AT LAGUNA de Los Condores near Chachapoyas in Peru, laborers cut down trees to gather lumber for a hacienda under construction. As they watch a tree fall, one of the men glances across the lake and spies a painting on a cliff face. Motivated by the possibility of discovering gold, the group treks around the lake until they find a site adorned with primitive rock paintings that lead the way to seven, above ground burial chambers.
They have discovered a burial site for the Chachapoyas, an isolated civilization that ruled the Northeastern lands of Peru from 700 AD until the Incas conquered them in the 15th century. The burial chambers yield ancient artifacts, ceramics, textiles, and more than 200 mummies.
But these men aren’t interested in the historical significance of their find. With gold in mind, they hack into the mummies and loot the site for several months until the police shut down their operation.
Eventually the valuable artifacts are recovered and the site is handed over to Peruvian archaeologist Sonia Guillén and her colleague Adriana von Hagen. While much of the scientific world marvels at the discovery of 200 mummies in their funerary garb, a smaller group of scholars turns their attention to what else is in the burial chambers— 32 well preserved knotted strings called khipu. Is this the clue they are searching for? Were the ancient inhabitants of Peru finally speaking to them from the past?
The mystery in the Andes is how the Inca—the largest empire in the Americas until the Spanish came in 1532—could rule and administer an area covering thousands of miles with no form of written word. At the peak of their power in the early 16th century, the Inca state stretched along the Andes from southern Colombia through Ecuador, Peru, and Chile, over to Argentina, and into the Amazon basin.
The rulers in Cusco kept detailed records of births, marriages, tributes, religious rituals, and other activities throughout their empire using knotted and dyed strings. For the last 500 years, historians and scientists have been trying to solve the mystery of the khipu and decode the information within. The most recent discovery of 32 khipu at Laguna de los Condores a decade ago, has provided valuable clues to the mystery of this ancient form of communication.
The word khipu is from Quechua, the language of the Inca, and means knot. A khipu is made of cotton or sometimes camelid fiber and varies in color from white to brown or green. When wrapped for storage, they resemble a string mop, but when they are spread out, they reveal a complex array of knots and strings. Most have one main cord on a horizontal plane with pendant cords hanging from it.
The pendant cords may have their own attachments called subsidiaries. Some khipu have as many as 10 or 12 levels of subsidiaries. Knots are strategically tied on the pendant cords and subsidiaries to represent numerical values.
Most of the 700 khipu in existence today are from the time of the Inca—the early 1400s to the early 1500s— but it’s hard to know exactly where they came from because most were dug up by looters and sold to museums and private collections.
Much of what we know about the khipu and the type of information it stores comes from the Spanish chroniclers. Through drawings and text they described how the khipumakers or khipukamayuqs kept records on all aspects of Inca life including census data, landholdings, and legal proceedings.
“If one reads the chronicles you continually come across references to the khipu whether they are talking about an Inca myth or Inca dynasty or social structure,” says Gary Urton, a Harvard professor and creator of the Khipu Database Project.
“Repeatedly one comes upon references that say all the Inca really knew was recorded in the khipus.” Urton refers to a translation from one of the most well-known Spanish chroniclers, Garcilaso de la Vega who describes khipu practices:
Although the quipucamayus [khipu-makers/keepers] were as accurate and honest as we have said, their number in each village was in proportion to its population, and however small, it had at least four and so upwards to twenty or thirty. They all kept the same records, and although one accountant or scribe was all that would have been necessary to keep them, the Incas preferred to have plenty in each village and for each sort of calculation, so as to avoid faults that might occur if there were few, saying that if there were a number of them, they would either all be at fault or none of them (1609-1617).
The chroniclers relied on the khipukamayuqs to give them information about the Inca state, especially in the early days of the conquest when the Spaniards were trying to understand the history of the Inca and the vast territory they controlled. In a 2003 article in Latin American Research Review, Galen Brokaw, a professor of romance languages and literature at the University at Buffalo in New York, describes a passage from the Discurso sobre la descendencia y gobierno de los Incas signed by a Fray Antonio in 1608,
“The viceroy Cristóbal Vaca de Castro gathered together the oldest inhabitants of Cuzco and the surrounding area and asked them to give an account of their origins and history. The responses that he received varied greatly and often contradicted each other. Perceiving the frustration of the Spaniards, these ancianos (elders) suggested that they seek out the old khipukamayuqs. They explained that prior to the arrival of the Spaniards, Atahualpa had attempted to revise Inca history by burning all the khipu he could find and killing the khipukamayuqs. Those that escaped had been hiding in the mountains ever since. The viceroy sought out these khipukamayuqs and had them brought to Cuzco with their khipu. He then posed the questions to them, and had their accounts translated and transcribed.”
Another prolific chronicler named Felipe Guamán Poma de Ayala included drawings of khipu in his book Nueva corónica y buen gobierno (c. 1615). He depicts treasurers with khipu, native officials carrying khipu, and a lord in front of his warehouse with an accountant and a khipu. One drawing shows a rolled up khipu—with the Spanish word carta (letter) above it— in the hands of a chasqui, one of the long distance runners the Inca used to deliver messages throughout their sprawling empire.
What puzzles khipu scholars is that although the Spaniards relied on the khipu, they made limited efforts to understand them. Galen Brokaw explains that because the Spaniards came from a society that used alphabetic writing to communicate, they found it hard to recognize a series of knots and strings as an important form of communication. As a result, they failed to preserve or study the khipu to the same extent they did with the pictographic texts of the Maya and Aztec civilizations.
Gary Urton says that the Spaniards deemed alphabetic writing to be a superior form of communication to the “primitive” khipu, so they were relatively disdainful of it. He laments that although there are about 2 dozen Spanish transcriptions of the khipu, researchers have yet to find the “Rosetta khipu” that would provide a match between a transcription and a khipu. “The wildest dream of the khipu student is to find the ‘Rosetta khipu,’ ” he says.
The big challenge is how do you move from classification and structure to meaning and some kind of interpretation?”
In 1996 an amateur historian from Italy claimed to have found what Urton sought when Clara Miccinelli, a descendant of Neapolitan nobility, announced that her family archives contained a khipu along with a corresponding Spanish translation. But because she has granted the international scientific community only limited access to the materials, khipu scholars are skeptical of the document’s authenticity and do not include it in their compendium of khipu knowledge.
Perhaps because the Spaniards didn’t quite understand the khipu, they destroyed many of them and even declared them “ungodly.” So as the khipukamayuqs died, their knowledge of the khipu died with them. It would take nearly 500 years for a serious study of the khipu to begin. Early in the 1920s Leland Locke, an archaeologist at the American Museum of Natural History in New York, began his quest to decipher the code of the khipu.
His groundbreaking discovery came when he realized that the top cords of the pendants he was studying recorded the sum of the values on the pendants joined together by the top cords. He and other Andean scholars declared the mystery of the khipu solved: they were simply a series of calculations represented by knots on strings.
Interest in the field of khipu research resurfaced in the 1960s when mathematician Marcia Ascher and archaeologist Robert Ascher began their analysis of 191 khipu. “There is a logic to the khipus so that the colors of the cord, the spacing of the cords, the places the cords are attached, all of these things carry meaning in a symbolic system,” Marcia Ascher explains.
In their book Mathematics of the Incas: Code of the Quipu, published by Dover Publications in 1981, the Aschers concluded that in addition to arithmetic, the numbers on the khipu stood for non-quantitative information that could include labels for names of people or places.
After the Aschers published their book, textile expert William J. Conklin added yet another layer of complexity to the khipu. An architect by trade who had turned his focus to the structure of Andean textiles, Conklin drew attention to subtle differences in how the knots were tied.
Some of the knots were tied to the left, while others where tied to the right. There were left-hand knots and right-hand knots along with left-hand spins and plies and right-hand spins and plies. Conklin suggested that not only was the placement of the knots important, but perhaps the direction the knots were tied was important as well.
The wildest dream of the khipu student is to find the ‘Rosetta khipu.”
In the early 1990s anthropologist Gary Urton began his study of the khipu. “They seemed to be the device the Inca made and used that encoded all of their most complex and precious information,” Urton says. After spending a year with a weaving community in central Bolivia, he wrote the book The Social Life of Numbers: A Quechua Ontology of Numbers and Philosophy of Arithmetic, published by the University of Texas Press in 1997. Urton set out to analyze the khipu knot by knot and string by string.
By 2000, he realized the only way to manage all of the information he had collected was to create a computer database, so he hired software programmer Carrie J. Brezine, a spinner and a weaver with a mathematics background who has studied Quechua.
Currently the database contains information on 350 khipu with a total of 30,000 strings. With the aid of the database, Urton and Brezine search for patterns within and among the khipu. Recently the two researchers discovered a connection that links seven of the 22 khipu found at Puruchuco, an archaeological site just outside of Lima.
“This is the first time that we have seen information actually passing from one khipu to another,” Urton explains. Until now each khipu seemed to record data in isolation, but the seven Puruchuco khipu were meant to be read together with the sums from one recorded on another.
“The big challenge is how do you move from classification and structure to meaning and some kind of interpretation,” Urton explains. “[My] greatest hope will be finding a connection between these structural characteristics of the khipu and information recorded in the Spanish documents.”
There is one group of khipus that Urton and other scholars believe they have deciphered. The largest khipu found by the workmen at Laguna de los Condores in Peru appears to be a calendar. It has 730 strings organized in to 2 groups (365 x 2), which Urton interprets as a two-year calendar.
“When you see strings organized like that you can say with some confidence, ‘I’m looking at a calendar,’” Urton says. “That means that each string can be translated into a day.”
The debate among current-day scholars centers on the question of whether the khipu is a form of writing. “If all they needed to do was record numbers, then you would have a system that was much more economical, such as one of just white cotton strings tied in a decimal hierarchy,” Urton explains, “but there is tremendous variation in terms of color, spinning and plying, and the directions of knots.”
In Signs of the Inka Khipu, published by the University of Texas Press in 2003, Urton suggests that each decision the khipukamayuq makes contributes to the information conveyed in the khipu.
Similar to a computer’s binary code, the making of a khipu involved 7 steps where an either or choice had to be made such as the spin and ply direction of the string (left or right) and the direction of the slant of the knot (like a Z or an S). With the additional choice of 24 variations of string colors, this system of communication could allow the khipukamayuqs to convey some 1,536 units of information (2 6 multiplied by 24), which is nearly double the amount of hieroglyphic signs the ancient Egyptians and Maya created.
If Urton’s theory pans out, the Inca will have created the only known three-dimensional written language, a task so complex that it would take 500 years for the rest of the world to even begin to understand it.
On the other hand, the Aschers present a different hypothesis. They suggest that instead of alphabetic writing like we use, the khipu contains concept writing that relies on numbers to convey information that is non-quantitative.
“Numbers in a certain layout are not magnitudes,” Marcia Ascher says. She explains that numbers can be used as labels. For example, a social security number or a driver’s license number is a label for a person, and if you put together a birth date, a telephone number, a zip code, and a social security number, you are telling the story of someone’s life.
In addition, familiar number patterns can be used to tell a story. Anyone who knows how to read the scoreboard of a baseball game, can tell the story of the game by reviewing the numbers alone. Ascher says that khipu have a multitude of number patterns and some are sums and some are not. The question is what do the patterns stand for? This method of using numbers to convey non-quantitative information is very different than what most societies are used to.
In a 2005 paper in Science magazine, Urton and Brezine suggested that just such an arrangement of knots used as labels is found on several of the Puruchuco khipu. Galen Brokaw comments, “For people like us who have alphabetic mentalities, the way we think about writing is conditioned by the medium that we use. The problem is that our alphabetic medium makes it difficult for us to understand the khipu.”
Although most of the khipu that remain today are from the Inca,William Conklin points out that the khipu are only part of a 2000-year history of Andean textiles. He has conducted detailed analysis of a group of Middle Horizon khipus from a time period 700 years before the Inca.
“The Inca quipu form must have gradually replaced the more complex, colorful, but redundant and repetitive quipus of the Middle Horizon period,” he wrote in the Annals of the New York Academy of Sciences in 1982, “The mathematics of the record-keeping Incas seems thus to have evolved from the art and ritual of earlier times.”
The promise of the khipu is access to information, statistics and historical accounts about the Inca, written by the Inca in their own voice. The beauty and complexity of this unique form of communication awes all who study it.
“One of the reasons we have not yet deciphered khipu is that we have not yet observed and recorded their information nearly carefully enough,” Conklin wrote in Narrative Threads: Accounting and Recounting in Andean Khipu, published by the University of Texas Press in 2002.
“In order to believe that the Andeans could actually read information from the complex khipu graphic fabric, it is necessary to remind ourselves of the amazement the Andeans expressed (at the moment of contact in Cajamarca) upon hearing the Spaniards miraculously verbalize language as they moved a forefinger along a line of our own subtle graphic code in the text of the Bible.”
Published in Americas magazine, September-October 2006, by Chris Hardman