Inca Huasi (ancient lake)

Transcription

DR. BARREIRO: Engineering a world with strings attached: the place of the khipu in building the Inka Empire. Eminent scholar, friend of our museum, Dr. Gary Urton will now share the fruits of his research and insights in a presentation on the khipu, the knotted mnemonic device of the Inka people. Gary Urton is the Dumbarton Oaks Professor of Pre-Columbian studies and chairman of the Department of Anthropology at Harvard University. His research focuses on a variety of topics in pre-Columbian and early colonial Andean intellectual history, drawing on materials and methods in archeology, ethnohistory, and ethnology. He is the author of many articles and editor of several volumes on Andean/Quechua cultures and Inka Civilization. His books include: The History of a Myth, The Social Life of Numbers, Inca Myths, and Signs of the Inka Khipu. A MacArthur Fellow as well, he is the Founder/Director of the Harvard Khipu Database Project. Welcome Gary Urton. DR. GARY URTON: Thanks very much Jose for that kind introduction and to begin I just wanted to thank the Smithsonian Institution for inviting me here to take part in this very important set of events related to the planning of an exhibit on Inka engineering. I do just want to say to begin with that when I was doing my doctoral dissertation field work that Ed Frankamont who was just honored by John Ochsendorf was doing his fieldwork in the village across the way and I spent much time with him and learned a tremendous amount from him. I also want to say that I learned about John Ochsendorf's work about ten years ago and in my book it's the most exciting work that's been done in the Andes in a very long time. If I can get my image up here. Can someone just… In Secretary Clough's introductory comments this morning he was talking about how the Inka had accomplished all of these fantastic engineering constructions and accomplishments without a writing system and that is the case. They did on the other hand, have a recordkeeping system and I've been involved in the study of the recordkeeping system for the past 20 to 25 years or so. I'm usually invited to speak on the recordkeeping system when we talk about it and wring our hands about whether it was a system of writing or not, and I just want to say I'm very pleased to come to a conference to talk about the khipu in terms of a straightforward recordkeeping and ask the question about the value of a recordkeeping system in relationship to maintaining calculations, maintaining data that related to this extraordinary world of engineering accomplishments that the Inka were so noted for and that we're seeing today. Just to give you a little bit of the context of the keeping of records in the Inka Empire, not to say that you don't have it already because from Ken Wright's presentation and from the presentation on Cusco and John's on the Inka, we can understand that there were an enormous number of engineering projects that were put in place by the Inkas and that were maintained by the Inkas over time and if one thinks about it for a short time you realize that this all involved a tremendous amount of calculation of measurement, of the recording and data, of the calculating of ratios and proportions and of calculating things for model to and scaling them up to the actual place they would be put in place in the environment. Some kind of recordkeeping system would have been essential and I'll get to that after once again, a little bit more laying the foundation or establishing the broader context for its use. We are talking about the Inka Empire which was the largest empire in the pre-Colombian new world stretching from the present-day border between Ecuador and Colombia and running some 5,000 kilometers along the spine of the Andes to a little bit south of Santiago de Chile. The Inka Empire was divided into four quarters; the name of the empire was Tawantinsuyu, the four parts divided together. To emphasize a point made by some of the other speakers, the Inka empire was situated within one of the most complex environments in the world spanning the region from the Pacific coastal desert to the west to the high Andes rising to some 20 to 22,000 feet above sea level and then dropping precipitously down to the headwaters of the Amazon in the zone called the Seja de Silva [phonetic] they gave off to the upper Amazon basin so this was a tremendously vertical environment and that verticality and that change of environments was compressed in some places within only some 200 miles from the Pacific coast to the headwaters of the Amazon in various parts of Peru and Ecuador so it was a very complex setting. One observation about this was made by my colleague at MIT and John's colleague at MIT, Heather Lechtman, when she talked about the nature of Andean technologies and noted that abandoning the western overemphasis on sophisticated hardware because you simply don't find that when you look at the engineering in the Andes, Andean systems of technology were heavily weighted on the side of management in two senses; management of the physical environment and environment that was and is the dramatically dominant feature of the Andean zone of South America and management of the social units of production whether at the community or the state level and we just saw an example of that with the building and the maintenance of the Huinchiri Keshwa Chaka, the bridge that John just told us about, but I want to talk about that issue of management because I think this is where the khipu came to play. It was probably the principle instrument of management in the Inka Empire. As I mentioned the empire was divided into four quarters, it was known as Tawantinsuyu, the four parts intimately bound together, the whole was bound together in terms of the movement of personnel and material by the Qhapaq nan or the Inka road system. There were two main trunk lines, one that ran along the coast, and one ran through the central highlands with roads that connected them at the certain strategic places and some of those were the spots of the bridges that John just told us about. That road system passed through the high Andes and down into the tropical forest, some images of which you see on the right and in the lower left, and you see a small stretch of the Inka road along the coastal desert and in some places these had two or three parallel tracks of road. The system was extraordinarily complex in terms of its construction and certainly in terms of its use in regulation of traffic moving along the road. There were a number of installations along the road. These were important for us to think about in terms of engineering, of engineering those buildings and of just putting all of the infrastructure in place for travel along the road, so we have large installations like the one on the upper right which is a site in northern Chile that has a very large compound building called a kallanka and then you have just tens and probably actually hundreds of small weigh stations called tampus, like those shown on the bottom right, where travelers could stop and refresh themselves and sleep for the night, change their llama herds whatever you do with llama caravans. That infrastructure all existed as a part of it and then of course we just heard from John the master of telling us about and bringing to us the message of the Inka bridges themselves. They actually made the whole thing possible by allowing travelers to cross those great chasms through the Andes. What's interesting to me here about the business of rebuilding the bridge now is that sort of work requires cooperation and it requires coordination and it requires management and that's the sort of thing that we're aware that exists to the present day but that existed in the Inka times as well and I want to talk more specifically about that in a moment. We do have then a whole set of buildings and various kinds of structures with their particular architectural features that were the central facilities of the Inka empire that were spread throughout the Andes and these buildings which are not terribly complex themselves but are quite sturdy, of course, these are a part of the products of engineering that we need to be aware of and take account of and that includes then also the extraordinary work in stonecutting which the Inka masons were masters at that we heard about this morning from the organization and the building of Cusco and that was so important. Again here just to emphasize that fact that in some of these large buildings some with internal spaces that span some hundred to 200 meters or so those buildings had to be built, the structures, the walls had to be built, and were built with great regularity in terms of the layout of the different features of the sites like the internal niches. This required measurement and the calculation of various features of the constructions themselves and I think those calculations were recorded on the khipu and we'll talk about that in a moment. The stone working of course was not terribly complex in terms of the tools so again, they were working stones primarily with stone on stone, abrading stone, working with water, thawing and freezing, wooden wedges that were driven into fissures in the limestone and then once the stones were cut and shaped then they were moved into place by manpower, by human power that was organized to move the stones across the landscape and that's a very important part of this system of management, managing the labor of doing all this. I happened to have the fantastic opportunity to see that kind of thing at work under different political economic circumstances when I was directing an archeological field school at Tiwanaku just south of Lake Titicaca in Bolivia in 2006 when the state was trying to reassemble the stonework that was the stonework on the summit of the main pyramid at the site of Tiwanaku so it has since been ravaged and a lot of the stones broken and thrown down from the top and they organized local labor to get those stones back in place and it was done then by groups of workman who came together in working with wooden staffs, then they pried up the broken stones from the lower part of the pyramid itself, raised them up and moved them onto sledges, moved the sledges using nylon ropes, the modern-day version of the grass ropes of course, the grass rope would serve equally fine here. Moving them onto the sledges, pulling them up from the lower level and getting them up to the top of the pyramid, and then of dragging it across the summit and then unloading it from the sledge, tipping it over and once they had it in place where they wanted it, slipping it off the sledge and getting it down. Then of course, once you accomplish a task like that then you have to celebrate things so they pull out the coca and the drago [phonetic] and everybody has a drink and has another chew of coca to get regenerated for pulling the next one together. That's the kind of activity that has to be organized, people would prefer to be out in their fields doing their own business but someone organizes this kind of work and this was a sort of organization that went into not only building bridges but building roads and all of the infrastructure of the building. In the Inka Empire this was done through a decimal system of organization. The Inka levying tribute in the form of work time so every subject in the empire was required to work a certain number of days every month for the state and those workgroups formed at the local level in workgroups of ten. You would have chunka, ten, a group of ten workers who could be put to a small task; five groups of ten workers could be put together into a larger group, workgroup of 50. Two groups of 50 could form a larger workgroup of a hundred and so on up to work groups of 500, 1,000, 5,000, 10,000, up to 40,000 or so. They had the organizational structure for bringing together workgroups of various sizes and set them to work doing the work of the empire. This was explicitly a context for the keeping of records and we have statements in the chronicles about how the Inkas kept track of time and they kept debits and credits. If you worked more than you were required to work then you got credit for it and so there were was a whole sort of accounting of time that was important in recordkeeping. Also for bringing those workers together you had to have an efficient census system to know the count of the population in this village and that village and organizing them into groups and we know from the chronicles that they did keep a census. They did yearly censuses and revised the whole thing every three to five years. It's not exactly clear. All that information was recorded on khipus, these knotted string devices that we're going to be talking about and all of those data went from the local level up to level of an administrative center and ultimately they would be sent to Cusco. The data could be shuttled around the empire as well because the Inka had a system of messengers called chaskis [phonetic] who traveled with the khipus with their information taking them from one place to another. Here we have a late 16th, early 17th century drawing in which a runner is running with one of these knotted strings but he has a little sign on it that says carta [phonetic], so to inform the reader that he's carrying a letter, right, so but what he's actually carrying is the knotted string record. These knotted string records would be produced locally. From there they would be sent to a central place, ultimately they would all go to Cusco and in the end they would be redacted or brought together, synthesized, aggregated into larger khipus, these very large knotted string devices. We want to talk about the khipu and you're looking here at a sort of potpourri of several khipus from what is the largest collection of khipus in the world. That's in the museum for Volkerkunde or the ethnological museum in Berlin. They have about 350 khipus there. There are some 850 khipus that exist that are - - from the end of the Spanish era. These devices are made of spun and plied either cotton or camelid fibers so llama or alpaca fiber spun and plied and often the alpaca fibers are dyed in brightly colored forms. I think the khipu was important in engineering for a variety of ways and I'll just talk about a couple of them here. One is a paradigm and I think the Inka road system which its main route and with the notable places, the named places, the places that would have all been understood as having a particular identify and probably, given what we know about the Inkas also, a hierarchical relationship one to the other I think was thought of our could be thought of as similar to the khipu cords which are these strands of spun and plied fibers that are knotted and so I think there was a relationship here between the sort of general structure of the khipu and the structure of the Inka road itself. The city of Cusco itself as we heard this morning was actually divided into two halves. There was a moiety system; there was an upper Cusco and lower Cusco. The whole valley itself was organized for ritual and political purposes by a system called the ceques system which was a series of 41 imaginary lines that went out from the center of the city and then organized sacred places called huacas, some 350 of them along those 41 lines and that was the sort of framework for ceremonial life and for the ritual calendar in the city of Cusco. You had a system something on this order and those ceques lines, those lines of orientation with their sacred places located along them I think would have been understood as similar to the khipus with their strings and their knots. Let's look at the khipu as a recording device. We know now we need to record numbers, we need to record them for the administrative system and for a variety of other purposes. Khipus have basically the form you see in this example which is from northern Peru, from Chachapolles [phonetic], has a main cord we call the primary cord that runs horizontally from which are suspended all these other strings which can be of a variable number from between two to we have some khipus with 1,500 cords so these are quite large things that take two people standing side by side to actually suspend the khipu from their hands. They have cords that are suspended from them that we call pendant strings. Lots of pendant strings carry what we call subsidiary strings or secondary strings and you can have subsidiaries of subsidiaries of subsidiaries and we have examples of six hierarchical levels of subsidiaries affixed to a single khipu so this was a hierarchical structure much like that hierarchical decimal system of organization that we saw. We have other kinds of structures in also like top strings that bind together the attachments and groups of pendant strings, I'll talk about this in a moment, and another structure we call the loop pendant. I won't get into that so much now. Here they're recording numerical values; they're recording all the quantitative data important for engineering and for other purposes like the census. Where we talk about numbers we have to note that unlike their contemporaries the Maya and the Aztec who used a vigesimal system, a base 20 system, the Inkas used a base 10 system like us. I've always thanked God I worked in the culture that did the decimal system rather than the vigesimal system, much easier to work with. For administrative purposes they spoke the Quechua language and Quechua was a base ten decimal system. The knots themselves are organized in tiers which are the tiers of a place value decimal system of recording decimal values in those tiers so that the decimal system works as our system works so we can write a number in a horizontal string. We know that the ones are to the right; the tens are to the left of those, the hundreds are to the left. The Inka system worked the same way but its rotated 90 degrees clockwise so the ones are at the bottom, then the tens, then the hundreds, then the thousands. They had several different kinds of knots they tied to record values in those different places. The majority of the knots were used to indicate one or another of the values between one to nine down in the units place down here. All the tens, hundreds, thousands, and ten thousands were signified by an overhand knot, just a very simple I call granny knot, overhand knot, and tied in a place value. Basically that's all the information you need and now you can go out and study your khipu and you know how to read the numbers on the khipu. They would tie different groupings of those knots then in different tiered configurations to assign numbers and its important here to note that they knew the value of zero and they used the value of zero in their calculations. In terms of the calculations the khipu then was not like an abacus. An abacus you move beads on rods. You can shift them up and down. The khipu you tie fixed knots so the calculations have to already be done and so the calculations we think where we had some chroniclers accounts where they talk about people dropping down on the ground and shuffling stones around or kernels of corn around and in some cases then we learn about counter stones with multiple slots where we think they were working with those stones working beads around them to do the calculations. One of our colonial chronicles shows a man holding a khipu with a board here with different slots and there have been about two tons of ink spilled on trying to analyze how that counter stone on this particular drawing would have worked. We have some khipus that give us evidence that in fact they were recording numerical values and these contain those top cords that I talked about. This is a drawing of a khipu in a museum in the Natural History Museum in New York. In this khipu you have groups of four cords that are bound together by a top cord and you have then the number recorded on the top cord that is the sum of the values of the group of cords that the top cord binds together so 612 overall recorded here, 612 recorded there. It's a system of sort of checking the calculations and we have several other examples of that. Now we know how to read numbers on khipus but our big problem and it’s the great challenge of khipu studies is so if you can say that string records the number 602 then 602 what. There is our stumbling block is how were they recording identities and we think one of the main ways they were doing it was through color. There's a fair amount of dyeing of camelid fibers and the Inka grew cotton in some four or five different colors and so the cotton khipus are spun in groups of cords in different color so color was very important but I don't have time to stop and talk about color now or we'll be here till tomorrow morning. But that was very important and we have to deal with that issue. There was a cadre of khipu keepers, people who were responsible for this; there was a hierarchy of them. The head khipu keepers were in the capital in Cusco and then mid and lower level ones were distributed out into the countryside. Census keeping was a very important part of the task they were charged with and I wanted to show you the analysis of one khipu where it seems to me we have a pretty clear example that this is probably a census khipu, although again we can't read it, we can't actually read the names of anything but what we can see is the organization of numerical values. The one I'm going to tell you about is one of six khipus that are tied together. We have several cases of this. This is from the museum for Volkerkunde in Munich and it's that khipu I'm going to tell you about. We'll draw a schematic diagram of the cords of that khipu and here they are. There's a total of 74 strings and they're organized in these groups for the reason that I'm going to explain to you. In terms of the way this table is laid out there are cord numbers so that at the top that's the first cord so that’s just arbitrarily beginning from one side of the khipu. We don't really know if we're looking at the right side of the khipu when we're here or if we should turn the khipu over so right and left is a big problem in the study of the khipu. Here's the first string, the second string, it has a subsidiary the third string, with a subsidiary fourth string, subsidiary fifth, sixth, seventh, eighth, ninth, 15, 16, 17, 18, 25, 26, 27, so I grouped these in a way that's going to look quite odd to you but I'm going to explain it to you in a moment. The cords are basically of two colors; they're AB light brown or MB medium brown so almost all of the cords are one of the other of those two, AB or MB. Then to go on then we're going to read the knot values, you know how to read the knot values and those are recorded in those blue/white boxes in those columns and then we're going to sum those and those are going to be recorded at the far right on the different columns. We know now the first cord we see here it contains a value 102, cords 2 through 4 with their subsidiaries they also sum 102 divided into the cords themselves and their subsidiaries and we see here the cords total 54, the subsidiaries total 48 so that's interesting you say. One cord totals the same as three cords and their subsidiaries. But then what's interesting is we'll step back and we'll take this group of figures right here as a sort of guide and then we'll look at this first group of strings that we set apart here, 5 through 9, 15 through 19, 25 through 29, they sum the same as the first cord itself. Its subsidiary carries 14 and it's the same as the strings 10 through 14, 20 through 24, 30 through 34 there. The next string has 13 and that’s this value. The next one has ten that's that value. The next one has 12 I think there's a bad job of adding here, the guy meant to write down or knot down 12 and he knotted down 14, then there's 24 here down to there and we actually have 104 registered here but I think it's referencing this 102 which is the base number that they're interested in. I think what we're looking at here is a khipu of one of these units of 100 so 100 was this level in the decimal organization. Oops, let me get back here, is this unit you have groups of 10, then groups of 50, then groups of a hundred. Hundred was a group very important. It was actually a name, a size group that was often related to the main social group in the Andes which some of you will know as the IU. The IU is alternately called a pachaca [phonetic] a group of 100 so we have the group of 100 here and I think that may be what's being accounted for in this khipu we have 102, we have a hundred down here and as we recall those groups of 100 are composed of 2 groups of 50 so appear again there's a group of 54 and a group of 48 equaling the 102. I think this is the way that kind of information that had to be pulled together, synthesized, and checked and rechecked so we have three different recordings of the value 102. That's how many people are in this accounting group, right? What did I do? Okay. I think that this was a cord of a cord keeper. Then we have other cases like that. I'm just going to give you a couple of examples just to show you how they were handling numbers, organizing numbers to arrive at the same value but using different groups of numbers, different groupings of numbers to arrive at the same value, something that one might want to do if you're establishing or calculating ratios or proportions et cetera. Here's a khipu that has groups of khipus divided into two. The first group totals 152, the second group totals 152 and you can see that the sets of numbers that they used to arrive at those two equal values are different so they're getting at a value of 152 and they're trying to see how you get there in different ways. Here a khipu divided into two big groups 247 for this group, 247 for that group. They may also be doing some kinds of checks and balances business here as well. This is a complicated one where you have on the first string 30 you have 32, the second string 37 and 26 totaling 63 and then that's on those two strings and then down here there's an organization of strings into 37 and 26 like we saw up here. Again, there's a lot of ways you can skin this particular cat I guess and get to these values adding sums and checking sums in various complicated ways. Then the last one I'll show you is this khipu, a beautiful khipu in a museum in Paris that has, if we blow up the figures here, we can see here's the color key and the first eight strings, strings one through eight, total 43,372. The remainder of the strings are in groups of nine from the ninth cord up to the thirty-second cord in these various groups and they total the same as we have up there. Again, it's showing we don't see the calculations but the calculations must have been done in some way and we can see then how they were able to record these numbers that were the result of a fairly complex numerical calculations that—and I think all of this was at the core of the recordkeeping and of the calculations that were central to this building of the Inka empire which was an empire based again, as Heather Lechtman points out to us, on management, managing the environment and managing social units and the managements of those social units was in terms of the census and administrative system. So I think we can understand the khipu as a critical tool in the practice of engineering in the Andes. Thank you very much.