consonants, we have the full consonants, and then these consonants are joined to form ligatures and conjuncts. This is the general structure of the Indian alphabet. 

Among the consonants, we have five different kinds of sounds, the Guttural sounds (ka, kha, ga, gha, ṅa, ha). Then the Palatal (ca, cha, ja, jha, ña, ya, śa), Retroflex sounds (ṭa, ṭha, ḍa, ḍha, ṇa, ra, ṣa), Dental sounds (ta, tha, da, dha, na, la, sa) and the Labial sounds (pa, pha, ba, bha, ma, va). All these groups of vargas of sounds are again divided into voiced and unvoiced versions, nasal sounds, semi-vowels and sibilants. This is the general structure of the consonant of the Indian script.

Now, if you take a look at the earliest Indian inscriptions, rather the Indian scripts from which all the modern, South Asian and Southeast Asian scripts have evolved, this is its appearance, in terms of the structure of the alphabet that we had in inscriptions. The earliest scripts appearing on the inscriptions of Ashoka, in terms of their structure, the guttural sounds, the Palatal sounds, the Retroflex, Dental these are the basic consonants of the Ashokan script. I am not getting into the details of the vowels because we will be seeing that in the next stage. So, from this onwards, unless we understand the actual systemic of this particular script, the earliest of the scripts of India, it will not be possible for us to understand what you call the evolution of scripts. This is of primary importance. 

I am going to the next one. This is the actual appearance of the script. The famous Rummendei Pillar inscription of Ashoka and it is a very good specimen for practising early Brahmi, you can read the whole script without any hindrance because it is written by the royal scribe and this structure, this systemic structure—I will explain what I mean by systemic structure later—this systemic structure prevails almost throughout the Ashokan dominion without any major deviance from the structure that we see here. I am showing this because this is the best example of the whole set of Ashokan inscriptions. We also have many other examples but this is the best one. Here it records Ashoka’s visit to Lumbini where the Buddha was born. The text of the inscriptions has recently been debated by some scholars, but the primary purpose of this inscription was to record Ashoka’s visit to the place of the Buddha’s birth. 

From the second century BCE onwards, after the decline of Maurya rule, we find gradual change in the early Indian Brahmi script, the historical script of India and these changes occurred in different terms of which the first, as Ahmed Hasan Dani has rightly suggested, was changes in the contemporary Greek palaeography that largely influenced the writing of the Indian subcontinent from the second century or first century BCE onwards. The first one was the development of palaeography in Greek script. Then, of course, the decline of the Maurya polity which actually was responsible for the monumental nationalistic Brahmi script that we see in Ashokan inscriptions. After the decline of the Maurya rule, this structure was abolished with not only polity but also other aspects of the culture of which script or writing is an integral part. 

Secondly, we have as I just said, the Ashokan inscriptions were written in Prakrit language. From the first century BCE onwards we have a gradual appearance of Sanskrit as an epigraphic language. By the middle of the second century CE we see that Prakrit is being completely replaced, almost completely replaced by Sanskrit. And from the third century CE onwards we have even the development of a special language, for writing inscriptions, which is usually called Epigraphical Hybrid Sanskrit or EHS. 

The fourth factor was the growth of provincial political cultures as I have already said. After the decline of the Maurya rule, we have several regional cultures. As it happens, after the fall of central powers, there were several local regional cultures under which script was appearing on inscriptions, the Brahmi script. And this also influenced rather the change of scripts from the first century BCE and, finally, the variation in the content and purpose of the inscription. But, this (pic. of Indo Greek coins in the slide) was to me the most primary factor in making the script, the Brahmi script, evolve. 

These are the Indo Greek coins on which you have Brahmi and Kharosthi scripts. Here I have plotted elements from Brahmi script. If you see the scripts, each of the letters (staring from Omicron to Epsilon), these are having from now onwards, from the first century onwards if you see all the scripts, they are having two tendencies. One is this circular form which has been converted to a square one and a second one you see if you compare this (Σ) with this (Ω); the ends of the letters, the terminal points of the letters, are having blots on it. This is because the scripts were written in their areas, in those geographical places, areas where Greek was the lingua franca where they used to write with reed pens, a kind of a pointed pen which when writing on a surface would result in blotting on papers at the terminal points of the letters. Whenever we write an akshara, the initial and the terminal points we have blots of inks from that reed pen. That reed pen technology was being converted on metal or stone or whatever be the surface of the inscriptions. That is why you see these blots. These blots did not happen automatically. These were actually copied from the inkblots that were characteristic of writing on surfaces like papers. 

You see this resulting in the change of the systemic of the script. In what way I am studying the systemic? If you see this inscription, the Rummendei inscription that we have just seen, if you take a snap, if you take a photograph of the inscription and we juxtapose that photograph with this one which is dated to the second century BCE, the famous Besnagar Pillar inscription, we will see that this (sa of Rummendei) has achieved this form (sa of Besnagar). It is a clear departure from this one, where you have the breadth of the script extending and the length reducing while here you have a rectangular outline, here (Besnagar) you have a complete square outline. And here, remember the change from circular to square in the Greek alphabet. The same is true for the alphabet ha. 

And now we have the formation of new ligatures which were absent conditioned by the necessities of the language used in Ashoka inscriptions. Now we have the appearance of new phonemes. So I have pointed out three major changes. One is the square form of the akshara, the angularity of the script and appearance of new phonemes. 

This is the Hathigumpgha inscription of first century BCE. The photograph is not very clear but here we see the first evidence of a complete change from Ashokan inscriptions. We have also some other minor examples but of the large inscriptions, Hathigumpgha is the foremost where we have the three changes that I have pointed out in the last slide. These are manifested for the first time completely on these inscriptions. 

Here (a donative inscription from Mathura) you have the evidence of that reed pen style. All the letters, if you see the letters, all the terminal points of the aksharas are having blots, having dots which are actually not original dots (in Indo-Greek coins). These have been made to be so because they were copying this, they were just imitating the reed pen style with which they were accustomed in their own region. And now you have this not only on coins but also on inscriptions (a donative inscription from Mathura). This is one inscription in Mathura where you have the same blot at each of the points and these blots are also responsible for the formation of what we now understand as matra. The head marks are also now developed. If you look at the vertical elements are having head marks always at the top that also resulted in the formation of the head mark which was gradually being converted into serif or matra. What were the results? The results were, the foremost result was the development of a number of regional styles. Now we have, unlike the Ashokan inscriptions, quite a large number of regional styles of writing characterising inscriptions all over the subcontinent. 

Then you have new symbols. Why? Because the language has changed, the introduction of Sanskrit has now necessitated the introduction of new phonemes as we saw for the first time with the Besnagar inscription itself as early as the second century BCE, from the first century BCE-CE onwards, we have a large number of such new symbols appearing on the inscriptions. Therefore this was the second major result. 

The third one was the introduction of head marks as I showed you. In Ashokan inscriptions, you had no headmark. The letters were clearly written but now onwards you have different kinds of head marks which were actually the result of imitation of the reed pen style, and then a large number of medial vowels. Why was it? Because the writing styles were diversifying under different regional schools of writing. A.H. Dani talks of these regional schools of Middle Brahmi from the first century BCE-CE onwards. 

Ashokan Brahmi which was earlier called Maurya Brahmi is now called Early Brahmi because it was not Ashoka himself who wrote inscriptions in Brahmi alone; other contemporary rulers, his descendants, they also used the same script. So it is better to call it Early Brahmi as a more generic terminology and he (Dani) preferred the terminology of early, middle and late, what was earlier thought as Maurya, Kushana, and Gupta Brahmi. So it is more or less generally accepted now that these three evolutionary stages be called Early Brahmi, Middle Brahmi and Late Brahmi. 

In Middle Brahmi, Dani detects these six regional schools of writing—one in Mathura, one in the north-western part of the country, one in central India and one in north-western Deccan, eastern India and southern India—six different regional schools which were characterising, which were showing, evidence on the changes that we just talked about. And this resulted in the formation of a script which is completely different from what you saw in the Ashokan inscription. If you take a snap of this and compare that snap with the Ashokan inscription, it is a complete change, the two snaps do not match. This is simply what I call a systemic change and not individual palaeographic changes. 

This curve is taken from this rectangular outline to that square outline and this curve is having this angular element instead of that curvilinear element. These are not changes. If you take a look at the script as a whole, if you see the structure of the consonants, if you concentrate on the structure of the consonants, this set is a complete departure from what we saw in the inscriptions. And this is the general structure, this is a normalised general structure of the middle Brahmi in Northern India. We have to remember that this is the example only of northern India. We have now at least as I just showed, six regional variations where you have different manifestations of this systemic change in different parts because of the development of the regional scripts.

From this time onwards, from the first century BCE till the fourth century CE, this is the second stage of evolution of Brahmi from Middle to Late. As we saw from the third to the first, the development from early to the middle, here is the next one. And here the factors that were responsible for the evolution of script are different from what you saw earlier. Here the first factor is the development of the cursive style. From now onwards because of extreme regionalisation of scripts, I will show you the nature of regionalisation, this is from the fourth century CE. 

In the fourth century, you have these changes, these factors finally affecting the script. These factors were affecting the script between the first century CE and the fourth century CE when we have the final appearance of the late Brahmi. Here, you have the first factor as the cursive style of writing that I first talked about when you asked me about the methodology of understanding scripts, then I talked about cursive style. Because of extreme regionalisation and also the formation of new symbols, people are now writing in cursive styles. They are not taking the pen out from the surface. So new forms, shorter forms of the letters which are completely different from the normalised versions, they are appearing on inscriptions. This was one. 

Then as I said, regionalisation, more symbols because of development of Epigraphical Hybrid Sanskrit now. And now the composers of the inscriptions have a curious fondness for flourishes. They are not writing ka simply as across or even as a cross with a headmark. They are not satisfied with that. Now they are making curvatures. Now they are making flourishes, designs on the simple appearance of the letter and then diverse material. Finally, this is the appearance of the script. 

Here you can see, as Dani shows, now you have different types of head marks characterising scripts in different parts of the country. You also have this fondness for flourishes like the medial vowels are not being added as simple angular lines. They are being devised as curvilinear designs and also you have a tendency of cursive writing as you can clearly see in this form of ha or as in this form of cha. These are head marks, flourishes, cursive style and then the medial vowels if you remember the points. These are the formation of the medial vowels unlike the Ashokan Brahmi or the Early Brahmi where you had a particular formula for adding medial vowels. Here you have different types of medial vowels being added to the letters. This was also a result finally of the regionalism in fact. 

Finally, you have this as evolutionary, from Early till late Middle Brahmi formation. And these evolutionary sequences are not visible only in terms of individual letters but also in ligatures where you have this sa finally appearing in this form (another variety of sa) via this.

And this is the regional nature of regionalism that we see in Late Brahmi. You have many styles. In Dani’s terminology, you have a script family from the Indus Zone, one from Mathura, one from Sanchi, one from Kaushambi, one eastern India, one inner Madhya Pradesh. We also have several other regional variations and there again systemic remains almost the same but the palaeography is completely changed. So now we have two layers of changes. One is the systemic and another is the regionalisation that is palaeographic.

So finally, in Late Brahmi, you have these as regional varieties of scripts, as many as nine sub-regional varieties as has been explained by Dani, starting from Indus Zone, Mathura, Sanchi, Kaushambi, eastern India, inner Madhya Pradesh, Eastern Malwa, north-eastern Rajasthan, south-eastern Deccan. You have these nine regional varieties of Late Brahmi which are appearing in fourth century CE. And this is the structure of the script. Here, I have shown the differential use of the medial vowels and here I have shown the structure of the consonants. Now again, if you take a snap of this and juxtapose this against the one that we saw in Middle Brahmi, these are quite different. They are completely separated from each other. And this is what I say, systemic. The whole structure of the script has changed. 

Remember that this is only the middle Ganga valley variety of the script. As you just saw, just seeing an enormous number of regional variations of which Middle Ganga Valley is only one. So you have normally two layers of evolutionary change, one is the systemic of the script and the structure of the script. The second one is regional palaeographic varieties or variations. And these are the ligatures that you have where all the factors, the cursive writing style, the formation or rather the appearance of flourish, the formation of new symbols, the formation of new medial vowels, and formation of new ligatures. All these can be seen in these different forms of the ligatures which are taken from the fourth and fifth century CE Gupta inscriptions of the middle Ganga Valley.

This is an example of the actual inscriptions from the middle Ganga valley, the famous Kahaum inscription from the Gupta period. See the head mark, the formation or the flourish, the differential use of the medial vowel and again the flourish of the medial vowels and also the cursive style of writing. Clear letter written as a continuous form starting from the top to the bottom and here the headmark has been intentionally created. This is not an original headmark. And this is one of the eastern Indian varieties of the same script that you have in the sixth century CE. Here you can see almost everything remained the same. If you see this and get back to the earlier one, you see almost everything is more or less the same except the headmark which was there; solid triangle here meets a solid part which is characterising the script in eastern India. So, you have regional palaeographic differences in the script. 

Now here you see finally the systemic change. This is Early Brahmi, this is Middle Brahmi and this is Late Brahmi. You see this one, and then place it against this, it is completely different. You see this one and place it against this one, it is completely different. So that is how from the third century BCE till the fourth century CE you have a linear evolution of Brahmi script in northern India. You have to remember that from Middle Brahmi onwards, this unilinear pattern is completely lost as you have several regional varieties of scripts but even within the regional varieties you can have unilinear patterns of ligatures.

This is the general methodology that we follow in understanding the evolution of scripts from the earliest of the Brahmi until its latest version in the fourth century CE. And this is how the script is making a final departure from the most mature version of Brahmi. (Haraha inscription of Isanavarman, 554 CE in the slide.) And the regional scripts, those were regional variations of the same script, the earlier ones, from Middle Brahmi till Late Brahmi but here it not only remains confined to the limit of regional variation of a particular script, it becomes a regional script itself. 

Here, as you see in the Haraha inscriptions of Isanavarman in sixth century CE, now you have a completely different script characterising the inscription and this marks the advent of Siddhamatrika from Brahmi, starting from the ligatures to the formation of the medial vowels and the formation of the letters themselves. Now we have a complete change in the whole structure of the script which shows a departure from Brahmi to Siddhamatrika. From now onwards we have a large number of regional scripts which were first noticed precisely by Abu Rehan Al-biruni in the eleventh century who talked of, as we all know, eleven different regional scripts of which the eastern Indian script Siddhamatrika, Vaiksuki and Gaudi were the foremost.

The video interview can be found here: