Scholars have long maintained that Inca civilization was unique because of its lack of a writing system. That dogma is now being challenged in a groundbreaking volume edited by Jeffrey Quilter and Gary Urton.

The issue under consideration is the degree of narrativity of the khipu, an object made of dyed strings knotted and hung from a horizontal cord. The khipu is well known as the device used by Inca accountants (khipucamayoq) for numerically recording human, animal, plant, and other resources. New research strongly suggests that some khipu facilitated the oral performance of dynastic history, poetry, speeches, and songs by their owners and makers.

But was this writing? Were khipu read (did they display a grammatical ordering of words able to be orally reproduced by any literate person) or were they consulted? Gary Urton explains khipu as “general signifiers evoking classes of objects, actions, places and times … [that] were given more nuanced form and substance by a khipucamayoq who would have brought to the [performance] information retained in his memory, as well as a range of creative, discursive practices for producing a narrative appropriate to a given place and perhaps audience” (p. 20). No unanimous conclusion is reached in this volume, but all contributors concur that information was encoded on khipu by means of patterned variation in the material used (the color of strings; their dyeing, spinning, and plying; the type and number of knots).

Gary Urton, Tristan Platt, and Rosaleen Howard each consider the relationship between language and khipu. Urton emphasizes that the syntactic, semantic, and overall grammatical properties of Quechua and Aymara are a necessary starting point for studying narrative khipu because “these were the languages that gave coherence and structure to the thinking, logic, and communication among local, regional, and imperial officials … who constructed these devices in the first place” (p. 192). Platt calls attention to regional variation in the cultural ordering of ethnocategories across the Andes and therefore in widely dispersed khipu. Howard suggests that the organizational principles of Quechua narrative performance and the structure of the Quechua language are compatible with creating and reading a narrative khipu.

Various authors interrogate the relationship between khipu and memory. Carlos Sempat Assadourian suggests that the Inca victor of a civil war killed as many khipucamayoq of the defeated claimant to the throne as could be found and had all their khipu burned. Khipu facilitated lawsuits of Andean communities against their Spanish colonial overlords (Tristan Platt). Under forced religious conversion Andean people recontextualized their khipu as pseudo-rosaries (Regina Harrison). Eventually, the colonial regime ordered all khipu destroyed. As Jeffrey Quilter trenchantly observes, “The control of information and information makers was the key to power” (p. 212).

William Conklin proposes an “information string theory” by which sequential narration could be accomplished in khipu construction. Maria Ascher demonstrates that the concept of numbers as labels is fundamental in decoding khipu; the formatting of numerical information provides the narrative frame for the khipu’s narrative content. Jeffrey Quilter makes a brilliant comparison between the telegraph, stenography, and khipu. He argues that "without any means of rapidly recording language as it was spoken or sung, the Inka would have had to rely on memory in order to later record the [spoken] texts as khipu" (p. 208). Robert Ascher claims that Inca khipu fit into “the concept of a writing group” (p. 105). He also calls attention to the simultaneous tactile, visual, vocal, and auditory acts involved in khipu performance.

Tristan Platt and Frank Salomon deal with colonial and contemporary khipu use in Bolivia and Peru, respectively. Platt exhorts scholars to contextualize khipu in “the living social context in which they were created and transformed as changing social objects” (p. 230). Salomon emphasizes khipu as accumulative histories corresponding to multiple makers and events, thereby containing components of differing date; he emphasizes “use-life.” Salomon interprets the display of “patrimonial” khipu (khipu that can no longer be read) in the community he studied as “currently functioning documents of civil legitimacy” (p. 295) and the “records of the minimal corporate constituents of a society” that “symbolize local autochthony,” unlike the imperial contexts of archaeological khipu (p. 299).

On the basis of ethnographic comparison Carol Mackey suggests that the khipu found by archaeologists in ancient tombs might have been copies of originals that remained in use.

Narrative Threads is one of the most intellectually exciting books I have read in recent years. It should be mandatory reading for archaeologists and ethnohistorians studying the Andes. It has keen significance for the readers of this journal interested in history and memory, discourse and knowledge, colonial resistance, cultural production, materiality, and performance.