with the purpose of conveying a message to somebody other than himself; b) a person perceives the visible sign and interprets it as a sign produced for the purposes of conveying a message; and c) the person attributes a meaning to the visible sign” (Mignolo, 1994). In this definition of writing or conception of the khipu as a semiotic system, there is no need to necessarily institute the representation of speech.
Lately scholars have made great strides in decoding the khipu system. The question under consideration is whether the khipu was a simply a mnemonic device that offered “cues” to the khipukamayuc, as the Spanish chroniclers claimed, or whether the system can be considered “writing.” New incursions have been characterized by a mathematical approach to the tactile–visual system of cords, knots, colors, and textures. Marcia and Robert Ascher in Code of the Quipu: A Study of Media, Mathematics and Culture (1997) have led the way. Scholars have also been keenly interested in the idea that khipus did not only encode mathematical knowledges, but were also capable of encoding narrative. Gary Urton, in Signs of the Inka Khipu: Binary Coding in the Andean Knotted-String Records (2003), characterizes the khipu as a “powerful system of coding information that was at home in pre-Columbian South America, and which, like the coding system used in present-day computer language, was structured primarily as a binary code” (1). Urton has examined the largest number of archeological and colonial khipu thus far included in any study. His historical and theoretical study leads him to think that the khipu – the system of knotted strings – was used for recording both statistical and narrative information. With this claim, Urton’s understanding of the khipu moves beyond mathematical studies and explores the earlier claims made by the mestizo intellectuals like Garcilaso de Vega, Inca (1539–1616), and Blas Valera (1545–97) regarding the khipu’s capacity to encode and store narrative information. Urton thinks that the khipu were constructed with “conventionalized units of information that could be read by khipu masters throughout the empire” (3). So it seems that the type of information stored in the khipu was at least of two kinds: statistical and narrative. Thus the khipu allowed for accounting and recounting or telling.
Like other modern scholars, Urton draws on the system of conceptualization and organization that is peculiar to the Andes. In an effort to bring to bear the Andean modes of thinking Urton introduces a new analytical idea: binary coding. This enables him to propose a “separation between the recording code and the script, or the ‘readable’ message, in the khipu” (162). He can thus conclude that the binary coding of the khipu “constituted a means of encoding paired elements that were in relationships of binary opposition to each other, and that, at a semantic level, these relations were of a character known in the literature as markedness relations” (162). Urton states that he has “sketched out a theory of interpreting the hierarchical and asymmetrical signs” of non-decimal khipu as the “architecture for canonical literatures [e.g., poetry, historical narrative] whose essential components would have been noted by the khipukamayuc and used as the framework … for constructing narrative recitations” (164).
The guiding idea here is that binary coding was one of the principal mechanisms and strategies for thinking in the Andes. Thus Urton looks for features of cords that apparently mimic Andean logical structures rather than depart from the Indo-Arabic arithmetic as an a priori assumption. Urton privileges binarism because it is widely recognized as the primary category of Andean thought and social organization. He argues that fiber working requires binarism from the very initial stages of spinning to cord and textile making. In this sense, cord-making mimetizes the logical operations that generate Andean order. For Urton the sign that a cord contains is not the cord, but rather the aggregate of binary combinations (left/right, cotton/wool, single/ double, colored/neutral) that construct the cord and function as bits of information. Urton’s theory is not wedded to a mathematical model and as such leaves open the possibility that the khipu cord could encode segments of speech, words, or even syllables. In this way the khipu would be capable of registering “writing” in the usual sense of visible signs that correspond to segments of speech.
One of the most important aspects of Urton’s research is that his method and arguments might finally put to rest the notion originally put forth by José de Acosta (1540–1600) and Bernabé Cobo (1580–1657) in 1653, and repeated throughout the centuries with respect to the khipu. Both argued that the khipus were simply a mnemonic device – not a system – that was used as a memory aid by the khipukamayuc. Thus, the intellectual capacity of the khipu depended entirely on the interpreter’s own abilities. This notion may have been developed in view of the fact that the Inca empire was multilingual, and neither Cobo nor Acosta could imagine how a khipu knotted in one part of the empire could be “read” in another if the languages spoken were not the same. The conception of “writing” as visible speech impeded the cognitive imagination of both scholars. Despite the fact that neither Cobo nor Acosta managed to explain how the khipu was “read” across the many languages spoken in the Inca empire, a problem that would have called for positing the existence of a system rather than simple “cues,” nor how the khipu served as the primordial tool in the governance of a huge and efficient state, their ideas remained unchallenged through the centuries. In fact they served to manufacture and cement the epistemological violence that characterized the colonization of the Amerindian cultures by Europe.
While Signs of the Inka Khipu has been widely regarded as a major breakthrough in Andean studies, Galen Brokaw writes that, despite the fact that Urton presents compelling archeological evidence for the conventionality of the khipu system, he nevertheless does not present enough ethnographic evidence to support the argument about the conventionality of the binary features, nor about the computer-style binary code (Brokaw, 2005: 574). Further, Brokaw argues that Urton “conflates the referential and the poetics” or the structure of cultural interactions (577). For Brokaw it does not follow that “Andean Cultures organize the world into binary categories … a homologous structure characterized the operation of reference itself’ (578). This scholar also finds it hard to imagine how the khipu could support two readings, one numeric and one binary. Brokaw believes that the numeracy direction, as pursued by the Aschers, will eventually result in a better understanding of the khipu than the binary-code model proposed by Urton (2003: 586–7). However, in “The Poetics of Khipu Historiography” (2003), Brokaw also attempts to make the case for the khipu as a system capable of storing narrative information. By comparing two colonial documents that certifiably claim khipus as their immediate source and khipukamayuc as their “readers” in the Quechua oral rendition of the contents, Brokaw is able to establish that there existed a khipu biographical narrative genre (112).
One of the documents Brokaw examines is the Primer nueva corónica y buen gobierno (1615) by Guamán Poma de Ayala (–1516). Contrary to almost all of the interpreters of Guamán Poma who have detected and commented on the European models operating in his work, Brokaw makes the case for a khipu-based historiographic genre as the guiding model in the first part of Guamán Poma’s extensive letter to the king (908 pages). Brokaw goes as far as hypothesizing that “much of the information about indigenous Andean history that appears in the Nueva coronica was collected either directly or indirectly from khipus” (116). In his study of khipu poetics Brokaw concludes that “undeniably the khipu employed a set of highly complex conventions capable of encoding semasiographic or even phonographic information that included highly stable genres of discourse” (141). He thus agrees, if not on the same grounds or with the same methodology, with the claims that Garcilaso de la Vega, Inca, in his Comentarios reales of 1606, made for the khipu. This revalidation of Garcilaso as a reliable informant is important because the ethnohistorian María Rostworowski (1983) found some of her findings at variance with the Incas, and concluded from there that Garcilaso’s work was not to be trusted, especially when it came to cognitive and narrative claims for the khipu.
Frank Salomon, in his Cord Keepers (2005), tackles anew the question of the colonial and the ethnographic khipus. His book is the most comprehensive study of both archeological and ethnographic khipu to date. Salomon points out that the archeological or pre-Hispanic khipus that have thus far been examined with radiocarbon dating show that by 600 CE Andean peoples were making highly complex khipus (11). Thus the art of khipu-making is not only Pan-Andean, but also indicates a deeply rooted continuous use and development of an art that precedes the