 when we saw the money supply from the summer and embarrassed you with the little introduction. I was first exposed to the great insights that Robert can bring to the subject by a paper they gave and you are not sure you are publishing on the coins now. I was not sure he was writing on the money supply that you know and you were not sure he would publish it in the final date of the takie fraam. ..on rai ein bydd, a wnaethon chi wedi bod yn gwneud yn ennill yn ysgrifftyn.. ..ynddoedd y maergylcheddau i'r hyn. Rwy'n meddwl y clywed â'r clywed, mae'n meddwl ffordd i'r meddwl... ..rydych chi'n meddwl i'r meddwl i'r meddwl i'r meddwl.. ..ynddoedd y meddwl i'r meddwl i'r meddwl i'r meddwl i'r meddwl i'r meddwl i'r meddwl i'r meddwl i'r meddwl... Ac yn y cyfnod o'r projekty hynny, yn ychydig i'r materio ac yn ei ddau'r cyfnodau cyfnodd, yn mynd i'r ddataethau, yn y penderfyn nhw i'r cyfnoddau ar y cyfnoddau a'r ymddangos am y teimlo'r anoddau oherwydd mae'r cyfnoddau ar ymdau cyfnoddau ar y gwaith. Nawr, byddwn i'r brlyw i'r gwiredd ychydig iawn yn gwybod, ond oedd wedi dod oed yn gyfnod i'r gweithio cyfferthai, ond rydyn nhw wedi gweld bod gan bosbtwch â'r gw再bloed. Rydyn ni wedi cymryd g weerbodd i ni. Rydyn ni wedi gweld yn fwy ddim yn gwybod i'r pas. Wrth bod yn ymweliad ychydig i'ch pwysig ac mae'r pwynt yw yn sefydledd yn ôl rydyn nhw'n gwybod i'r ffordd yma. Felly tynnu amser, rydyn ni wedi cael ei wneud i'r cyllideb yn fwrdd arlaug, logodau gwahanol, yw'r cullid yn rhaid. Rwy'n deithas nhw, rydyn ni'n tiriau a'u gwisio allanod. Rhaid o wneud yn fwy, yn rhaid rhai. Rhaid o bobl technicol ar amser yw'r cyflosion. Mae'n rhaid o wneud i wneud ar arnau o rhaid. Ie, so Michael is very kindly going to just move the slide forward when I do the same here. As the blurb on the website suggested, I'm going to be talking about a quantification problem, essentially the problem of how we count coins and what those counts mean in a wider historical context. And I'm going to be doing it in a relatively narrow focus and sort of exploring the methodological issues. And I'll say a bit more about what I'm going to talk about today. But I wanted to say yes, both Daniel and Gethyn are going to intervene at points in the talk to discuss particular aspects of interest. I've also got several other colleagues who aren't from the project who have agreed to be here today. Joe Cribb, who's the former keeper at Coins and Metals. Shalindra Bandari, who's the present, you're called keepers at the Ashmolean, aren't you? So it's currently senior assistant keeper at the Ashmolean. So Joe and Shalindra are the two foremost experts on Oriental Coins in the UK. And they will simply intervene when they think it's appropriate. Shalindra, in particular, has written quite a bit about the subject I'm going to talk about as well. And Golraheem Khan from Peshawar University is visiting me at the moment and works extensively on coinage in Pakistan, Afghan border area, particularly the material at Taksila, which he's written some of the standard treatments on. So next slide. This book is essentially the stepping off point for all of this. This is Early Medieval Indian Society by R. S. Sharma. R. S. Sharma is well known for his advocacy of a theory of feudalism in the mid to late 1st millennium. One of the things that he believed happened with his advent of feudalism is that the economy was demonetised, that increasingly market exchanges were replaced by feudal obligations. You didn't work the land because you were given wages. You worked the land because you were required to by your overlord and you were tied to it. And this is from an introduction to a chapter. He calls it Porcity of Metallic Coinage and he gives this little account. We are so obsessed with problems of political history that the obvious relevance of coins to economic activities does not bother us much. It needs to be stressed that coins so dramatically altered the nature of exchange and economic transactions that once in use they contributed immensely to trade, influenced the mode of taxation and established impersonal relations between consumers and sellers and between the state and its employees. Far-flung areas with different levels of economic development were bound together in a network of exchange relations leading social and cultural interactions often cemented by political domination. So Sharma is very interested in coins and one of the things he does in this chapter is to try and demonstrate that after the Gupta dynasty the number of coins in circulation goes down and therefore by extension the amount of money. Next slide. So a little bit just to make clear because there are limitations of expertise in what I can speak to here. What this talk is about how we can count and quantify numismatic data. This is the kind of exceptionally dull nerdy question that only numismatists really care about and even only some numismatists. But it's enormously important. Sharma is resting the history of the late 1st millennium on a count of the number of coins. So how we count coins matters. It's also about this formula here, mv equals pq, which if you've done any economics class you will know is the quantity theory of money. Fisher's formula for the quantity theory of money. It's q for Fisher. The formula varies, but it's essentially a claim by economists that there is a relationship between the amount of money in circulation and the amount of market transactions and the value of them. It's going to be about whether or not cross regional trends existed in South Asia in this period. And it's going to be asking the question whether or not any of this means anything. What this talk is absolutely not about. I will not be offering a definition of feudalism or discussing whether or not feudalism happened. That is a much larger question. What I'm talking about here has a role to play in informing that sort of question, but it's not a question I'm qualified to discuss. I won't be discussing Marxist theory. That's partly because I'm not qualified to discuss it, and mostly because it's immensely dull. I'm not going to be trying to answer the question if a profound social change occurred in India in the first millennium. What I will do, and I hope we can all be in agreement on this, I am going to assert that the evidence that we have available as historians at the end of the first millennium looks radically different to the evidence we deal with at the beginning of the first millennium. If you work primarily on the first century's AD and you switched working on the eighth ninth century, the source material you encounter feels and looks different. That, I take as an empirical given for the whole talk, I take that to be the starting point for Sharma's hypothesis, and that essentially what he's offering is an explanation for why that takes place, and his explanation is there is a profound social and political change, so the source has changed. I'm not going to be presenting any new theories about what happened. This is, at its core, going to be a talk about the methodology. The first thing that I'll do is actually do exactly what Sharma quite rightly criticised for, which is worry about politics. I have a whole bunch of maps here. These are from John Huntington's site where he does maps of South Asia. These are base maps. If people can pass them round and maybe share between one or two of you. I have some nice thick marker pens you can play with, not enough for everybody, but do pass them round. What I would like everybody to have a go at in teens, if you want to, is to draw a map on top of this base map, a political map of India in 550 AD. In 550 AD, which states were where? Which rulers ruled over what territory? I couldn't do this myself, but I think this is always a useful exercise. Have a quick go. We'll give everybody five minutes in which to have a go at this. Don't put your names on it. You do not have to accept responsibility for whatever borders you jot down. Partly for my own interest, I'm genuinely interested in what a group of people who are all specialists do with a map in terms of borders for 550. Go on, have a go. I think one of the reasons this is potentially interesting is because if you asked a Romanist to draw the borders of the Roman Empire in 150 AD, they'd be able to put a line straight on a map. They'd know which bits of territory were inside, which bits were outside. If you asked somebody who studies European history to draw a map of Europe in 1913, they'd be able to put the borders on. My experience of doing this task before is often in South Asian studies. We just don't think in these terms. You can do this as teens. You don't all have to do it on your own if you feel self-conscious. I think we'll restart there. I'll be interested to get the maps in. I see Sheldon has been very enthusiastic on his map. So some people have given this a go in the past, well, recent past. This is a map by Thomas Lasso, covering 550, which I think is completely wrong, but it's interesting. I think one of the differences between this by an amateur essentially doing things for Google Wikipedia stuff is the way that all those borders meet. Whereas looking around the room, it's quite noticeable in this room that most of the borders don't. Most of us don't think of these states as being rammed up against each other with modern style borders. This is a one from Wikipedia from 590. Again, I think that that purple is somewhat optimistic. The next one is from about 600 by Huntington himself using this base map. That's a bit later. The only one I could find for 550 was the first one by Lasso. That little exercise out of the way just to get everybody thinking about this period in the 6th century. The thing, of course, that maps in the 6th century is by about 500, the Gupta Empire has essentially collapsed as the hegemonic power of North India. Nobody historiographically is writing about a hegemonic power again until Harsha in the 7th century. For this 100 years in the literature, there's no hegemonic power. Now we'll come to another one of those statements. This is empirically true. The coin in the top left is not a gupta coin. The coin in the lower right is. I would say that it's an empirically true statement that everybody in the room has seen more examples of the type of coin shown in the bottom right than they have of the type of coin shown in the top left. I would say that's true even of the specialist numismatists. Do either of my specialist numismatists want to disagree with me on that point? No. They are, I mean, gupta coins are far more familiar. Far more familiar. The paucity argument, which is the central argument underlying Sharma's recently statement of his feudalism hypothesis works like this. I have seen fewer coins of this period type dynasty. Therefore there are fewer coins extant. Therefore fewer coins were made. Therefore fewer coins circulate at the time that that coinage was made. Therefore there was less money. That's the paucity argument in that show. The very first one of those steps is true. I have seen less of those coins. All of the subsequent steps do not follow. Sharma then uses it to make his main argument. He counts the number of coins that are in collections. This is the graph at the bottom behind me. This is going to take a little bit thinking round my apologies for it. The graphs go from oldest on the right-hand side to newest on the left-hand side. Which is slightly counter-intuitive. The first time I gave a presentation on this I drew the graphs that way round. I have no idea why I did it, but I then subsequently felt obligated to make sure all the other graphs correspond. All the graphs work from oldest on the right-hand side to newest on the left. At the bottom are Sharma's counts of items in public collections. You will see lots of punch mark coins, lots of Cushan coins, a lot less than everything after the Cushans, a lot of sultanate coins. This is his argument. There's a lot less money in the late 1st millennium. The graph on the top is from John Dale. He has Ikshavakesh. The graph on the top left is by John Dale. It is a counter of the number of hordes per year. This is per year of issue of the coins that are in the hordes in Uttar Pradesh. The pattern is slightly different, but the basic shape is still there. There are lots of punch marks. There's a lot less of everything in between. It goes back up under the sultanates. So, quantity theory. The quantity theory works like this. The assumption is that the money in circulation chases the things you want to buy. So, if there's more money in circulation, then the amount, the price of the things you want to buy goes up. This observation was made by Europeans in the 16th, 17th century, when they showed a dramatic rise in prices with the influx of silver from the New World. It can vary a little. So, if the coinage you've got, if the money you've got is moving around faster, that functionally increases the money supply without actually putting more money in place. So, if you take your money and put it in a bank account and it doesn't do anything, then it's not really adding to the money supply. So, this is known as velocity. It's the second term. So, the money supply for economists is how much money there is and how quickly it changes out. M for money, V for velocity. And the effect it has is dependent on how big the economy is, Q, and what the price of everything is. So, the assumption economists make is that basically people's behaviour stays essentially the same. Velocity doesn't change very much. Economies don't change in size very quickly. That's often modelled by taking the GDP of the country, which alters but doesn't alter very fast. So, changes in money, in the amount of money, altered prices. So, when you hear that a government is being reckless because they're printing money, this is the risk, inflation occurring. However, the ancient economy is not universally monetised. Only certain sections of it are. And so, what Sharma is doing, he's not doing this formally to talk about the quantity theory, but what he's doing is he's assuming that in the ancient world, if the amount of money changes, the price doesn't have to alter. Instead, you could simply change the size of the monetised economy. A section of the economy could stop being monetised if the amount of money goes down. A section could become monetised if the amount of money goes up. And this is what is essentially positive, that the money supply is proportional to the size of the monetised economy. And when the money supply falls, the size of the monetised economy falls, and it's replaced with non-monitised transactions. Land grants being a major part of the thesis. And then next slide. And there have been some critique of his position. Shailendra might wish to say something because this is his article on the screen. So Shailendra wrote of a lengthy article critiquing the basic underlying attempt to measure the money supply by counting the number of coins in collections. Do you want to say something? I'll tell you a little bit on how the method of counting coins, because first what Sharma did was, as Jo knows, he came to British Museum and he said he started counting trades. And there were lots and lots of Gupta coins, because Guptas have been collectible. Gupta coins have had good press. They look beautiful, they're gold, they have lots of nice imagery, there was this classical, slightly nonsensical paradigm about the golden age of Guptas. So it was people collected Gupta coins more adequately. And so as a result there are lots of Gupta coins in museum collections, and whatever came after Gupta coins was not really attributable. So it was not collected. Collectors like coins that they can attribute and put it in little slots. So as a result, most of the museum cabinets after Guptas were tentative, because they were not attributable coins. So what I tried to say in the article here, the first point was that lack of attributable coinage does not mean lack of coinage. So if you can't attribute a coin, it won't be collected because collectors are not interested. And museum collections therefore are not really any empirical basis of the actual currency that was in circulation. Museum collections are formed with lots of biases. There is money bias, there is people who like certain coins. When Joe was the keeper, Cushan collection increased, and the rest did not. So they haven't got any more coins, for example, for eons, because there is no driver for this kind of coin. There we are. In fact, those coins show you an example from Vietnam that has no examples of it. So that was the main thing. That's why the other sort of things which Robert will talk about, I hope, is the kind of inadequacy of numismatic knowledge about certain things. So you see on a coin the name of a Gupta emperor, and then you think that it is a Gupta coin, but in reality it is not. It is just a continuation of Gupta coins, which sort of went on beyond the Gupta coin. So even in case of Gupta coins, there are lots of silver coins, for example, from Gujarat, which now we know with other methodologies are actually posthumous, but they all got sort of moved to Gupta cabinets from the post Gupta cabinets. So that sort of created a lot of skewed picture in terms of numbers represented within cabinets. So that is basically what my point to say in that article was. And also then I presented certain ideas of if you actually want to quantify money in circulation, what are the methods available, can be used. So that's where I stopped. When Benares Sharma came to the coin room, I went round the collection with him, and all the time he was counting the coins, I was saying exactly the same things to him, so he had no excuse because he'd been told. There's a really good example. In the collection, at that time, we had one coin of Maltam, which wasn't identified. This is the early Islamic emirate of Maltam. Since then, I've seen about 5,000 examples through port. But for that period, as far as we knew at that time, there was no coin engine in Maltam. Now we know it was an extensive coinage. It was just collecting practice in the southern Punjab and seeing people who were collecting coins. In the northern Punjab, that's the share of people collecting coins. That was the reason. So what I did or started doing earlier this year in relation to other things that were going on in the projects, I began trying to narrow the focus down on the general presentation Shalendro made, and look at just a very narrow period, but across the whole region. The problem is that mumismas tend to work vertically through time. So we tend to pick a point or a series and follow it without much regard for what's happening around the edges. The kind of cross-regional pattern Shalendro was identifying was only really visible if you cut across mumismatic series. So I decided to look at the period from 500 to 600. This is the period immediately after the government collapsed. It's the very beginning of when all of this is taking place. I tried to survey all the coins. If you want to move on to the next slide, these are the coins that I found. I wrote this down into a survey and I shared it with some of the other people on the team for them to respond to. This will literally just flick away behind us showing coins from the 6th century. Some of them might be very late 5th or very early 7th. As Shalendro said, we can't always date these things very accurately. The first and most obvious thing is that there's quite a few of them. There's no absence of coinage in this period. They're also quite varied and I'll say something about this. These are not types. This is not one of every type. This is pretty much one of every dynasty from the 6th century. In a few cases two from a dynasty. But essentially at that level, if we ran that presentation it was one of every type. It would still be running several hours from now. This is pretty much what Shalendro has just said. We know what the problem with counting public collections is. Skully, interest, rarity and aesthetics affect the number of coins that are there. Public collections are essentially a reflection of our interest in the source material, not the amount of the source material that's out there. To put that in perspective, the guptor coins that are extant compared to those horrible little base gold examples I showed at the start where I said I've seen more of the guptors than I have of those, the guptor coins are probably an order of magnitude smaller in terms of extant examples. But nobody has seen more of the base metal types than they have of the guptors. This brings us to the next slide. The fallacy fallacy. In formal logic fallacy fallacy is the assumption that because you have demonstrated somebody's method is incorrect, you have demonstrated that their conclusion is incorrect. Usually when your method is wrong your conclusion is wrong because most of the time there are lots of possible conclusions to pick from and you're just picking random if your method is wrong, so you're probably going to be wrong. However, this isn't one of those cases. Either cross regionally the money supply went up, or cross regionally it went down, or it stayed the same, or there isn't a cross regional pattern. So there's only basically four options. So if they're all equally likely, Sharma has a 25% chance of being right, no matter how bad his method is. So demonstrating that Sharma is wrong in his method is not the same thing as demonstrating that Sharma is wrong in his conclusion. And in his defence, because I think it's important, when he began writing about this in the mid-1950s, he did not base his argument on coins. He based it on an analysis of other forms of evidence and assumptions drawn from Marxist's theory. And he subsequently claimed that the evidence that he encountered confirmed his opinion. He sought to test his position by looking at the coins. So it's reasonably plausible. Maybe he's got the wrong method on the coins, but maybe he was right to begin with. Maybe, you know, if you measured the coins correctly, you'd get the same answer. I don't think she won't get the same answer. But... There are other aspects of his theory, which one can max it out a little bit. Yes. Also, in defence of Sharma, he openly states that his premise is Marxist's. Yes. And he says, here's the frame, and let's see if it works. And he's willing to be, he's perfectly open to be, refusing. And he does state his premises very clearly. And I think that's kind of important. We don't want this to be a Sharma bashing session. Sharma's work is very good, and did Michael's right. He openly states what his assumptions are, and then he seeks to test those. We think the way in which he tested his conclusions about the coins is wrong, but he is at least attempting to test it. So it's a good practice to begin with. Thank you. So, how can we count the number of coins? Well, we talked about collections. We can count hordes. This is what Dale does. Remember, Dale got basically the same results, right? Slight differences in nuance, but essentially the same results by counting numbers of hordes. We don't just have to count numbers of hordes. We can count the production of coins within hordes. How many coins produced in a particular period are found in hordes. We can count how many coins are deposited based on hordes. I'll come back to that one in a moment. We can count the number of dies. We can count how often coins are mentioned in other sources. We can count assemblages of coins from archaeological sites. In principle, we can measure things like velocity, commodity, money, credit. These are all important. I'll return to those. So from this point onwards in the talk, this should be thought much less of a talk. So if there's a point you want to make, shout out, and we will stop and discuss that point. Could I make one point here? There's one difference between the two findings of Sharma and Dale. There may be others, but there's one clear one where there's a difference in Cushan, because the hordes shown by Dale are much smaller, whereas the collections according to Sharma of Cushan are much higher. That, of course, can go down to the collectible nature of Cushan coins. But that's one clear difference between them. It is, and I think actually that's a thank you for bringing that up, because that's a point worth making, that the different things that skew our data are quite complicated. That difference is not down to collecting practice. That difference is down to the fact that the place that has recorded its hordes in a systematic way is Uttar Pradesh, and Cushan coins are not made in UP. They're made somewhere else. All of those coins are imports from somewhere else. The amount of coins in that is surprisingly high, given that the coins shouldn't be circulating there at all, because they're not mostly within the Cushan empire. Yes, but it's a good point that there are differences in the details of that count. The first one I'm going to do is give me an opportunity if we go on to the next slide to talk about how robust the data is. This is one of the intrinsic problems with any large data set. The underlying data has to be robust. How do you ensure that it actually is, short of physically checking every single coin to make sure the identification is correct? This is a set of data collected by Chattopadhy of inscriptions that mention coins from what he refers to as South India. I think, given the general area, he's referring to everything essentially maharashtrone lower when he's collecting data. He says that his data is not comprehensive, as much as he can, but that he's probably missed things. He does it to give the variety of different monetary terms that are operating in these inscriptions in the context of surveying South Indian coinage more generally. He doesn't include it for this purpose. What I've done is adapted it into a chart which shows how many inscriptions, when he's collecting a coin, he records for each century down to the 10th century. Here's a point I was making. It's essentially the same pattern. We get a little peak, so at the end of the cushions, guptas, then nothing, then we get a huge rise afterwards. Details aside, the details matter, and they are different, so we get the same pattern from counting the number of inscriptions as we do from counting hordes as we do from counting coins in collection. The point of this was to talk about how robust this data is. Is he right? This is where Daniel very kindly looked to the paper and told me last night, that he was completely wrong. On one particular inscription that Daniel happened to know. Basically, it's just something that I happened to stumble into a couple of years ago, and it rang a bell when I read it in Robert's paper. It's an inscription that is supposed to say that a queen donated some land to a Brahmin, and in addition to the land, he donated gold bullion, or gold bars, to him. I was looking at that inscription years ago for a different reason, but it sort of struck me as weird, and looking at the text, it seemed that there was absolutely no mention of bars of gold in the text. So basically what we have is in the De Omagari transcript for line 14, those of you who can read their magari, can see Suvarna si laka aia, which the editor called, I think, the editor of the inscription amended or altered to Suvarna salaka aia, and then claimed that it meant bars of gold. Now this is problematic to begin with, because the inscription doesn't say salaka, it says si laka, which is quite different. Even the inscription contains lots of what you might call spelling mistakes or non-standard orthography. So okay, let's assume this single word contains two spelling mistakes and si laka aia had been intended. Still, then the sentence wouldn't work. The syntax of the whole sentence would not be in good order. I've put in transliteration below that a bit of context for what we have in the inscription. You can see, or you can believe me, that it's definitely a land grant. The entire context is about land donation terminology. And what it says is that she, the screen call shower on me, gave a land to this Brahmin. There is nothing to say in addition, which is entirely something that's been added by the editor. There's no in addition. The land is the logical object of the sentence. The gold bars, even if we accept that it's gold bars, is not the logical object. It's a singular genitive form and you would expect a plural genitive if you were to say it was 50 gold bars. So anyway, what I'm saying, can we have a glimpse? I just want to highlight, I think what is coming is highlighting the written inscription, where you can see that just before the yellow bit, it is, the reading of the inscription is not dakshina, which was another assumption, or another commendation by the editor. It's dakshini. It's not a dakshina, a sacrificial reward. It's dakshini, it's to the south, or to the right. And after that, it's suvann shiplakaya, which was correctly read by the editor, but he meant it to something else. So can we go to the next slide again? What I'm suggesting, yes, okay, so we're just highlighting the points I've been talking about. So what it says is that I've donated something to the south of a pre-existing agrahara, of a pre-existing Brahmanical land, and I'm assuming, now can we go to the next slide, that we need a much simpler commendation, which is simply to amend shiplakaya to shiplakaya, which would mean that it's the name of a region. If you translate, if you don't amend suvanna shiplakaya, then you can just say it means golden rocks, or something like that, which is very plausible as a name for a village, or a spot, or a stretch of land, and I think it's a much less intrusive way of dealing with the text, and it gives us a much more logical result. So that's about to go to bullion. Where is this from? It's from a place for him in Iberby, somewhere in Maharashtra, I was just looking around to see where it is. I think it's close to bullion. I think it's in Pune now. Could be. Could be. No, the plate itself is down there. Okay, okay. Oh, is it? Right. So, okay. Great, thank you very much. So... Ah, now the reason that that particular inscription caught my eye is that is the... depending on how you read the names of the kings involved and the names of kings in other inscriptions and the letter forms from somebody who analysed them in the 1930s, right, then that inscription is in the 5th century. So it was the closest one to the 6th century in the list that I had. So I looked at it more closely and this is a good example of the robustness problem. The Chateaupage had depended on Dixit's reading and had therefore given its gold bars. So this is not a coin. This is possibly a commodity money, right? Possibly just bullion, right? I'd read it and seen Savanna and a term after it and gone to myself, ah, the editor has misunderstood and this is presumably a name and so this is a gold coin. The editor had excluded that possibility on the grounds that there were no gold coins that could have circulated there, right? But he didn't know enough about the gold coins, so I was willing to do that. But as Daniel has just shown, we were both completely wrong and there just isn't a reference to money in the inscription, or it's entirely falsely appearing in the data set. Is this a money rush recruiter? Yes. It's a rush recruiter. One more thing, even if the emulation to Shalaka is correct, I mean Shalaka does not sound like a gold bar and Shalaka would be a splinter. It's a long and thin and usually pointed. It's not the way I would expect Julien to look. Will you say why? Why I wouldn't expect Julien to look like that? I mean, if I told you that that Shalaka has a wire, it's not a bar. Why? Well, to me, it would suggest something rigid and a gold wire would be extremely pliable, so probably not. Anyway, it's more Shalaka, so we can look at what happens. So the point was to just say, our underlying data sets are not particularly robust in this regard. Some of the project is working towards solving, isn't it? You do so far, yes. So... So far, garbage in, garbage out. Absolutely. Re-enacting the exclusions of the archives of old Thucybian critique. So... So... Next slide. So the question that will always need to ask service what is it we're counting, right? We kind of had this point when we talked about Dale's data being Uta Pradesh Horns, and the Cajun coins are there, but that's not where the Cajun coins are made. So what exactly should we read from their presence in one place? So what is it you're counting? You're counting a region, are you? Now ideally, as a new mesmer test spot, we would like to count on minutes. We would like to look at each individual mint, manufacturing centre, and say what's happening here? Do we have more coins from here, less coins from here? This overcomes a lot of our problems. A lot of our problems are down to not comparing like with like. It's all very well to say there's only 2,000 of these coins and there are 15,000 of those. But if there are 15,000 small copper coins and 2,000 large gold coins then the 2,000 large gold coins is still a lot more money than the 15,000 copper. Now neither Sharman nor Dale's account takes any regard for the size or denomination or metal content of the coins. It doesn't take account the sort of context in which they appear. This group of coins gives you an opportunity to do that. These are what are known as the central provinces type of the Gupta Empire and they begin under Kumara, under Kumara Gupta and the last Gupta Emperor to issue them as Buddha Gupta so essentially right at the end of the empire the empire collapses Buddha Gupta is making this type of coins and Han King Toramana makes virtually identical coins. And then a whole group of kings by the Malkaris make them and then get some coins associated with Harsha. They don't have Harsha's name on them they say Shiladitya but the critical point about these is they're almost certainly made in the same place. This is one mint for the vast majority of these making coins it was making coins under the Guptas it carried on making under the Malkaris making them under Harsha so for 200, 250 years this single mint is making coins. They're the same weight they're equally ugly they right most people most people are not interested in collecting them look at the most recent catalogue type catalogue of the Gupta coinage and see how little space is dedicated to these as against the gold coinage of the same empire they are not being systematically distorted relative to each other in the kinds of ways the other data is so in principle you could count these unfortunately the only publication on these coins of the post Gupta is the publication of the single hoard of 1906 by Byrne and that's it so that's not enough data and one of the assumptions is that we just don't, that's it that's all that's ever been found and maybe they were very very rare I'll just show you where they kind of fit in in the next slide so here they are here this blue line so because we know at the end there almost certainly being issued a coinage we can probably assume that's where they were being issued or very close by right the way back to the beginning all of the other different series being issued in North India through the 6th century shown as bars running down small dotted lines if they begin before the 6th century small dotted lines if they continue and this one here is a mint next slide unfortunately most of them are not mints this is an abortive attempt to try and sort the present type catalogs of Alcon coinage into the mints which issued them right we have these amazing type catalogs they delineate 500, 600 different types for the coins but they haven't resolved the issue of which coins were issued in the same place how many different mints so the Alcons are not like with the Malkawi coins because they're a whole conflation of different series are they one mint, are they two mints are they half a dozen different mints under different policies very hard to compare and just to illustrate again Michael said Michael said this is an old point that the way in which the archives form determines our interpretation of them and those are affected by things that have nothing to do with the evidence I went back through the reports for Uttar Pradesh and I discovered six horns in the treasure trove reports of Uttar Pradesh the 1906 horn which was found in the tower is not in the treasure trove it wasn't treasure trove it wasn't treasure trove but I found six horns in those reports which consisted of several hundred coins described as silver drama now the interesting thing about those horns is every time a group of coins that could be described as silver drama is found but it's not one of these so it's Indo-Susanian Gadiopisa the horns correctly describe it as Indo-Susanian Gadiopisa they even correctly name west and satrap these six horns are unique in being anonymous and the only explanation I can think of since they all cluster immediately around Kanash is that these horns are in fact the same type of coins the central provinces style coin they back up to about I think two and a half thousand coins essentially unreported because they're not making it from being being reported as treasure trove to appearing in the literature they are anonymised in that process because when they pass through somebody looks at them and goes it's not worth giving any more information that these are silver dramas and so they disappear out of the so dead horse thoroughly beaten we'll move on to one of the ways in which we can skip ahead and say something about production and this is the die study this is a sculpture from 10th century it's about there it's from a temple in Rajasthan and it was originally identified as blacksmith at work but that's very unlikely given the arrangement what was interpreted as a bellows on this side is probably a bag and this is a myth producing coins the person in the centre is swinging the hammer the person on the left is holding the reverse die this is the top die with a pair of tongs so the die is a long column probably made of bronze or iron with the design the reverse design of the coin on the bottom of it the design of the front of the coin is on the anvil itself you strike the top and it impresses the design into the coin and then you remove the coin and put it in the bag it's useful to have a third person doing this if you've ever done this coins get stuck on the dies and you need to lever them off with a knife so if you've got a third person around just flipping the coin off that speeds the process up this matters because every single one of those dies has to be engraved by hand you literally etch into it at one to one scale with a chisel to produce the design it's a kind of job it's a bit like engraving a seal and that process with the person there hitting with the hammer it's quite violent it's violent enough that a French team who reconstructed this said that they were able to embed small silver coins into wooden posts nearby if they angled their dies right or wrong as it might be what you're seeing here are the people at the kind of front end of the mint generating the coin they're doing it with these handmade dies and they're hitting the dies with the hammer and the dies break so they have to replace them with new handmade dies which means we can connect every coin that is made to the two tools that produced it and because there are two tools one underneath and one above when one breaks and it's replaced we can form a chain connecting that back to the previous tool we can say this die was used with this die and that die was used with this previous die and so on back through a sequence die studies are the most powerful tool numismatists have available for studying coin production because they literally lay out the process that these three people are engaged in for you to see and they're quantifiable if you want to make more coins you have to make more dies now there's some uncertainty in this your coins have to be roughly the same size they have to be made in the same sort of metal if you're making copper coins you're going to break your dies much faster than if you're making gold coins if you're making gold coins you'll break them faster than if you're making silver coins if your coinage is large and the relief is deep you'll need more force your die will break quicker so you can't compare coins that have different production procedures but if the coins have the same productions procedures then the number of dies you used is a good indicator of how many coins you made and if you have a really good die study a really good die study not the type of die study then you can even detect changes in procedure so you can even test for that possibility and these kinds of die studies exist earlier in South Asian history and later than the period under study but we also have a little bit of data from the 6th century so three groups of coins the type at the top is an issue of Ramaditya who was a king ruling and sinned sometime early to mid 6th century AD there's sinned marked on the left this is the one I said I would show you earlier there are no examples of this coin in the British Museum we have no Ramaditya coins so people just weren't interested in coinage from that area the coin below is is this a yasha is this a draa is this a draa it's a yasha it's a yasha there you go there's the difficulty of attribution played out mine for you yes this is part of a group which has received the clumsy title the DNNVY series after the names of the kings who issued the coins they're believed to be a small state somewhere in the Punjab who succeed to the Kidderites in that area and are ruling from the late 5th to the early 7th century Joe has another opinion on who they are would you like to voice it now we don't know who they are the Kidderites are not in that area yes because they came after because they came after no because the Kidderites are further north so they can't be succeeded so this is part of the problem of unwrapping the political and the coin at the very bottom is a coin from Santata imitating a gupta type this is part of a very long series of coins that we believe runs from about 500 to about 800 this coin is relatively early at the sequence so it belongs sometime in the mid 6th century so all of these coins are at the same date and we have data for die studies for all of them now the Ralliedigia coins I have a catalogue of which I haven't published 50 or so examples it says on the slide 55 examples from 37 dies so based on the number of examples I've seen, the number of dies that we've found we can calculate how many dies were originally used 113 if we know how long they're issued for there were happening to guess but Ralliedigia appears to be a single king so we can guess any number that seems reasonable for a single king depending on how we want to inflate or analyse the numbers we can generate a number of dies used per year and so this is an index of the number of coins that were made right, that's what I was waiting for so the first common is how long the coins were issued for second common is the number of coins divided by the number of dies that we've found we're talking about the obverse dies here the die that appears on the anvil underneath the coin that's then gives you a prediction for the number of obverse dies that were originally used 115 113 and then you divide that by the number of years you would think they were made over to give you a number of dies employed each year which is an index for the number of coins that were being made so the Punjab group we don't have any die data for the sixth century but for the coins that come in the seventh the person who just published a new examination of this group examined a small group of them and found 12 obverses from 16 coins now what's really interesting about that is he concluded it was a very large series on the basis of that piece of information it's actually quite a small sample I wouldn't put too much weight on it but it actually implies it's quite small as a series issued over a long time the number of dies is very small the sanatata coins nobody's done a die study of these but no man and Bose who just published a new catalogue of them did and this was what put me on to it would mark that they had seen virtually no dies die duplicates in all the coins that they had examined so I ran through that catalogue and discovered that certainly as far as the catalogue's concerned yep that appears to be the case there are only three die duplicates in the entire 160 in coins that they have now that might be slightly inflated because of course they're trying to show a couple of each type rather than multiple coins of the same king so maybe they excluded some that were actually die duplicates but given that they themselves had already testified that they'd seen virtually no die duplicates this seems to be broadly in keeping those are issued over a much much longer period so obviously while the number of dies that we're estimating is enormous compared to sinned the actual number of dies per a year isn't that large an increase we'll go on to the next one we can convert this back into our production chart to give ourselves a visual indicator of size and so on the left you can see the scale of the sinned motarn issues on the right scale of the San Matata and those are the Punjab issues there so the Punjab issues are tiny they're insignificant compared to but unfortunately we have no data for the canals issues we have no data for the eastern Ganges issues I think we have like one two coins of the later later cultures so we don't have enough coins to do this we have more and more refinement here than we do here here we can only look at the whole services and go here I can break it down by the reins of particular kings which is a useful check as well that I haven't wild the misestimated 25 years for Baladetius right and his figures are a broad in keeping with them the interesting thing about those is add all those back together and you're in the same order of magnitude as the Gupta issues from the same region so what's happened in this period is that you've had political fragmentation and political fragmentation has made the coins invisible nobody studied these are literally there are no tight catalogs of the sin at all when R.S. Sharma came I think there were about three senators probably about 30 on the Punjab it's just so impressive so it's so reading the British Museum collection as a record of production compared with that is just the two are completely different and in fact the second of these seminars we had in the project was the one I gave on Gupta coins in general and in that I demonstrated exactly the same thing for the transition from the Cushan to the Gupta period that contrary to the graphs or production action rose an order of magnitude between the two empires so it's the same pattern we're seeing again once we can look at the production levels they don't look like the number of extant coins they don't look like the amount of coins that we actually have so move on to the next one so back to the paucity argument we've skipped over that first step in the paucity argument and gone to how many coins were actually made so we've been able to skip the first two and go straight to how many coins are actually being made but that doesn't alter the fact that we looked at at the beginning which is the next two steps don't follow just because you're making a certain number of coins doesn't mean that the number of coins in circulation of the period they're made goes up or down and just because the number of coins in circulation goes up or down doesn't necessarily mean the amount of money available so the dye study gets us so far and short circuits some of the problems but it doesn't provide a definitive answer to the question it's also worth saying that what I did there was to cheat by picking out the three series where I could actually do something interesting as Shellendra suggested at the start for the vast majority of coinages in the sixth century we don't have dye studies we don't even have type catalogs in most cases we're in the same situation we are with the central provinces type that we have one publication often just on a hoard or a particular find that would be the case for nearly half the series I looked at in that survey I think another thing is that as I said in the beginning they are not attributable coins there are coins if you go to say for example coin data in Rajasthan there are lots of indosysanian type coins and they are massive orders thousands of coins 40,000, 50,000 coin numbers in the hoard but they are not attributable so collectors are not interested how many indosysanian coins can one have in the collection if you are example one is going to be sufficient for the entire series to be represented in your collection you're not going to collect any more so that bias is always there they're not very beautiful either not totally if at all and of course that affects the scholarship so one of the things that I thought I might do is that no we don't need to go on a slide yet one of the things I thought I might do is those silver central provinces times they're so perfect for measuring whether the production is going up and down because they have individual kings names you can put them in order you can date them fairly accurately they're all of the same type I thought well I can run through those and see if for the number of dies but I can't because the museum, the British Museum was perfectly happy to have 12 of them an Oxford was happy to have 15 and when Burns published the hoard he was happy to show 16 images so the two and a half thousand coins that are out there I can see images of 30 of them so I know there's a diolink between some of them but how many? In a bag probably In a bag I'll show you a bag I would be very surprised if anybody has looked at them subsequently that would be my assumption they're literally sitting in a basement in a bag with nobody having I've gone to an Indian Museum and been shown a bag like this where somebody's pulled it out and there's a carrier bag which is beginning to biodegrade and they put it on a desk in front of me and that's the carrier bag it arrived in the museum and they've been like well they may have been distributed they might have been there are there are a bunch of coins running on the market in the early 90s there was a whole lot of these coins and most of apart from the old ones which are in the past or in the end most of the later ones were acquired out of that they turned up here in the market and a bit of our stuff our coins from that also came from a collection of Indo-Sythian coins which were acquired which was a collector's seat and he probably has acquired from from the same long time from the old stuff from him We don't know where the coins are but we do know that the tea is in the next round Aha That was about to ask isn't this usually the point we break so, yeah, we'll break there then I mean just the last thing about before we go to tea is that what Robert did for those three little blobs on Sin, Kashmir and Samadata you can obviously do that kind of analysis on other series for example a good example where there are lots of coins that would be Visnukundin I mean there is massive amount of data there you can actually see dice on But the Visnukundins don't have a good type catalogue yet No, but there is not because they only have one type There is no type Visnukundin type But there are variations there are small variations and they look like they might be regional that's one of the arguments that's been put so they're almost a group One of the things is you can do a dice study but you often need somebody to have done the basic leg work of dividing them into types and recording the finds and publishing syllogies before you can do the dice study And they would do that on the basis of pooling the information from lots of different isolated collections Yes, so that's what's happened with Samadata I could, the Rhanoditia I had a catalogue of but the Samadata did not only do that because somebody had done a type catalogue which was the product of several scholars and quite a bit of work But that is a type catalogue so the dice I think it's just not very visible because they only illustrate two coins and one of those rubers I've seen 30 or 40 of them done whereas in that catalogue there are two so I think the resources are probably there to do more detail So let's have the tea