Quipu is the Spanish form of the Inca (Quechua language) word khipu (also spelled quipo), a unique form of ancient communication and information storage used by the Inca Empire, their competition and their predecessors in South America. Scholars believe that quipus record information in the same way as a cuneiform tablet or a painted symbol on papyrus do. But rather than using painted or impressed symbols to convey a message, the ideas in quipus are expressed by colors and knot patterns, cord twist directions and directionality, in cotton and wool threads.
The first western report of quipus was from the Spanish conquistadors including Francisco Pizarro and the clerics who attended him. According to Spanish records, quipus were kept and maintained by specialists (called quipucamayocs or khipukamayuq), and shamans who trained for years to master the intricacies of the multi-layered codes. This was not a technology shared by everyone in the Inca community. According to 16th-century historians such as Inca Garcilaso de la Vega, quipus were carried throughout the empire by relay riders, called chasquis, who brought the coded information along the Inca road system, keeping the Inca rulers up to date with the news around their far-flung empire.
The Spanish destroyed thousands of quipus in the 16th century. An estimated 600 remain today, stored in museums, found in recent excavations, or preserved in local Andean communities.
Although the process of deciphering the quipu system is still just beginning, scholars surmise (at least) that information is stored in cord color, cord length, knot type, knot location, and cord twist direction. Quipu cords are often plaited in combined colors like a barber pole; cords sometimes have single threads of distinctively dyed cotton or wool woven in. Cords are connected mostly from a single horizontal strand, but on some elaborate examples, multiple subsidiary cords lead off from the horizontal base in vertical or oblique directions.
What information is stored in a quipu? Based on historical reports, they were certainly used for administrative tracking of tributes and records of the production levels of farmers and artisans throughout the Inca empire. Some quipu may have represented maps of the pilgrimage road network known as the ceque system and/or they may have been mnemonic devices to help oral historians remember ancient legends or the genealogical relationships so important to Inca society.
American anthropologist Frank Salomon has noted that the physicality of quipus seems to suggest that the medium was exceptionally strong in encoding discrete categories, hierarchy, numbers, and grouping. Whether quipus have narratives embedded in them as well, the likelihood that we'll ever be able to translate story-telling quipus is very small.
Evidence for the Quipu Use
Archaeological evidence indicates that quipus have been in use in South America at least since ~AD 770, and they continue to be used by Andean pastoralists today. The following is a brief description of evidence supporting quipu use throughout Andean history.
- Caral-Supe culture (possible, ca 2500 BC). The oldest possible quipu comes from the Caral-Supe civilization, a preceramic (Archaic) culture in South America made up of at least 18 villages and enormous pyramidal architecture. In 2005, researchers reported a collection of strings twisted around small sticks from a context dated to approximately 4,000-4,500 years ago. Further information has not been published to date, and the interpretation of this as a quipu is somewhat controversial.
- Middle Horizon Wari (AD 600-1000). The strongest evidence for the pre-Inca use of quipu record keeping is from the Middle Horizon Wari (or Huari) empire, an early urban and perhaps state level Andean society centered at the capital city of Huari, Peru. The competing and contemporary Tiwanaku state also had a cord device called a chino, but little information is available about its technology or characteristics to date.
- Late Horizon Inca (1450-1532). The best-known and largest number of surviving quipus are dated to the Inca period (1450-Spanish conquest in 1532). These are known both from the archaeological record and from historical reports-hundreds are in museums around the world, with data on 450 of them residing in the Khipu Database Project at Harvard University.
Quipu Usage After the Spanish Arrival
At first, the Spanish encouraged the use of quipu for various colonial enterprises, from recording the amount of collected tribute to keeping track of sins in the confessional. The converted Inca peasant was supposed to bring a quipu to the priest to confess his sins and read those sins during that confession. That stopped when the priests realized that most of the people couldn't actually use a quipu in that manner: the converts had to return to the quipu specialists to obtain a quipu and a list of sins that corresponded to the knots. After that, the Spanish worked to suppress the use of the quipu.
After the suppression, much Inca information was stored in written versions of the Quechua and Spanish languages, but quipu use continued in local, intracommunity records. The historian Garcilaso de la Vega based his reports of the downfall of the last Inca king Atahualpa on both quipu and Spanish sources. It might have been at the same time that quipu technology began to spread outside of the quipucamayocs and Inca rulers: some Andean herders today still use quipu to keep track of their llama and alpaca herds. Salomon also found that in some provinces, local governments use historical quipu as patrimonial symbols of their past, although they do not claim competence in reading them.
Administrative Uses: Santa River Valley Census
Archaeologists Michael Medrano and Gary Urton compared six quipus said to have been recovered from a burial in the Santa River Valley of coastal Peru, to data from a Spanish colonial administrative census conducted in 1670. Medrano and Urton found striking pattern similarities between the quipu and census, leading them to argue that they hold some of the same data.
The Spanish census reported information about the Recuay Indians who lived in several settlements near what is today the town of San Pedro de Corongo. The census was split into administrative units (pachacas) which usually coincided with Incan clan group or ayllu. The census lists 132 people by name, each of whom paid taxes to the colonial government. At the end of the census, a statement said the tribute assessment was to be read out to the natives and entered into a quipu.
The six quipus were in the collection of the Peruvian-Italian quipu scholar Carlos Radicati de Primeglio at the time of his death in 1990. Together the six quipus contain a total of 133 six-cord color-coded groups. Medrano and Urton suggest that each cord group represents a person on the census, containing information about each individual.
What the Quipu Say
The Santa River cord groups are patterned, by color banding, knot direction, and ply: and Medrano and Urton believe that it is possible that the name, moiety affiliation, ayllu, and amount of tax owed or paid by an individual taxpayer could well be stored among those different cord characteristics. They believe they have so far identified the way the moiety is coded into the cord group, as well as the amount of tribute paid or owed by each individual. Not every individual paid the same tribute. And they have identified possible ways that proper names might have been recorded as well.
The implications of the research are that Medrano and Urban have identified evidence supporting the contention that quipu store a great deal of information about the rural Inca societies, including not just the amount of tribute paid, but family connections, social status, and language.
Inca Quipu Characteristics
Quipus made during the Inca Empire are decorated in at least 52 different colors, either as a single solid color, twisted into two-color "barber poles", or as an unpatterned mottled group of colors. They have three kinds of knots, a single/overhand knot, a long knot of multiple twists of the overhand style, and an elaborate figure-of-eight knot.
The knots are tied in tiered clusters, which have been identified as recording the numbers of objects in a base-10 system. German archaeologist Max Uhle interviewed a shepherd in 1894, who told him that the figure-of-eight knots on his quipu stood for 100 animals, the long knots were 10s and single overhand knots represented a single animal.
Inca quipus were made from strings of spun and plied threads of cotton or camelid (alpaca and llama) wool fibers. They were typically arranged in only one organized form: primary cord and pendant. The surviving single primary cords are of widely variable length but are typically about a half centimeter (about two-tenths of an inch) in diameter. The number of pendant cords varies between two and 1,500: the average in the Harvard database is 84. In about 25 percent of the quipus, the pendant cords have subsidiary pendant cords. One sample from Chile contained six levels.
Some quipus were recently found in an Inca-period archaeological site right next to plant remains of chili peppers, black beans, and peanuts (Urton and Chu 2015). Examining the quipus, Urton and Chu think they have discovered a recurring pattern of a number-15-that may represent the amount of tax due to the empire on each of these foodstuffs. This is the first time that archaeology has been able to explicitly connect quipus to accounting practices.
Wari Quipu Characteristics
American archaeologist Gary Urton (2014) collected data on 17 quipus which date to the Wari period, several of which have been radiocarbon-dated. The oldest so far is dated to cal AD 777-981, from a collection stored in the American Museum of Natural History.
Wari quipus are made of cords of white cotton, which were then wrapped with elaborately dyed threads made from the wool of camelids (alpaca and llama). Knot styles found incorporated in the cords are simple overhand knots, and they are predominantly plied in a Z-twist fashion.
The Wari quipus are organized in two main formats: primary cord and pendant, and loop and branch. The primary cord of a quipu is a long horizontal cord, from which hangs a number of thinner cords. Some of those descending cords also have pendants, called subsidiary cords. The loop and branch type has an elliptical loop for a primary cord; pendant cords descend from it in series of loops and branches. Researcher Urton believes that the main organizational counting system may have been base 5 (that of the Inca quipus has been determined to be base 10) or the Wari may not have used such a representation.
- Hyland, Sabine. "Ply, Markedness, and Redundancy: New Evidence for How Andean Quipus Encoded Information." American Anthropologist 116.3 (2014): 643-48. Print.
- Kenney, Amanda. "Encoding Authority: Navigating the Uses of Khipu in Colonial Peru." Traversea 3 (2013). Print.
- Medrano, Manuel, and Gary Urton. "Toward the Decipherment of a Set of Mid-Colonial Khipus from the Santa Valley, Coastal Peru." Ethnohistory 65.1 (2018): 1-23. Print.
- Pilgaonkar, Sneha. "The Khipu-Based Numeration System." ArcXiv arXiv:1405.6093 (2014). Print.
- Saez-Rodríguez, Alberto. "An Ethnomathematics Exercise for Analyzing a Khipu Sample from Pachacamac (Perú)." Revista Latinoamericana de Ethnomatemática 5.1 (2012): 62-88. Print.
- Salomon, Frank. "The Twisting Paths of Recall: Khipu (Andean Cord Notation) as Artifact." Writing as Material Practice: Substance, Surface and Medium. Eds. Piquette, Kathryn E. and Ruth D. Whitehouse. London: Ubiquity Press, 2013. 15-44. Print.
- Tun, Molly, and Miguel Angel Diaz Sotelo. "Recovering Andean Historical Memory and Mathematics." Revista Latinoamericana de Etnomatemática 8.1 (2015): 67-86. Print.
- Urton, Gary. "From Middle Horizon Cord-Keeping to the Rise of Inka Khipus in the Central Andes." Antiquity 88.339 (2014): 205-21. Print.
- Urton, Gary, and Alejandro Chu. "Accounting in the King's Storehouse: The Inkawasi Khipu Archive." Latin American Antiquity 26.4 (2015): 512-29. Print.