Вы находитесь на странице: 1из 345



Textual Struggles, Education, and Land in the Andes

Denise Y. Arnold
With Juan de Dios Yapita

University of Pittsburgh Press

Published by the University of Pittsburgh Press, Pittsburgh, Pa., 15260
Copyright © 2006, University of Pittsburgh Press
All rights reserved
Manufactured in the United States of America
Printed on acid-free paper
10 9 8 7 6 5 4 3 2 1

library of congress cataloging-in-publication data

Arnold, Denise Y.
[Rincón de las cabezas. English]
The metamorphosis of heads : textual struggles, education, and land in the Andes /
Denise Y. Arnold with Juan de Dios Yapita.
p. cm. — (Illuminations)
Translation of: El rincón de las cabezas.
Includes bibliographical references and index.
ISBN 0-8229-4280-1 (cloth : alk. paper)
1. Indians of South America­—Education—Bolivia.  2. Aymara Indians—Education. 
3. Indians of South America—Andes Region—Languages—Writing.  4. Quipu.  5. Inca
textile fabrics.  6. Indian literature—Criticism, Textual.  7. Transmission of texts—
Andes Region.  8. Literacy—Bolivia.  9. Peru—History­Conquest  10. Spain—
Colonies—America—Administration.  I. Yapita, Juan de Dios.  II. Title.  III. Illumina-
tions (Pittsburgh, Pa.)
F3320.1.E4A7513 2006
302.2'24408998324—dc22 2005031096
For Elvira Espejo and other intellectuals of the Andean highlands,
and to the memory of Anne Paul
Our ancestors never, never ever knew how to read and write,
they never ever learned, did they?
For our ancestors then, the book was the natural world,
everything that really existed in nature, our ancestors read
in accordance with the season, with the time of year .....

Professor Domingo Choque, Academy of Aymara

Language and Culture, Pusisuyu
Con t e n t s

List of Illustrations, ix
Acknowledgments, xi

Introduction, 1
pa rt 1 . t e xt ua l s t ru g g l e s
1. Andean Textual Polity, 19
2. Colonizing Texts and the Struggle over Meanings, 38

pa rt 2 . t h e r o s t ru m o f h e a d s
3. Land, Seeds, and Letters:
The Cycles of Production and Reproduction, 69
4. Cycles of Metamorphosis: The Children as Enemies, 87
5. Warriors and Weavers: The Pathways of Learning
in the Community, 110
6. The Cycles of Libations in School Rituals, 133
pa rt 3 . a n d e a n t e xt s a n d t h e i r i n t e r p r etat i o n
7. Cycles of Memory: The Inka’s Voice, 161
8. Cycles of Sound: Prayers and the “Rain of Letters,” 183
9. The Corporeality of Kipus: Toward a Mathematics Incarnate, 207
10. Kipu, Number, and Writing, 225
11. Textual Logic in the Andes, 244
12. Toward an Andean Textual Theory, 273

Notes, 291
Bibliography, 303
Index, 323

I l lu s t r ations

Fig. 1. Map of communities mentioned in the study 13

Fig. 2. The Coya’s belt in the American Museum 22
Fig. 3. Catechisms in the form of boustrophedon and spiral 48
Fig. 4. Drawings from Jakhüwi 4 (p. 29) 59
Fig. 5. The cover of Jakhüwi 1 with people of paper and ink 60
Fig. 6. Drawings from Aru 1 (p. 26). A child on the floor, while a
group of children traces his body with pencil on paper 61
Fig. 7. Photo of children sacrificing a sheep in a multimedia presentation (2004) 64
Fig. 8. Photo of the parade which began in Uma Jalsu (Where the
Water Comes Out), Livichuco, 5 August 1998 76
Fig. 9. Photo of the parade which began in Uma Jalanta (Where the
Water Goes In), Livichuco, 6 August 1998 76
Fig. 10. Photo of the march in front of the school rostrum,
Livichuco, 6 August 1998 77
Fig. 11. The model of communal education developed by the comunarios 90
Fig. 12. Photo of the tongue and throne motif 101
Fig. 13. The ideal couple of weaver and warrior 127
Fig. 14. The school ritual sites in Livichuco 140
Fig. 15. Photo of the anxata ritual in the hamlet of Livichuco 140
Fig. 16. Photo of a child butting his head against a ram 141
Fig. 17. Inka usnus 149
Fig. 18. A Moche usnu 149
Fig. 19. Photo of schoolboys with their knitted caps hanging
down behind, like trophy heads 157
Fig. 20. Photo of a typical reading corner 162
Fig. 21. The hand and kipu counting 214


Fig. 22. Left- and right-directed operations 215

Fig. 23. Quechua and Aymara numbering compared 216
Fig. 24. An Aymara chinu with its primary and secondary cords 218
Fig. 25. The movement of the libation knots toward the pen knot 222
Fig. 26. Female and male knots, according to Don Domingo Jiménez 223
Fig. 27. Libation sequence on threads and knots 238
Fig. 28. An Inka kipu with a wooden handle 241
Fig. 29. The pulu and k'illpha borders of a woman’s weavings 247
Fig. 30. Photo of the pulu border tied by the holding thread
(ch'ukurkata) to the loom pole 248
Fig. 31. Detail of a pendant kipu thread with the larger quantities above 259
Fig. 32. Kipu with baby knots and the flow of sexual substances 261
Fig. 33. Direction and borders in the textiles of men and women 268
Ac k n ow l e dgments

This book would not have been written without the help of many persons
and institutions. We offer thanks to PIEB (Proyecto de Investigación Estratégica
en Bolivia), especially Godofredo Sandoval, and the General Directive for Inter-
national Cooperation of the Dutch Foreign Relations Ministry (DC-OZ/DGIS)
for the opportunity to carry out the Multimedia research project during 1998,
whose published result, El rincón de las cabezas: luchas textuales, educación y tie-
rras en los Andes (2000), forms the basis for this book. Thanks also to the junior re-
searchers in that project, U. Ricardo López G., Luisa Alvarado Cruz, and Nelson
D. Pimentel H., for help with chapters 5, 6, and 10–11, respectively.
In Livichuco (ayllu Qaqachaka, Avaroa Province, Oruro Department), our
thanks go to the authorities and ayllu members for their hospitality during our
visits in 1997–1999, in particular the corregidor Don Juan Wayllani Maraza, the
school mayor Don Teodoro Maraza Choque, and the ayllu leader (jilanqu) Don
Eriberto Maraza. Thanks also to the members of the APSU (Artesanía Para
Seguir Unidos) project and other participants in the Livichuco workshop, espe-
cially Don Juan Maraza, Don Feliciano Maraza, Don Santiago Inka Maraza, Don
Donato Inka Maraza, Doña Isabel Mamani, Doña Sabina Mamani, Doña Antonia
Espejo, and the late Doña Agustina Mamani. We also thank Mary Ostergren of
the U.S. Peace Corps for her hospitality in Challapata, and likewise her succes-
sors, Laura Pusateri and Alicia Díaz.
In the Branch School of Livichuco, we thank the teachers Samuel Gisbert,
Florencia de Gisbert, and María Cayo, for sharing their experiences and hopes
concerning the current educational reform and for permitting us to participate
in their civic commemorations during 1998. Among the schoolchildren, we
thank especially William Espejo and Arsenio Wayllas Colque, and Elías Chuki-
cha Wallpa, Eloy Wallpa Maraza, and Roberto Choque Wallpa, for sharing their
repertory of tales and prayers. In the central school of Qaqachaka, we thank the
teacher Leocadio Choque Poma and the former headmaster Isaac Wayllani for
clarifying ideas concerning Qaqachaka educational history and present-day edu-


cational practices. We also thank Amado Cahuana, director of the Litoral Col-
lege, San Miguel (Carangas Prov., Oruro Dept.), for his valuable comments, and
Trifón Choque, director of CEDPAN (Centro de Estudios y Desarrollo de los
Pueblos Andinos), for providing us with reports on their projects over the last
In Qaqachaka pueblo, we thank Elvira Espejo Ayka, her mother Doña Nico-
lasa Ayka, and aunt Doña Lucía Quispe Choque, for explaining the philosophy
behind various ideas concerning Andean education and child development. El-
vira also helped during the preparation of this book with many suggestions and
clarifications. While this book developed as a work concerned with reading and
writing, in the same room Elvira’s weavings grew to express a more Andean
reality—an inkuña cloth for her coca leaves, from whose multiple levels emerged
several tigers and a lion; three headbands (in Paracas style), different belts, and in-
numerable tullmas for her hair braids—in short a whole world for comparison.
In the valley community of Mitma, Aymaya (Rafaél Bustillos Prov., Potosí
Dept.), we thank Don Domingo Jiménez Aruquipa, his wife Doña Teresa Kuchu,
his son Don Julían, and daughter-in-law Doña Jacoba Tijlla for their hospitality
and the many conversations they held with us in visits to La Paz.
We thank Filomena Nina and Juan Carvajal of the National Technical and
Pedagogical Service Unit (UNST-P) in the Bolivian Education Ministry for com-
ments on the preparation of the educational reform modules, and Carmen de
Urioste B., director of the UNST-P, for inviting us to the “Workshop on achieve-
ments and difficulties in the process of normalization of the Aymara and Que-
chua native languages, as proposed by the Educational Reform.” Thanks also to
Carlos Coello and Clemente Mamani for commenting on the educational reform
modules in Spanish and Aymara, respectively.
We thank the Literary History Project based at the University of Toronto,
Canada, especially Elizabeth Monasterios at SUNY, New York, for giving us the
opportunity to study the “Institutional modes of production and dissemination
of texts in the Andes.”
The following colleagues shared invaluable bibliographic information and
commentaries on work in progress: Xavier Albó, Margot Beyersdorff, Gordon
Brotherston, Ricardo Cavalcanti, Claudette Kemper Columbus, Lindsey Crick-
may, Bartholomew Dean, Sabine Dedenbach-Salazar, Penny Dransart, Teresa
Durán, Carlos Fausto, Jo Murphy Lawless, Robert Leavitt and his wife Miriam,
Aurolyn Luykx, Guillermo Mariaca, Jenny Palmer, Anne Paul, Silvia Rivera, Ina
Rösing, William Rowe, Frank Salomon, Brian Street, Aparecida Vilaça, and Mar-
celo Villena. We give special thanks to Stephen Nugent for commenting on the
final manuscript, and to Stephen Scott for his painstaking editorial skills.

At the University of Pittsburgh Press, we thank Nathan McBrien, John Bev-

erley, and Sara Castro-Klarén for their editorial support, and Gary Urton, Galen
Brokaw, and an anonymous reader for their comments.
In the Institute of Aymara Language and Culture (ILCA) in La Paz, we thank
Margarita Tito, Emiliana Ylaya, Rosario Rocha, Lorena Peñaloza Bretel, and the
late Carlos Urquizo Sossa for their transcriptions; and Norberto Copana, Nilton
Callejas, Edson Quispe, Marcia Rodríguez, and Roberto Espejo for help with the
drawings and layout. Finally, special thanks to Dany Mena for help with endless
chores, and to Ian Marr for improving the English translation and for his host of
bibliographic searches.
All photographs are by the authors, unless otherwise indicated, as are all

Words—affirms Octavio Paz—may be used to designate a reality or to disguise it.

Karen Claure, Las escuelas indigenales

This book examines how the centuries-long struggle for sovereignty in the
Andes is played out in conflicting ideas over the nature of writing. It traces at
one extreme the colonial confrontation over the expression of divine power, per-
sonified in the fateful encounter at Cajamarca in 1532 between the Inka leader
Atawallpa, whose textual polity, or textual basis of government, was founded in
cloth, and representatives of the Spanish Church, whose textual authority was
inscribed in the Holy Scriptures. At the other extreme is the everyday reality of
modern neoliberal politics in Bolivia, where the current educational reform pro-
gram’s more secular intent to replace traditional reading and writing practices—
in the hands of weavers, local title bearers, scribes, and rural teachers—with al-
phabetic writing is driven by the demands of globalization.
The terrain between these two extremes introduces the reader, perhaps unfa-
miliar with the Andes and its woven basis of government, to a portrayal of how
human life and notions of personhood become transformed in the passage from
woven to written systems of communication, and thence into a modern infor-
mation economy. This portrayal reveals how the Andean populations enlist weav-
ing knowledge as an unexpectedly subversive device in their resistance to current
global changes, such as educational reform, and why officialdom continues to be
deaf to their demands.
In this play of globalization and resistance, we trace how the neoliberal re-
forms of the 1980s and 1990s (more flexible labor markets, state cutbacks in
health and education, land reforms) tend to be accompanied by “multicultural”


policies, especially in regions with large indigenous populations. Under the ap-
parently innocent guise of bilingual intercultural education, neoliberal policies
are able to impose state hegemony by dividing a common cultural terrain into
archipelagos of competing linguistic and cultural groups, deprived of the sover-
eignty they enjoyed before the Spanish Conquest. This, in turn, undermines the
emerging democratic demands of majority indigenous populations in countries
such as Bolivia, where 62 percent of the overall population identify themselves
as indigenous, according to the latest census (2001). In practice, the familiar colo-
nizing principle of divide and rule is driven by the political control of writing.
Describing the history and methods of the repeated attempts in the Andes to
control regional writing practices engages us in both practical issues of literacy
education and wider philosophical questions of cultural self-determination.
Like Brinkley Messick in The Calligraphic State (1993), we examine textual
polity as a political entity and a discursive tradition. Drawing on Weber, Messick
coined the terms “textual authority” and “textual domination” to describe some
of the social and political processes involved in articulating, through relations of
power, the authority of certain forms of writing and kinds of texts over others. In
these processes, textual domination intersects with other dimensions of author-
ity and the relations of specific modes of production. Our overall task is to under-
stand the interlocking of textual polity, social order, and the forms of discourse in
an Andean context, and its cultural consequences.
In examining the initial confrontation over writing during the colonial period,
we reconstruct certain aspects of the traditional Inka textual culture, where knot-
ted kipus and exquisite weavings held sway, and trace its destruction with the
Spanish Conquest. In the present-day remodeling of scriptural practices, we ex-
plore first what alphabetic reading and writing has meant historically for Andean
rural populations, and then how this history shapes their present attitudes toward
the current educational reform. In between, we show how the centuries of con-
flict between different reading and writing practices, as well as the gradual emer-
gence of new hybrid forms (the result of the colonial contact between different
textual practices), are shaped by the reactions of Andean populations and their
representatives to the affairs of state that affect them.
Historically, the scope of this textual struggle pitted entire states and their
representatives against one another in matters of administration and fiscal mea-
sures; jurisprudence and jurisdiction; political, territorial, and linguistic divisions;
and ceremonial and ritual organization. Even now, the ongoing phenomenon of
interpenetrating sovereignties still erupts into violent disputes, such as that of
October 2003, when the rebellion of the Aymara populations of La Paz that over-
threw President Sánchez de Lozada began in the teacher training college of Wari-

sata. While such oppositions have been commonly viewed in terms of rural An-
dean communities versus the state, the arguments presented in this book reveal
other ways in which long-standing disputes between sovereignties color regional
variants on a common theme, namely the question of land—the Indian question
par excellence.
One of the most significant institutional sites where this textual struggle has
been played out is the school. In Bolivia, a multinational and multilingual coun-
try, the indigenous nations, languages, and writing practices have suffered fierce
repression in the classroom for almost two centuries. Many studies have empha-
sized the hegemonic nature of schooling, with its multiple forms of oppression
of indigenous peoples, a masculinist stance toward gender relations, and so on.
Fewer have taken into account the role of the school as an everyday site for con-
testing power and hegemony. In a historical sense, schools—where we can trace
the play of power over centuries—are sites where social memory and alternative
models of power relations are worked out.
In this sense, there is much to be gained from an ethnographic reading of
schooling from a rural community’s point of view. The educational arguments
they present here respond to the stark silencing in the current Bolivian educa-
tional reform proposals of the voices of Andean populations (community mem-
bers, children, rural teachers), and the whole gamut of regional schooling and
reading and writing practices.
From the historical perspective developed in the early chapters, we argue that
present educational reform proposals, based in the attempt to wipe out illiteracy,
ignore the very textual practices that have much to contribute to this process in
a positive and constructive manner. Instead, notions of schooling, writing, liter-
acy, mathematics, text, meaning, interpretation, and comprehension founded in
European-criollo thought (and influenced by multicultural trends from outside),
are imposed, while the historical roots of illiteracy are ignored.
We identify a further problem in the discourse of these outside trends con-
cerning bilingual intercultural education. While recognizing the propositional
nature of “interculturality” in theory, in practice this position simply subjects di-
verse indigenous nations to the hegemony of universalism. We argue that the
liberal posture of the reform hides yet another colonizing intent to undermine
Andean languages and cultures, this time by molding them through alphabetical
writing in a wider pedagogical project that seeks to impose the values of a knowl-
edge presumed to be universal concerning being and knowing, corporeality, gen-
der and sexuality, and intergenerational transmission.

Methods and Sources

In order to express the multivocal nature of these ongoing struggles for sover-
eignty, we adopt a two-tiered approach regarding sources and methodologies: the
first from external debates, and the second from within the region of study.
First, we seek to identify textual concerns of broader comparative relevance,
drawing on recent studies focusing on the Americas, for instance the work of
Nancy Farriss, Serge Gruzinski, Gordon Brotherston, and T. A. Abercrombie,
as well as those from farther afield, for example, the work of Clifford Geertz in
Indonesia, Brinkley Messick in Yemen, and Benedict Anderson’s groundbreak-
ing study of the textual formulation of “imagined communities” in nineteenth-
century nation building.
Farriss’s Maya Society under Colonial Rule (1984) and Gruzinski’s Conquest of
Mexico (1993) adopt a dialectical perspective that poses the point of view of In-
dian communities as a whole against Spanish hegemony. These broad sweeps are
one of the riches of historical studies of this kind but also show their limitations.
For his part, Brotherston, in the Book of the Fourth World (1992), adopts a model of
historical continuity, whether in Abya Yala (Latin America) or Turtle Island (the
United States). In the Andean context he simply speaks of Collasuyu and its on-
going literary tradition (into modern times), viewing the Spanish Conquest as a
mere interruption in a much longer historical durée. In relation to these, Messick’s
work on Yemen is a closely argued ethnographic comparison between the tradi-
tional Islamic writing practices of the provinces and the modernizing demands
of a central bureaucratic state. He, too, focuses on schooling as an institution me-
diating the textual transformations taking place.
More recently, Abercrombie’s Pathways of Memory and Power (1998) develops
a postcolonial model of Andean historical ethnography that describes “hybrid”
textual reformulations resulting from transculturation. His model draws on the
linguistic process of pidginization when languages come into contact to describe
the “interculture of colonial borderlands,” and he rejects what he views as the
nostalgia-driven anthropological romance that fails to take sufficiently into ac-
count this new postcolonial space (1998, 24). But, as Abercrombie is aware, the
pidgin’s existence limits the understanding between the parties, just as it limits
our own understanding of how local polities might reconstruct their own iden-
tities “as opposed to” hegemony. In this sense, the rich redefinition of mestizaje
he develops plays into the hands of only one group of social actors, while ignor-
ing what Stuart Hall (1996) has called the “strategic essentialism” of historical or
modern identity politics currently being played out in the new social movements
on the periphery.

Our task is to avoid recreating essentialist models of lo andino by developing

a theoretical model complex enough to account for the cultural, textual, and his-
torical variability across the region and its implications for long-term relations of
domination and resistance. Taking into account Foucault’s mapping of layered
textual interpretations and Bourdieu’s “fields of cultural production,” we devel-
oped a model of textual practices in what Galen Brokaw calls “textual contact
zones” (2003, 113), parallel to the linguistic model of languages in contact.1 Our
formulation of this textual model helps us understand, in turn, the limitations of
Abercrombie’s pidgin approach.
Studies of languages in contact propose that contact results in changes, or
“interference,” in the languages’ phonological, syntactic, and semantic systems.
This linguistic model was criticized some time ago for its cavalier attitude toward
the complex dynamic processes between languages, and the term “interference”
was replaced by the more neutral term “transference.” There were also debates
concerning the degree of influence of the mother tongue in the learning of a
second language, as compared to other grammatical influences at a more uni-
versal level. In the case of Spanish acquired as a second language by speakers of
an Andean maternal tongue, the results have been called “interlects,” or “semi-
languages” (in the case of the Andean Spanish spoken by rural teachers), although
all of these intermediary and hybrid variants have been shown to be a continuum
or a series of “interlanguages” between the mother tongue (L1) and the language
being learned, or destination language (L2).2
Taking these debates into account, our contrastive study of textual practices
in contact compares the characteristics of transference at different levels of inter-
action. Here, we find an analogous range of influences, for example of the ma-
ternal practice of weaving (TP1) on the acquisition of a second textual practice
(TP2), alphabetic writing. We explore ways in which the more universal concepts
of writing, voice, and text have realizations in both weaving and braiding (as TP1)
and alphabetical writing (as TP2). We argue that such comparisons are necessary
to understand the historical encounters between distinct yet related practices, for
instance the specific regional reading and writing practices of the Inka state (and
other Andean polities) at the moment of the Spanish Conquest, as well as the Eu-
ropean practices of the time, while not overlooking all the variants on these that
have occurred as a result of contact.
Within these preliminary theoretical concerns, we draw on the work of the
French philosopher Jacques Derrida, in particular his reformulation of writing
in Of Grammatology (1976), in our search to redefine operative terms and catego-
ries. This enables us to broaden the definition of writing to embrace both Eu-

ropean and Andean writing practices. From the perspective of “grammatology”

(the study of writing systems), we pose the following questions: To what extent
might we call textual practices based in cloth or knotted threads “writing”? What
aspects of Andean weaving (or knotting) activities might be considered as equiva-
lent to writing?
We do not intend to reduce any levels of signification to writing (or even met-
aphors of writing) because of the danger of erasing context-dependent commu-
nicative functions. Rather we take into account the etymology of the term “text”
explored by Walter D. Mignolo (1994a), in the sense of “making something,”
more specifically “weaving it.” We also respect the dynamic sense of this term as
“gathering together,” “interweaving or interlacing” things, for example voice and
writing. “Text,” in this sense, evokes the idea of something woven that forms part
of a web.3 The term “textual practice,” from Leinhard (1992b), describes the set
of secondary practices (literature, song, music, dance) that derive from the han-
dling of textiles in the Andes and that are based on cognitive, logical, conceptual,
and classificatory schemes homologous to those of Andean weavings.
Despite the fact that Derrida’s study of writing systems derives from Euro-
pean metaphysical preoccupations, it casts light on many philosophical dimen-
sions of Andean textual practices. He and his contemporaries found in writing
the origins of the notion of being, and, beyond this, of European preconceptions
concerning identity (Self ) and its relationship with the Other. Looking for the
“trace of being,” Derrida found the “textuality of being”; he found in writing
the textual bases whence the notions of Self and the Other derive. Above all in
Of Grammatology, he discovered in the origins of writing something hidden from
Western history, a submerged violence that underlies fundamental questions of
subjectivity, identity, and the Other.
For Derrida, this submerged violence derives from the “inert” nature of writ-
ing that can only be activated by the voice. He argues that the foundation of writ-
ing in vocal practices (what he calls the “reign of the Word”) is a key but un-
perceived aspect of writing in the West that needs investigating. Therein lies its
religious power, founded in the metaphysical notion of Logos: the Word of God.
This leads him to question the Western pact between being, Logos, and voice.
In addition, Derrida’s reconfiguration of the relation between writing and
voice challenges the position of Saussure and Saussurean linguistics. From Der-
rida’s perspective, Saussure preconceives language through the filter of writ-
ing (what he calls the problem of the written word) and not from oral sonority.
Saussure encounters sense in an anterior play between voice and phone, as a mini-
mal sound unit. So when Saussure contends that writing is a representation of the
voice (including this preestablished play), Derrida calls this “metaphoric” writing,

as opposed to “concrete” writing. For Derrida, Saussure’s work limits itself to

phonetic writing and never really approaches the question of the “inert writing”
(écriture) with which the voice interacts. Derrida derides the whole of linguistics
for this fundamental flaw (which he relates to certain Western philosophical pre-
suppositions) and its failure to explain satisfactorily the origins of language and
The strong sense of submerged violence implicit in Derrida’s view of writ-
ing colors his view of both the origins of writing and the colonial imposition of
this medium upon peoples considered illiterate. Derrida criticizes the Western
world’s arrogance for thinking of writing as if it were a form of knowledge supe-
rior to the orality of other cultures. This is why he undoes with such ferocity (in
his chapter on “The violence of the letter”) Lévi-Strauss’s structuralist idea that
orality comes before text, grounded in the mistaken assumption that the first na-
tions of the Americas had no “writing.”4
In Derrida’s view, no community can be without writing in its broadest sense
(what he calls “arche-writing”) since this emerges with the origins of language
itself. Even in the way that so-called oral cultures draw on genealogical memory
to generate proper names, a writing-like process of inscription takes place within
an institutionalized and socially agreed classification procedure, and certain types
of signs are used. From the moment that the proper name is “erased” or “substi-
tuted” in this way, “there is writing” and also a “subject.” In this sense, for Der-
rida, even speech is structured as writing.
For Derrida, then, the notion of “writing” as the proper basis of communica-
tion should include all examples of templates (glyphs, marks on ceramics, foot-
prints in a landscape) with which the voice interacts to produce the dynamic play
of “text.” Thus Derrida includes as writing a much broader set of textual prac-
tices than is usual in the West, for example the designs inscribed on gourds by
the Nambikwara, a group Lévi-Strauss lived with, and he accepts the Chinese
definition of writing, wen, as something that also includes glyphs, marks in stone,
wood, hide, prints in the earth and the constellations, in other words the whole
conglomeration of signs in any given territory, both above and below.
From this perspective, the broad definition of writing for Andean peoples
would include the patterns they make by throwing coca leaves, the images in-
scribed by lightning upon rock, their designs in ceramics, the tracings left by
dancers (or animals) upon the earth, the historical ritual sites organized by a
pattern of pathways (called ceques), their gestures made by “writing in the air,”
glyphs upon hide, and also their alphabetic writing on paper.5 In each case, writ-
ing constitutes the basic template with which subsequent vocalizations interact.
This compels us to revert to the etymology of “text” (Latin textus: textile, texto,

“to do, or weave”) to describe this “vocalizing” dynamics, with its power to revive
inert writing within a more regional logocentrism.
Derrida’s work sheds light on other questions. If many European precon-
ceived ideas about being and the relation between identity and the Other under-
lie the origins and actual practices of alphabetic writing, then we might seek in
the origins and practices of weaving equivalent preconceived ideas for Andean
peoples about being and the relation between Self (naya-nanaka) and the Other
(jupa-jupanaka), something particularly pertinent in the context of present-day
political and educational concerns with interculturality.
Nevertheless, Derrida’s very regional agenda of seeking the origins of writing
in Western Europe does not offer many clues as to how Andean peoples might
relate Self to Other, or voice to territory. Nor does it permit us to read easily the
relationship between Andean territory and its own forms of writing. This is why
the Andean textual theory we finally develop, while inspired by Derrida’s gram-
matology, necessarily draws on the more regional metaphysics addressed by the
Brazilian anthropologist Eduardo Viveiros de Castro in his notion of “ontological
depredation” as a pervasive characteristic of Amerindian populations and a key
to their ideas of Self and Other. In brief, this notion proposes that Amerindian
peoples constantly replenish their own cultures (and notions of Self ) from the
outside by appropriating vital aspects of the Other, in a context of struggle.
This important perception about indigenous notions of consumption, while
not addressing the question of writing directly, nevertheless has implications for
understanding the violence underlying textual practices, including indigenous
notions of writing. Such relations have depredatory and destructive aspects (in-
corporating the forces of the Other) as well as constructive ones (reabsorbing the
forces of the Other to reinforce the Self ); the articulation of these relations con-
stitutes an Amerindian ontology of being.
Returning to the theme of comparative writing practices, both Derrida’s ex-
ploration of grammatology and Viveiros de Castro’s notion of ontological dep-
redation help explain how the deliberate transference of a conceptual domain
from maternal textual practices (making textiles, reading kipus, and so on) to-
ward a Hispanic one (alphabetic reading and writing) are more than examples of
interlanguages or intertextual interference. Rather, they demonstrate a deliberate
transference by Andean populations, a resorting to transferences as a form of in-
tercommunication, and beyond that to a constructive appropriation of Otherness
as a compensatory textual strategy to enrich the Self.
In this sense, the linguistic definitions (and its biological counterparts) of in-
termediary stages between codes, as expressed in the rather contemptuous ter-

minology of semilanguages, interlects, pidgins, or hybrids, seem overly theoreti-

cal. The danger is that they do not take seriously the situated contexts of learning
the mother tongue, acquisition and use of other languages, and textual practices,
or even cultural practices.
For our purposes, it is just as necessary to be aware of the situations in which
these hybrid forms are created by Andean populations, whether in rural com-
munities or in their encounters outside these limits, where communal matters
still take precedence. In these situations, Andean populations have had their own
reasons for how they have managed the dominant language (L2) or textual prac-
tice (TP2). This may have given way gradually to the development of interlan-
guages or intertexts, with expansion toward urban centers through migration, or
the widespread periodic appropriation and use by Andean populations of other
spaces. But in no way can we find in the self-estimation of Andean populations a
derogatory attitude toward these intermediary forms or toward their own bi- and
We argue, then, that the context of intertextual formation has much in com-
mon with the motivations for creating Red English by indigenous peoples of the
United States6 or the regional Spanish of Andean populations,7 which both resort
to the grammatical features of the maternal tongue (L1) as a compensatory strat-
egy that permits the reproduction in practice of a whole set of prior values and
tacit cultural knowledge to achieve the perpetuation of identity, even within a
neocolonial setting. In our case, the similar application by Andean populations of
features of maternal textual practices (TP1) to the dominant one (TP2) is a strat-
egy that challenges official attempts (past and present) to impose literacy criteria
from outside by incorporating them into the dynamics of their own communica-
tive practices.
In this strategy of resistance, Andean peoples seem to approach reading and
writing with the intention of avoiding it in the narrow Western sense, but of ap-
propriating alphabetic writing as an expression of the Other, in their own terms.
We intuit that this same configuration of ideas underlies a properly regional ex-
perience of interculturality in a colonial setting. As a part of this same strategy,
we find that social memories are ordered by the relentless venture to reinstate
the regional systems of authority of a separate sovereignty. Even where domains
of sovereignty are overlapping, interpenetrated, hybrid, or destructured, regional
social memory refers to certain preestablished temporalized poles, more often
than not the Inka state apparatus, as points of reference on which to ground a
new generation of nexuses of collective articulation. We found these regional
modalities of formulating identity in our earlier study of oral tradition, River of
Fleece, River of Song (Arnold and Yapita 2001), where regional reproductive prac-

tices, ideologically molded through the discursive practices of shamans and the
songs of midwives, still control a historically grounded institutional setting of an-
imal census practices. And we find them in the institutional setting of schooling
in the modalities of interpretative practices concerning the nature of authority in
reading and writing.
In a second approach, we temper the detailed analysis of this question
through local ethnography, where the colloquial understanding of local discourse
is posed against the overpowering body of knowledge created by historians and
academia. Here we fill the documentary and juridical lapses in official accounts
with evidence for the structuring and authority of alternative means of jurispru-
dence and its jurisdiction.
Our approach here goes beyond the regional point of view, as expressed in
the kinds of discourse that anthropology usually scrutinizes (ethnohistorical, sha-
manic, ritual, and ceremonial), to reveal the basis of regional authority in found-
ing texts, and the constitution of authority through a doctrinal corpus with its
own set of arguments and assumptions. We trace the way in which formal tex-
tual thought is grounded in these authoritative texts of regional jurisprudence.
In short, we scrutinize the philosophical and juridical bases of a regional textual
polity, including the pertinent institutional sites where these are played out, above
all the school.
Using mainly anthropological and linguistic approaches, we analyze school-
ing in the Andes to include not only what happens in the classroom, but also
the historical relations between the community and the state, with its nexus of
social, economic, political, and ritual obligations. Through interviews with com-
munity members, we demonstrate their appreciation of schooling as part of a
tributary network with a much longer historical trajectory, and of the school pre-
cincts as part of a series of ritual and political sites that bind the community to
the state, not only the modern mestizo-criollo nation-state, but also the memory
of former Andean states.
From this perspective, a principal concern of community members regard-
ing schooling is their rights to land, historically mediated by writing. Only with
schooling could community members read and write and so defend their land
titles in courts of law. While this broader juridical framework of schooling served
as the basis for historical land claims in the recent past, at present it serves as a
forum for wider claims concerning cultural and linguistic identity. We suggest
this is because schooling in the community is concerned with instruction in the
wider sense of the ways in which indigenous peoples transmit their own values
through generations. While schooling has centered until now in the learning of

European reading and writing practices, there is today a growing awareness in

rural communities that Andean textual practices have much to contribute to the
Interestingly enough, this brings the concerns of community members in
line with the new pedagogy that has arisen as a result of similar experiences in
the rest of the world. One feature of this is the insistence that teaching literacy as
such is not enough; rather its definition should be broadened to include the so-
cial dimensions of the whole repertory of practices and competencies that pupils
experience in their world. In the Andes, indigenous groups propose that school
programs must include alternative writing practices, indigenous literatures, and
the multimedia and multimodal techniques of new technologies.8 In particular,
new proposals insist that the task of education should be a reconceptualization of
pupils’ needs in terms of the multiple textual practices demanded by the variety
of communication media (oral, visual, textual, tactile, choreographic, numerical)
that exist in one and the same community.
This means rethinking the function of language in terms of communication
media. In a real attempt to extend and transform the definition of literacy, the
acquisition of reading and writing would form just part of the wider processes
of the construction of meaning. Similarly, paper and ink can be seen as just in-
stances of cultural artifacts, on a par with weavings, kipus, music, song, video,
animations, film, dance, multimedia, and so on.9
In Bolivia, as in the rest of the world, the communicative demands of mod-
ern workplaces require urgent training in new and alternative literacy forms, for
example, in the new kinds of visualization produced by computer software, with
its combinations of symbols, texts, and images.10 Information and communica-
tion technology has abandoned the traditional notions of grammar, lexicon, and
semantics in favor of semiotic systems that bridge speech, writing, and reading.
This is leading to the development of new “grammars of meanings” in which
combinations and units form “comprehensible totalities.”11 In chapter 1, we ana-
lyze regional textual practices using an information technology approach, with
the aim of showing that far from being a brake on the development of the coun-
try, these practices include many aspects to be exploited in a future approach to
Recent studies of Andean mathematics,12 like similar studies of other parts
of the world, demonstrate the need to consider the pedagogical, cognitive, mne-
monic, and didactic uses of numbers in their social, cultural, and corporeal set-
tings. The incorporation in schooling of advances in knowledge, such as the re-
cent studies of the kipu based on cybernetics and theories of communication,
would permit what Víctor Hugo Cárdenas—the Aymara-speaking former vice

president of Bolivia—has called “another route to modernization.” From this

perspective, we argue that it is not enough to translate the concepts of universal
math to Aymara and Quechua, as does the present educational reform.
This approach allows us to sidestep the complexities of the debate around
orality and literacy, and whether or not the introduction of writing has influ-
enced the cognitive and conceptual development of children in the region.13 We
opt instead to study specific historical and present-day contexts in a determined
region to document the contact between different textual practices.
We applied the field methods of dialogical anthropology14 in our search to
understand the local textual polity and its formal bases of thought and argu-
ment, and how these articulate with the regional history of schooling in gen-
eral, and the advance of the current educational reform in particular. To this end,
we carried out interviews (formal and informal, structured and semistructured,
the greater part in Aymara) with the community members concerning their at-
titudes to different educational situations (state education, communal education,
and communal educational rituals), in a comparative framework. Details of the
qualitative and interactive methodologies we use have been described extensively

livichuco and other field sites

The principal body of fieldwork for the study was carried out in the ru-
ral community of Livichuco in ayllu Qaqachaka (Abaroa Province, Oruro De-
partment) in 1997–1999. (An ayllu is a kinship-based community, usually with a
founding ancestor in common.) Having been involved in the region since 1984,
we were invited to work there by the comunarios (members of the ayllu commu-
nity) and their community leaders and schoolteachers. Other comparative stud-
ies were carried out by our students in Aymara-speaking communities in Oruro
(N. Carangas Province, Saucari Province), La Paz (Ingavi Province), and Potosí
(Rafaél Bustillos Province) where they had been brought up, or made present-
day contacts. These periods of fieldwork were interspersed by brief visits to the
teacher training institute of Caracollo (Oruro Department), where Livichuco
teachers are trained. Thus the evidence and interpretations presented here come
from a wide area of the Southern Andean highlands with a total population of
about 640,000 (see figure 1).16
Livichuco is a relatively quiet corner of Qaqachaka within the wider land
struggles that embroil its neighbors, the Laymi and Jukumani, and that charac-
terize the region as a whole. An ample bibliography on the zone includes work
of our own and others.17 With its immediate neighbors, Ventilla and Qachüta,

River Desaguadero

Fig. 1. Map of communities mentioned in the study.

Livichuco has some thirty families. Located in the minor ayllu of the same name,
Livichuco (sometimes called Araya or Arriba) is one of the six constituent ayllus
of Qaqachaka. Some nongovernmental organizations working in the zone, with
an Indianist fervor to vindicate the ayllu past, now call Qaqachaka a “major ayllu,”
although it is more commonly known by the comunarios as a former ecclesiastical
annex (añiju) of the Toledan reduction town of Condo Condo, which was formed
when the sparsely settled Andean populations were grouped together in towns,
with the aim of indoctrination and tax collection. Situated on the modern bound-
ary between the departments of Oruro and Potosí, the region of Qaqachaka was
the focus of a particularly ferocious wave of fighting around the year 2000. How-

ever, this aspect of Qaqachaka life is not new; oral history recounts how the Qa-
qachakas, as a former part of Qharaqhara (and Killakas Asanaqi) were warriors
of the Inka, a tradition that survives in social memory to this day.
Livichuco hamlet, at four thousand meters above sea level, has a pastoral
economy (llamas, sheep, and alpacas) and with the climate change has begun to
produce bitter potatoes and onions for sale. It is some two hours by truck, the last
part on dirt roads and mountain tracks, from the regional market town and cen-
tral municipality of Challapata.
There is marked out-migration to urban centers for cash income and “in or-
der to give better opportunities to the children.” But although the comunarios, es-
pecially the men, travel frequently to cities, staying up to two years in productive
places such as Caranavi or Cochabamba, a strong communitarian sense compels
them to return once in a while, especially at formal fiestas, to reclaim their land
Qaqachaka has a rich and dynamic tradition of song and music, weaving and
braiding, dance, and oral history. Historically, Livichuco was the principal tambo
or postal way station of Qaqachaka, and an elaborate history and oral tradition
is centered in this building complex, which served as the school until 1999. In the
last five years, the APSU project (Artesanías Para Seguir Unidos) has stimulated
textile production, as well as the rescue of natural dyes and the gradual recupera-
tion of the camelid flocks, in particular the males, lost in decades of the constant
fighting and generalized rustling by enemies. The sale of weavings in urban cen-
ters and now on the international market is creating a significant source of mon-
etary income.
According to the 1992 census, 99 percent of the population in Qaqachaka
speak Aymara and 46 percent (mainly the men) speak Quechua,18 which justified
the application there of bilingual and intercultural education. Serving a popula-
tion of 4,252 (according to the 2001 census), the educational district of Qaqa-
chaka has a central school (núcleo central) and eighteen branch schools (escuelas
seccionales). The current educational possibilities include basic and intermediary
levels in the central school and since 1999 the baccalaureate. The branch school
of Livichuco has around 90 pupils and 3 teaching staff, while the central school
has more than 700 pupils. The whole ayllu, with more than 1,200 pupils and 120
teaching staff, is the largest educational unit in the department of Oruro. Its size
and population density presents a wide array of educational examples.
Until 1999, Livichuco’s old school precinct had just two classrooms and two
rooms for the teachers and their families. It should have entered the educational
reform transformational program in 1998, but administrative problems in the
Challapata Municipality concerning the construction of the new school com-

pound (with the reform design) and teachers’ lodgings postponed this for a year.
During our fieldwork period, the Livichuco school was still a part of the im-
provement program, with traditional techniques of teaching and learning (in
multigrade classrooms for grades 1–3 and 4–5), besides a number of different
civic activities, both local and national, that form part of the general school year.
The beginnings of change could be felt most insistently in the retraining courses
for teachers organized by nongovernmental organizations working in the zone.
During this period of fieldwork, we searched for regional experts in differ-
ent fields in order to deepen our knowledge about the textual practices they use.
Among them were Don Donato Inka and his son, Santiago (wise ones and his-
torians of the place), Don Domingo Jiménez (traditional regional authority or
mallku mayor, historian, and wise one), Elvira Espejo Ayka (midwife, weaver, and
art student), Doña Sabina Mamani, Doña Agustina Mamani, and Doña Lucía
Quispe (all midwives and wise ones), Doña Antonia Espejo (mother and weaver,
with experience in school duties), and Don Feliciano Maraza (father and braider
of fine slings). We include only some of their voices here, confident that they rep-
resent the tendencies and perceptions of many rural communities, above all the
great unease among older people, the holders and transmitters of culture, when
faced with the realities of modernization.
Throughout the book, we give special attention to the Aymara terminol-
ogy used in these interviews. The Aymara orthography in this book follows Juan
de Dios Yapita’s phonemic al-
phabet of 1978, adopted as
1 . A ymara consonant chart
the official Aymara alphabet
in Bolivia in 1983, the only Occlusives
change being the use of h for Simple p t k q
aspirated sounds. In speech, Aspirated ph th kh qh
all accents fall on the penul- Glottalized p' t' k' q'
timate syllable; the elision of Affricates
some final vowels in speech Simple ch
marks the direct object and
Aspirated chh
has additional prosodic func-
Glottalized ch'
tions. We generally use the
Fricatives s j x
Aymara variant that predom-
Laterals l ll
inates in Qaqachaka ayllu and
Nasals m n ñ
neighboring regions of Oruro
Semiconsonants w y
and northern Potosí. See ta-
Vibrants r
ble 1.19

Although there are many phonetic vowels in Aymara, the three phonemic
vowels are written as /a/, /i/, and /u/. Pronunciation approximates the sounds
of vowels and consonants in Spanish and English, with the following exceptions:
ä Elongated a, pronounced as in the English father
chh Aspirated ch, pronounced as in the UK English chips
ï Elongated i, pronounced as in the English peace
j Pronounced as the h in the English half
k Pronounced as the k in the Spanish kilo
kh Aspirated k, pronounced as in the UK English cat
ph Aspirated p, pronounced as in the UK English pipe
q Pronounced further back in the throat than the k; no English equivalent
qh Pronounced as in an upper class can’t in UK English, with a long vowel
th Aspirated t, pronounced as in the UK English tease
ü Elongated u, pronounced as in the English prune
x Pronounced as in the Scottish loch
' Denotes a glottal stop, a quick cutoff of the preceding sound in the
back of the throat

Textual Struggles

A n d e a n T extual Pol i ty

In approaching Fourth World literature in its great diversity of origin and form, we need to
settle the matter not just of grammatology but also of text.

Gordon Brotherston, Book of the Fourth World

The institutions that constitute civil society functioned as passageways that channel flows of
social and economic forces, raising them up toward a coherent unity, and, flowing back, like
an irrigation network, distribute the command of the unity throughout the immanent social
Michael Hardt and Antonio Negri, Empire

The world’s remaking in the 1990s through information technology and commu-
nications has been unforeseen and all-encompassing. An electronic interface has
enveloped daily life and manners of conducting business, with new communica-
tion media (computing, digital display, optical fibers, mobile telephones, e-mail,
visible languages, geographic information systems) and textual practices (writing
onto a computer screen, manipulating a PlayStation, or weaving a technotextile
made of fiber-optic threads).
However, these new directions in technological development are rooted in
the past and linked to countless institutions. One aspect of this technological dy-
namics is the role of communication media, and the textual practices associated
with them, in the reproduction of distinct kinds of polity and the relations of


contrasting modes of production. As various authors have noted (Althusser, Wil-

liams, Bourdieu and Passerón, and more radically McLuhan), communication
media are, at a profound level, the media of production. In their capacity to com-
municate pertinent information between different levels of society, according to
McLuhan’s famous dictum, “the medium is the message.”
Andean societies developed textual polities founded in cloth.1 Their princi-
pal textual practices of weaving, braiding, and the knotting of threads (to pro-
duce what in Quechua is called kipu, or quipu, and in Aymara chinu) gave rise
to a multitude of other practices (dancing, painting images, oratory, making li-
bations, praying, singing, playing music) and regional textual forms (cloth and
braid, glyph and song) that stem from common weaving elements. All of these
diverse practices are generated from predetermined additive sequences based in
cloth. As communication media, such textual practices reproduce the modes of
production of Andean societies grounded in particular territories. Their principal
textual origin is fleece, the product of grazing flocks on extensive pastures, and as
many have observed, the very quality of Andean fleece is a direct consequence of
grazing animals on lush grasses and clear waters.2
In contrast, the textual practices of European reading and writing, as com-
munication media foreign to the Andes and only introduced with the European
invasion, help reproduce the modes of production of Hispanic society. Continu-
ing the European tradition, their textual basis is paper, a product from distant
woodland territories, which aided the bureaucratic functioning of the colony
and, nowadays, of the nation-state.
Understanding the ways in which Andean populations historically managed
this relationship between local productive and textual spheres by continuing cer-
tain aspects of their own practices while appropriating European writing, cap-
tures the reality of the present clash between different textual practices in the

T e x t u a l P o l i t y a n d A lt e r n at i v e F o r m s
of Writing in the Andes

First of all, in the absence of alphabetic writing, how did Andean polities
manage the multitude of populations, territories, and activities under their do-
minion? To what point can we apply Derrida’s philosophical considerations to
the Andes, where textuality and writing were originally based in cloth and knot-
ted threads? As a preliminary answer, we suggest they did so through an archaic
form of “network production” based in cloth, yet still grounded in territory.3
Let us trace how this functioned, basing our arguments on studies written to

date. Undoubtedly, certain insights into the workings of Andean woven polities
have emerged from the interweaving of information sciences, cybernetics, and
semiotics. The semiotic study of cloth locates the different signs and symbols of
textile manufacture to referents in the physical and conceptual world. In essence,
the relationship between the organizing language of textiles and speech suggests
homologies between the formal structure of textile designs and the syntax and
organization of discourse in spoken languages. The seminal essay by Verónica
Cereceda, “Semiologie des tissus andins” (1978) opened a decade of studies of
this type.4
A difficulty of this semiotic approach is that it tends to generate abstract and
closed models, erasing the contexts in which weaving actually takes place. But
there are certain advances: Cereceda’s approach challenges Saussure and sup-
ports Derrida, reconfiguring the relationship between the activity of weaving and
speech. Instead of perceiving in writing a secondary representation of voice, Ce-
receda perceives in cloth, as a form of writing, the primary dynamics of voice.
Her “text” of study becomes the interaction of textile with voice.
As part of the same semiotic tendency, in their essay “The Weaver’s Eye”
(1992, 51–53), Edward and Christine Franquemont (both weavers) and Billie Jean
Isbell intuit that the basic principle of weaving—whereby a small nucleus of in-
formation is repeated rhythmically and symmetrically in order to fill space and
time—organizes all other fields of activity: conceptual, social, and physical. This
suggests that the strategy (or dispositif in Foucault’s terminology) through which
a polity founded in cloth configures its rule has a fractal-like nature, whether in
its systems of communication or means of social control (of subjectivity or the
peoples under its dominion). A common woven logic would also facilitate the de-
ployment of this strategy.
This means abandoning the semiotic signification of texts as closed codes
(whether written or woven) to examine the dynamics of communicative strate-
gies and practices. Writing is just a part of a whole gamut of techniques and prac-
tices in the flow of information, from its production, storage, and retrieval to its
dissemination. In present-day information technology and communications, the
storing of information (codified in electronic circuits, characters, icons, and so
on) occurs in the memory systems of computing. But before information tech-
nology, alphabetic writing stored information (codified in units of letters, words,
sentences, and discourse) in deposits based in paper.
By contrast, the majority of rural weaving activities in the Andes even today
are based in fleece, and information is stored in materials (cloth, kipu, braids)
derived from fleece. Likewise, the units of writing—letters, words, sentences,
and discourse—find an analogous form in cloth. Contemporary Andean weavers

Fig. 2. The Coya’s belt in the American Museum (B/4642) (in Murúa [1590] 1946,
reconstructed in Desrosiers 1986).

compare the interlacing of warp and weft with a series of written letters, as did
their Inka counterparts many centuries ago. Consider the following example (fig-
ure 2) taken from one of the belts of the Inka empress (Coya),5 where cloth em-
bodies information while serving as its medium of construction.6
Such items of dress might be organized along similar semiotic lines in many
parts of the world. However, Andean textiles also served (and still serve) as the
medium of more complex systems of communication. Abundant evidence in
the literature shows how Andean peoples administered their economies through
cloth. Weavings codify and store information on local sites of production, in-

cluding details of local ecologies and the social organization of landscapes, flora,
fauna, and avifauna.7 As maps, textiles encode sites of local topography and the
lines of communication interconnecting them.8 Our own work in the Southern
Andes found a persistent pattern of “thread-knot” that traces the alternation be-
tween movement and rest, whether of the gods, of humans in their daily produc-
tive tasks and festive breaks, or animals and plants.9 In addition, weavings such
as the mantle (which doubles as a carrying cloth) relate these earthly pathways
to their celestial counterparts.10 As an incipient “informatization of production”
(Hardt and Negri 2001), woven logic differentiates here between sites of produc-
tion and the pattern of routes connecting raw materials and labor to intermedi-
ate deposits or sites of consumption.
In the same region, textile production (obtaining quality fleece, dyeing it, set-
ting up the loom, counting the warp yarns, and then weaving) is organized into
a grammatology along three main axes: gender (the division between male and
female), logical taxonomy (of position, and the interrelation between the whole
and its parts), and age groups (according to relative age, and the genealogy of
primary elements and their derivates, or offspring).11 Elements of these textile hi-
erarchies can be traced historically at the level of visual languages. Mary Frame’s
seminal essay, “The Visual Images of Textile Structures in Ancient Peruvian Art”
(1986) deals with logical taxonomy in the structure of the linked and sprang tex-
tiles of the southern coast of what is now Peru. These early weavings, twisted
by finger, still had no weft. She argues that underlying patterns of yarns (their di-
rection, lay, and points of articulation) constitute the metalanguage (or dispositif)
that organizes the designs of the surface structure.
Anne Paul’s studies (2000a and b) of the Early Horizon and Early Intermedi-
ate periods of Paracas culture (ca. 100 BC–200 AD) on the same coast suggest
that such patterns reinforce the relation between garment and wearer. Her work
identifies elements of the visual language of weavers of that period, when the
body was perceived in terms of an alternating play of gyrations (around neck,
waist, and knees) expressed through the patterns and placement of enveloping
garments. She holds that key forms in this play of gyrations (like Foucault’s dia-
grams) reproduced fundamental aspects of social hierarchy in Paracas culture,
including gender and age relations.
Such ideas begin to suggest how Andean and European forms of writing (in
the broad sense) are ontologically different. An emerging body of scholarship re-
veals how textiles, as extensions of the person, embody aspects of being in their
terminology and iconography.12 As vital sites for constructing identity and sub-
jectivity, weavings fulfil an important policing function in social control, allowing
the tracing of wearers’ ethnic identities, tribute categories, and social status. The

body itself is perceived as something woven, while cloth as a kind of prosthesis

integrates body and mind.13 As living beings, textiles embody what René Devisch
(1993) has called “synesthetics,” that is, a diffuse body-centered esthetics. Andean
textiles differ fundamentally from alphabetic writing in the ways they embody
notions of being, knowledge, and social memory and control.
Another weaver, Sophie Desrosiers (1997), traces how weaving and culture
intersect according to the parallels between textile logic and cultural logic. Her
idea shares features in common with the social semiotics developed by G. Kress
and T. Van Leeuwen in their “grammar of visual design” (2001). These authors
reject the usual semiotic notion of an intrinsic relation between signifier (form)
and signified (meaning) in favor of a motivated conjunction of signifiers and sig-
nifieds “formulated in relation to the sign-maker and the context in which the
sign is produced” (2001, 7).
Like Derrida, Kress and Van Leeuwen view forms of visual representation as
essentially separate from language (forms of speech). Only in early forms of al-
phabetic writing does visual representation take over language as a means of re-
cording. In other cultures, such as in the Andes, these remained quite separate.
For them, artifacts such as kipus permitted Andean cultures to encode verbal lan-
guage, together with other aspects of culture best represented in a visual (and tac-
tile) form. Although there are important homologies between visual and linguistic
structures, there is no priority of one medium over the other; the visual has not
become subsumed to the verbal as its primary form of representation (2001, 19).
From this perspective, the textile logic of images that express the patterns of
textile structures might also guide their vocal production in narrative. A historical
example of this possibility is John Rowe’s conjecture (1980) that the iconographic
codification of early Chavín culture (ca. 1200–300 BC), with its baroque ornamen-
tation, was related directly to narrative structures and coded in visual units ho-
mologous to simile, metaphor, and metonymy. (He gives the comparative liter-
ary example of Old Norse court poetry, particularly the Icelandic kennings.) Mary
Frame (1986) similarly suggests homologies between weavings and narrated sto-
ries in Andean coastal textiles. These are expressed through spatial and temporal
markers of character (possibly also of voice) in the juxtapositions of textile de-
signs, for example in the use of the evocative snake head image as a structural el-
ement where yarns cross.14 Gail Silverman Proust’s contemporary ethnographic
work in Q'eros (Peru) concerning the expression of narrative images (above all the
Inka’s head) in certain textile designs reinforces this possibility (1994, 113ff.).
In this way, woven elements shape discourse and its units, stylistics, and liter-
ary features, including the manifestation of textile voice. We indicate elsewhere
certain homologies between the three- and four-yarn units of ancient textiles and

their associated design motives and the three or four characteristic voices of con-
temporary Aymara tales, in which tripartition has an expressive and emphatic
quality. Embedded in indirect reported speech (the equivalent of “saying, they’d
said, they say”), these voices are ordered in time, informing us of the identity
of the original authoritative source of the tale even though they are long dead.
Other modern studies indicate the “intertwined” nature of Aymara sentences
that in critical junctures can transform into “hidden sentences” within the weave
of the text, seeds of ideas that come to fruition later on.15
A woven underlay patterns discourse forms and processes in its image. This
might be why the basis of Andean discourse is most often a braided dialogue be-
tween at least two people. Just as in musical performance, one person guides and
the other follows, filling in any void in oral memory. This same woven under-
lay seems to organize suffixes as discourse markers that follow the thread of the
narrator’s point of view throughout a tale, organizing the pragmatic coherence
(or “evidentiality”) of the text. Similarly, one finds woven links between themes
(what Huanca calls “opening and closing illustrative topics”) and woven connec-
tors between stories, all of them sewn into one rhapsodic cycle.16 As communica-
tive devices, the names of elaborately decorated braids (k'anata) make rhymes in
the wedding songs of Qaqachaka. But their presence in song also replicates the
braiding together of couples in an interlacing of hands, which in turn forms part
of a wider braiding of families with their animals and food produce, in different
ecological niches.17
A similar woven logic pervades the organization of knotted threads. Used
as accounting devices in other parts of the Americas (for example by the Mapu-
ches), they reached the most complex forms in Tiwanaku and other early An-
dean cultures.18 It was under the Inkas, however, that kipus were systematized
into templates for state rule, administration, and planning, a development that
made possible the immense reach of this Andean empire.
Various authors interpret Inka kipus as counting systems related to Inka ad-
ministration. But J. V. Murra, in his classic essay of 1975, demonstrated that kipus
functioned both mnemonically in registering numbers and also through a sys-
tem of local categories with a logically predetermined ordering. In the past few
years, other studies have attempted to widen kipu analysis by examining the logi-
cal coding and decoding of their woven language (by color; knotting; thickness
of yarn and knots; direction of spin or lay; and whether they are twisted to the
right, S, or left, Z).
Some researchers propose a common language between kipu and weaving.
For example Silverman Proust (1994, chap. 6) perceives a relation in Q'eros (Peru)
between kipu knot size and quantification and modern textile equivalents in the

width of stripes of everyday woven bags, a homology reiterated in the color code
used. Other studies delve into kipus with a view toward their possible narrative
structures; Martti Pärssinen (1992) especially proposes that kipus had a phonetic
reading. Gary Urton (2003), on the other hand, perceives in the insistent binary
language of kipus a meta-coding device, which he compares to the ASCI code of
computer logic. Still others propose a dynamic relation between kipu, voice, and
In light of these possibilities, Andean kipus are said to have served as the basis
of modern information technology, for example in the present use of kipu design
in hardware organization, or as the model for bar codes in the global market.20

Textile Dynamics

In this other kind of writing—woven and braided, traced on the ground, or

vocalized in a distinct manner—both grammatological (writing-centered) and
grammatical notions are expressed differently than in alphabetic writing. Unlike
written descriptive grammars with their strong standardizing tendencies, An-
dean folk grammars are disseminated at woven, visual, and oral levels. The North
American linguist Bruce Mannheim (1986) argues that the Chomskyan notion of
a “completely homogenous speech community” is the fantasy of scholars and
standardizing movements in the West. Drawing his ideas from Jakobson, he en-
counters the real grammar of Andean languages not in linguistic texts but in the
pragmatics of popular practices based in such textile devices, for example, in the
organization of Quechua popular songs. Instead of a priori definitions (based in
written Latin), Andean grammars function through a common visual-oral lan-
guage of rotation, reflection, repetition, parallelism, sliding or staggered design,
and duality.
Gordon Brotherston expresses the same idea in a different way. In his Book
of the Fourth World (1992) he argues that if writing is based ultimately in the dy-
namics of voice, then the grammar of orality must be based on vocal expression.
Drawing on Derrida, Brotherston rescues the Greek term gramma to describe the
expression of writing according to its relationship with voice. For instance, a song
line is a measure of gramma in the sense that the line is sung in one single breath.
Likewise, as we shall see in later chapters, each passing of the weft in weaving in
a sense “vocalizes” the warp, which acts as its written support.
Beginning with Dell Hymes (1981), a whole generation of linguists and an-
thropologists interested in the ethnography of communication has attempted to
express this vocal dynamics on paper. If the characteristic pattern of orality is
tracing threads of sound in the air, then they seek to express these threads of

sound on the page. For them, the notion of prose is only possible in a written reg-
ister of language that overlooks the original vocalization, whereas orality always
follows its voiced gramma.
Apart from simply emitting threads of sound (like furrows or lines in the
air), orality gives an importance to this sonic emission of language (and its units:
phones and syllables) quite distinct from the grammar of written language. Mod-
ern Saussurean linguistic theories, based on the significance of the larger units
of language according to a mode of analysis that excludes vocal dynamics, is not
much help in understanding the perception of these units. More useful are medi-
eval theories of language, for example the phonological theories of the grammar-
ian Antonio de Nebrija (written in 1492), and Julia Kristeva’s observations about
the Sanskrit grammar of Pānini (1989). These authors point to the sacred nature
of sound quality as a fundamental element in vocalization. For Nebrija, the qual-
ity of sound is founded on the Christian Logos of the period. For Kristeva, the
Sanskrit sphota is an instance of sacred Indo-European language, “wherein mean-
ing bursts forth, spreads out, germinates and gives birth to itself.” Similarly Pla-
to’s reflections on chora express a vocal rhythm that precedes and underlies figu-
This dynamics based in live performance (versus the written word) is a feature
in common with other textual practices throughout the world. Saskia Kersen-
boom’s exploration in her book Word, Sound, Image (1995) of the “life of the text”
among the Tamil of Southern India helped us appreciate other forms of textu-
ality. There, as in the Andes, the representation of a text, its meaning, and the
proof of its comprehension are all produced through performance. The object of
a text is not to “read it” but rather “witness it” or make it present in the world. In
the Andes, similar ideas concern the dynamics of the text and its life, genealogy,
and corporeality.
In Writing without Words (1994, 22), Elizabeth Hill Boone proposes that in a
medium such as the kipu, the dynamic interrelation between the elements gives
them meaning. In contrast with phonetic and logocentric writing, the kipu for
her is a meaning-centered system that functions according to the codification of
the relative position of elements that are more conventional than iconic (color,
texture, form, and size, and form of knots and threads). Basing her arguments on
Ignace Gelb’s earlier study of writing (1952), Boone terms such meaning-centered
forms of writing a “semiography” (from the Greek semasia, “meaning”), citing as
examples the semasiographic systems in the West such as musical scores, math-
ematical formulae, and the international symbols found at airports. In each case,
meaning is revealed through a common conceptual network of graphics, icons,
and symbols. She suggests that the kipu functioned outside linguistic references,

perhaps at a supralinguistic level, by means of “codes of knowledge” shared in a

common Andean cultural domain, but where different languages were spoken
(1994, 15–22).
Archibald Hill (1967, 94) calls such systems (whose meaning is given by in-
ternal conventions and structures) “discourse systems,” a definition that captures
the capacity of weavings and kipus to generate an ample vocal discourse from a
shared background knowledge, in spite of giving a minimum of information.
As meaning-centered or discourse systems, weaving and braiding provide
conceptual bridges between the more colloquial notions of writing in Europe
and the Andes. This possibility goes against the positions of Walter Ong and Jack
Goody, who hold that cultures without writing are incapable of expressing or
thinking in the abstract. Goody in The Domestication of the Savage Mind argues
that only writing allows the possibility of abstract thinking, through the tech-
nique of recording and organizing information in lists inscribed in paper.22 As
we shall see, making libations based on the configurations of a kipu is also a way,
both abstract and practical, of recalling in speech a list of categories and inscrib-
ing them in memory, in a writing-like way, whose elegance lies in the possibility
of generating a maximum of discourse from a minimum of information.
So how might Andean textual practices (voice, weaving, and kipu) be inte-
grated into the wider communication and control systems of which they form a
part? Understanding these dynamics is crucial. As Marx himself said, only when
the interconnections of the whole flow with incessant renovation does each so-
cial process of production become, at the same time, a process of reproduction
([1869] 1973, 543).
In place of alphabetic writing, a ubiquitous logical ordering of kipus and tex-
tiles must have been instrumental in registering information flow between differ-
ent levels of the Inka empire. A key for understanding the empire’s spread lies in
the dynamics of kipu or textile use, ignored until now. Therein we find not only
the notions of voice, body, memory, being, and knowledge, but also the dynamic
relation among all these to territory.
To understand this relation, the corporeal site of vocal production must be
linked with its dissemination. Until now, this matter has not been researched ad-
equately in Andean studies of vocal practices. Oral genres, their taxonomies and
characteristics (for example in the works of Jesús Lara), the social and cultural
context of their production, and the major and lesser units of their organization
(whether as part of the poetics of libation making or of song) have all been stud-
ied.23 Now we need to study the dynamics of voice from a regional and a corpo-
real point of view. We need to ask: How does vocal performance link the body,
through texts, with social memory and Andean terrain?

The dynamics between the different levels of organization found in rural

weavings and other writings outside the immediate ambit of modern Europe is
based in tangible and often corporeal units. The pragmatics of its operation de-
fies the abstract models of space and territory so favored in the Western world.
We must turn our perspective from dry electronic networks toward more organic
and corporeal ones. In parallel, we must relate the language of regional writing
dynamics, and the very notions of voice, to this organic world.
Jean Aitchison’s concept of “the language web” (1996), at once social, cultural,
and biomorphogenetic, proposes the kind of organic network that would coordi-
nate activities at a linguistic level. When we think of Andean textiles as a kind of
original “written” support, then the linguistic dynamics between different levels
of organization might find homologies in the levels of vocalization. In this sense,
each woven garment, as an expression of the speech of its weaver (Saussure’s pa-
role), concerns a more individual perspective toward a particular woven territory
and its life forms. At a higher level, the different textile genres (aguayo, punchu,
wallqipu, istalla), like different languages (Saussure’s langue), express the aspects of
a common woven territory. At a higher level still, the different voices intertwine
in a common language of identity in a recognizable woven territory. Through
this deliberate interlinking of woven aesthetics, vocalization, and territory, the
Inka state was able to control the different ethnic groups under its dominion.
This same hierarchical encoding of voice, number, color, and design en-
ables weavers to “think through cloth” in organizing and articulating productive
relations at different levels (between person, family, community, the state, the
gods), of lands and herds, throughout the body politic. At times, the symbolic-
analytical dimensions of woven practice functions communicatively to exchange
information in highly creative ways. In River of Fleece, River of Song (Arnold and
Yapita 2001, chap. 6) we show how the weaver-singers of Qaqachaka, conscious
that textile ontology derives from the fleece of their flocks and the produce of
their lands, develop song episodes that communicate to others the conditions of
production for generating quality fleece by grazing their herds on good waters
and pastures.
It seems to us that these weaver-singers attribute to fleece, as a communica-
tion medium, an expansive quality based in the organic material of which it is
composed, in a conscious expression of the material basis of fleece as the pri-
mary reproducer of the local modes of production. They recognize that the el-
ementary particles of fleece are made up from a germinal seed element, which
then generates another stemlike element. These elements intertwine in quan-
tity to generate a woolen covering to the whole territory. The human activity of
weaving reproduces these elements. The movement of the weft is compared to

the boustrofedon movement of plowing furrows in a field (as were the lines of
writing on the pages in our field notebooks). Textile designs (like seeds) reiterate
this germinal nature of fleece (and other vegetative) structures, from which they
are constituted. These develop in time through the intervention of other nour-
ishing elements (so other designs, stripes and figures, express images of water
and rains, dung and urine) to germinate and then expand the elementary units of
growth (in roots and shoots) into a vegetative-animal covering.24
All this is driven by human intervention in the daily farming and herding
chores (physical labor, weeding) facilitated by a homologous network of paths
leading to fields, expressed in certain weaving figures, and longer pathways to the
warmer valley lands, where highland produce is exchanged, expressed in other
figures. Then, at a cosmological level, this network of earthly pathways has its
celestial counterparts, in the Great Path of the Milky Way, and its tributaries, re-
membered on festive occasions in the songs to the animals.25
Productive spaces, along with the elements of production, are thus replicated
in textile spaces and their figurative elements. Innumerable studies confirm that
textile stripes, pampas (fields or meadows), and other designs replicate localities
themselves, the paths that conjoin them, and the produce resulting from local
lands.26 Different kinds of territory are reiterated in the terminology of these tex-
tile parts. In the Qaqachaka region, some relate more to farming; others more
to herding. In their function as maps (of ayllus, with their rivers and boundaries),
the inhabitants of Coroma (Potosí, Bolivia) have used their weavings juridically,
to prove their claims for land.27 At the same productive level, generative notions
centered in cloth agree with genetic science; an initial “seed” unit develops in the
textile sense according to the generative rules (codes, transition rules, genetic al-
gorithms, and mutations) of cellular automata.28
In all this, there are universal factors of realization, taken for granted by those
who practice these textual forms but ignored by those who only pay attention to
the meaning of written texts (on paper). A key element among these is the cor-
poreality of the productive processes, whereby the sonority of voice is consid-
ered to have generative power, like seeds that sprout, and breath, an animating
character. In practice too, the weaver-singers of Qaqachaka literally mix fertiliz-
ing breath with the seeds of the corn beer (chicha) they drink, to direct the power
of vocalization in their songs to the animals, like seeds, toward the future germi-
nation as new offspring of the harvest.29
For Andean populations that practice farming and herding, the basic units of
germination, seed, and stem are primordial. They transpose this organic imagery
to the structure of fleece, weaving, and kipu making and its logic to alphabetic
writing. At a higher level of organization, the Inka state resorted to knotted kipus

(much older than the Inka empire itself ) to “bind” local communities into its im-
mense territory. Kipu dynamics seem to have worked in much the same way as
that of other textiles. Inka state kipus ordered the documentation of imperial
units (in people, animals, food products, and other goods and chattel), registering
the flow of tribute from the far ends of the empire toward its center and of rec-
ompense to the periphery in turn. We know from colonial chroniclers and mod-
ern historians that the working of the empire depended on physical labor rather
than tribute in kind, although on occasions there was tribute in children, for ex-
ample in the sacrifices of the Qapaqjucha.
Included in these dynamic processes were the access and handling of territo-
rial waters needed for farming and grazing. Evidently, the kipus of Chachapoyas
measured the water at a local level, just as did the Waruchiri kipus examined by
Salomon.30 In this sense, a vital part of their workings concerned the flow of liq-
Another part of kipu dynamics concerned head counts (animal, plant, and
human). In his Book of the Fourth World, Brotherston reminds us that the Inka
kipu system was essentially a pastoral discourse, a register of heads of cattle,
where the economic origins in animals were inseparable from the roots of em-
pire.31 Later, we shall see that kipus were equally a farming discourse, which dealt
with the “heads” of farming produce. As for the circulation of persons, Tristan
Platt and others hold that kipus registered the transactions of matrimonial alli-
ances, and they must also have indicated the flow of manual labor, of fiscal obli-
gations and loans, and of people and their goods.32 Finally, during warfare, kipus
must have registered the flow of heads (as war trophies) and of available military
reserves (as a contribution in blood).
On the other side of the coin, something “owed” (or still to be contributed
in time or quantity) in the dynamics of the system was considered a crime. The
Andean notion of jucha as something owed was transposed in the colonial period
toward the Christian meaning of “sin.” But, the importance of its original sense
was at the heart of institutional ways of confessing outstanding obligations orally
before the highest ayllu authorities (the guilt was evidently cumulative), and of
noting them on the kipu knotted cords.
Under new guises, kipus continued in use for many decades after the inva-
sion, even serving as the textual foundation for translations into the new writ-
ten mode of documentation.33 Evidence suggests that the famous two-thousand-
page-long letter to the King of Spain attributed to Guaman Poma (Nueva corónica
y buen gobierno, ca. 1613) was based on the reading of kipus and other mnemonic
devices.34 Guaman Poma’s insistence on ordering things (whether written or
drawn) according to an Andean hierarchy (as if he were following a predeter-

mined logic), suggests that the direction, colors, textuality, and type and sequence
of knots marked not only the quality, gender, and class of objects annotated but
also the spatial logic of Andean hierarchy, as Murra intuited. Even today, it is pos-
sible to hear a wise one, such as Don Domingo Jiménez, “read” his Aymara tales
with directional suffix markers, as if he were reading a kipu (or its Aymara equiv-
alent called chinu).35
In a new kind of analysis, some scholars are considering the written transcrip-
tions into Spanish of kipus from the early colonial period. The parallel reading
of these two communicative media, kipu and alphabetic writing, has stimulated
these researchers to “penetrate the code” of the kipus (just as the Rosetta stone
facilitated the decipherment of Egyptian hieroglyphics) by seeking their gram-
matical units.36
Kipus were not only used for bureaucratic matters. The recent discovery of
colonial drawings of a “literary” kipu attributed to Father Blas Valera, a mestizo
Jesuit, implies the mnemonic use of a syllabic model of organization, based on
suffix position and rhythm, to record the famous “Hymn to the Rain” of an Inka
Súmac ñusta Pulchra Nimpha Fair maiden,

Toralláiquin Frater tuus Thy brother

Puiñuyquita Urnam tuam Thine urn

Paquir cayan Nunc infringit Is now breaking.

Hina mantara Cuius ictus And for this cause
Cunuñunun Tonat fulget It thunders and lightens
Illapántac Fulminatque And thunderbolts fall,
Camri ñusta Sed tu nympha But thou, royal maiden
Unuiquita Tuam limpham Their clean waters
Para munqui Fundens pluis Shalt give us rain;
Mai ñimpiri Interdunque And sometimes too
Chichi munqui Grandinem, seu Shalt give hail
Riti munqui Nivem mittis And shall give snow.
Pacharúrac Mundi factor The world’s Creator,
Pacha cámac Pachacámac Pachacámac
Viracocha Viracocha Viracocha,
Cai hinápac Ad hoc munus For this office
Churasunqui Te sufficit Has appointed thee,
Camasunqui Ac praefecit And has created thee.
Brotherston (1992, 78)

The hymn alludes to a prayer for rains addressed to four protagonists: the
princess and her brother, the coastal god Pachacámac, and the great god Viraco-
cha. The Inka Garcilaso, in his Comentarios reales ([1606–17] 1982, book 2, chap.
27) tells how he obtained this hymn composed by an Inka “poet and astrologist,”
based on the “knots” and accounts of certain annals in “threads of diverse col-
ors,” designed to encourage faith in Viracocha and his power over thunder, light-
ning, and thunderbolts, hail, snow, and rain. He mentions a comment by Blas
Valera that the verses, composed of four syllables, had a “spondaic” quality, as
if used for recording a sequence of libations. Garcilaso holds that the verses also
constituted a fable in which the “Maker put into the sky a maiden, daughter of
a king, who carried an earthenware jar full of water, to pour out whenever the
earth needed it.” If this is so, then the kipu structure could be read according to
different literary genres: hymn, narrative, and libation.
Aside from the value of these verses as Quechua literature and the matter of
deciding if the organization of this kipu does or does not constitute “writing” in
the narrow sense, we should take into account other aspects of its multigenre and
multimodal creation. What calls our attention is that the Quechua word p'uñu in
the third verse not only denotes a small jar, but also an elongated form of head,
just like those that Father Blas Valera drew in his papers to illustrate certain knots
on the same kipu. The Inka custom of shattering the heads of small children
on high mountain shrines, as a way of beseeching rain from Viracocha, is well
known. In this light, the question arises if poetry also dealt with tributary matters
through the violence of sacrifice.
The same historical document suggests that kipus expressed other kinds of
knowledge besides counting. While the most common kipus were used in a more
widespread language of daily accounting and communication, other cult kipus,
in a more controlled language, were used for safekeeping religious and caste se-
crets, and could be consulted only by the emperor, the Virgins of the Sun, priests
and philosophers (and presumably by Garcilaso and Father Blas Valera). Evidence
suggests that only the official kipu readers (kipukamayuq) could read each individ-
ual kipu, as a form of speech (parole).

Traditional Andean Forms of

T e x t u a l - T e r r i t o r i a l O r g a n i z at i o n

Andean textual polities founded in cloth developed very different kinds of

institutional support from those of alphabetic writing to assure the production
and reproduction of the territories enveloped under their charge. Under the Inka,
state ideology emphasized work for the multitude, especially agricultural pro-

duction and grazing, as ritual acts through which everyone contributed toward
the maintenance of equilibrium in the cosmos and the reproduction of society.38
Sovereignty prevailed through an ideology of hierarchical obligations between
the Inkas and their subject groups. The quasi-divine status of the Inka emperor
and Coya (Qhuya) was deployed through a language of love for their subjects,
materialized in exchanges of sufficient food and drink as recompense for tribute
in farming and herding labor and, in some cases, going to war. This formal hi-
erarchy based in gift exchange was a dominant mechanism through which state
power was enforced.39
As part of the fields of production and cultural action (in Bourdieu’s terms),
these exchanges would at times have been controlled through threats (and the
execution) of both real and symbolic violence, as new groups were incorporated
into the empire, or brutally defeated.40 The corporeality of this violence can be
sensed in the Quechua song tinyacusun collected by Guaman Poma:
Aucap umanuan upyason Let’s drink from the enemy’s skull
quironta ualcarisun Let’s put on their teeth as beads
tullunuan pinkullusun Let’s play the flute with their bones
Caranpi tinyacusun and the tambourine with their skin
taquecusun. Let’s dance in this way.
Guaman Poma, Nueva corónica ([1613] 1989, f. 314)

With Inka expansion, these institutional mechanisms of state power devel-

oped into nested levels of hierarchy. From the heart of the Inka empire in Cusco,
state power controlled a pan-Andean network that penetrated into chiefdoms, vil-
lages, and households. Under its aegis, the coded artifacts in stone and thread of a
woven economy were disseminated by the extensive communication systems of
roads and way stations or places for deposit of emergency supplies called tambos,
which stretched to the remotest corners of the empire. Institutional mechanisms
controlled space and time, regulating annual and other more extensive cycles ac-
cording to state administration goals.
These mechanisms also controlled literary genres and their expression, unit-
ing the whole Andean region into one single cultural center that stretched from
present-day Ecuador to northern Argentina, with far-flung outliers that reached
the depths of the Amazon basin and southern Chile.41 (We examine elsewhere
how Quechua, Aymara, and other Andean language groups, with their forms of
literary organization, were drawn into the same orbit of power relations.)42
This is why historical Aymara texts, like Quechua ones, register in many in-
stances the Inka’s greatness, a curious fact since the Inkas were present in the
Qulla nation of the Southern Andes (most of whom were Aymara speakers) dur-

ing a period as brief as fifty years. However, this praising of Inka status was un-
doubtedly a part of formal state politics which sought to erase the memories of
other groups in favor of their own and exercised strict control of public debate.
Moreover, they obliged the diffusion of the Quechua language while forbidding
the use of other languages, accelerating the incursion of Quechua for administra-
tive purposes in areas where formerly Aymara was spoken.43 At the same time,
this pan-Andean state formation recognized and encouraged other local differ-
Perhaps these Inka memories reminded Aymara speakers of their own his-
tory as a people, when the fleece of their herds bequeathed the very roots of
their literature. Brotherston suggests this in his Book of the Fourth World. Citing
Matienzo (1567), he describes how, historically, the Aymara speakers of the Qulla
nation affirmed that the Inkas took Qulla animals to found their own herds. As
one of the four quarters (or suyus) making up Tawantinsuyu (the empire under
Inka rule), the Qullas retained more rights in property over the local herds. In
this situation, there was deference toward the Qulla llamas during the Inka’s ini-
tiation and other ceremonies,44 and Aymara literary forms that lauded the herds
received state patronage.
Andean institutional ways of administering power relations must have shaped
the learning of key practices for reproducing this woven polity. By studying learn-
ing sequences we can ascertain modes of processing data and their hierarchical
relation within the same structures of power. We begin here with the Inka period
as the basis against which we can develop comparative studies of the present in-
stitutional modes of transmission of textual and numerical practices.
Garcilaso calls schools the yachay wasi or “houses of learning” of the elite un-
der the Inka state. He tells us how, in Cusco, the center of the empire, there was
a whole “school district” linked to the palace, principally for males of royal blood
and the sons of chiefs in regions under Inka domination.45 Murúa ([ca.1590]
1961, 107) describes how the course lasted four years, each year under a differ-
ent teacher (amawta, philosopher; jarawiq, poet and musician; willaq umu, priest;
warachiku, who taught the art of war; and kipukamayuq, the reader of kipus), as
the required level of reading and interpreting kipus demanded various years of
formal apprenticeship in an institutional setting.
Once trained, this chosen elite administered the woven polity of empire.
Guaman Poma ([1613] 1989, f. 358–59) details how the Inkas managed kipus at
a state level according to a division between the government (that gave orders)
and the administration (that collected accounts). The former put the empire into
order according to a vocal hierarchy measured according to its distance from the
original order given by the Inka’s voice; the latter according to a hierarchy of ac-

counting. Guaman Poma differentiates between the Inka’s secretaries (kipuq) who
“governed the whole empire with their cords,” and the accountants and treasur-
ers (also called kipuq) charged with the numerical accounts of the empire.
In the first rung of the hierarchy of the first group, a personal secretary car-
ried the “accounts of the words of the Inka,” and then a secretary of counsel car-
ried the “words of the Inka and the royal lords of Tawantinsuyu.” These were the
sons and grandsons of the great lord Waman Chawa, called Lliwyaq Puma, Apu
Puma, an indication that these were of the line of wise ones, chosen by lightning.
On a lower rung, a secretary “of the most excellent lord Viceroy, second person
of the Ynga” was selected from the sons of the “great Lords” (apukunas). Among
them there was a division of tasks between the qillqa kamayuq, “charged with
iconography,” and the killa wata kipuq, who carried the “account of the months
and years.” On a still lower rung were the “clerks of the chapter council or ca-
bildo,” “royal clerks” and “named clerks.” The latter went to the provinces with
the judges and mayors. For Guaman Poma, these men “wrote without lying,” be-
cause in the cords “they knew so much that they made me think that they were
in writing.”
Among the accountants there were, in a lower rank, the condor chawa or
greater treasurer, son of the apu, “he who carries the accounts of the people of
Tawantinsuyu, he who receives the income of the Inka.” This person counted
by combining a kipu with quinoa grains. In a lesser rank, there was “in each city
and town and village of Indians” a “greater accountant” and a “lesser accoun-
tant” who counted with a combination of kipu and counting board. For Guaman
Poma, these had such a “great skill” that “it was better than in paper and ink.”
Still today, the specialized readers of kipus and colonial texts (the so-called ti-
tle bearers) are selected people (more often than not wise ones) who acquire their
art during a long process of apprenticeship, according to the regional norms of
legitimate textual transmission, from father to son, and from guide to follower.
For the female members of the Inka royal families, there was another for-
mal institution: the aqlla wasi, “houses of the chosen ones” in which virgins were
trained in state cults, weaving, the care of the flocks charged to the cult, and the
preparation of food and drink.46 At a more regional and local level, formal ed-
ucation for the other people in the empire was of a familiar and daily “collec-
tive learning,” even though it was still evidently made up of three principal areas:
farming, crafts, and religion.47
From their textual basis in fleece, Andean textual practices were and still are
disseminated according to a predetermined logic and hierarchy. The same live
performance derived from supporting texts (whether poetry or libations based
on kipus or choreography based on weaving) guarantees the production and re-

production of their proper textual basis. It provides at the same time a mnemonic
way of reflecting on and storing knowledge, and a communication medium of
the dynamic movement of its elements from one level of the system to another.
In its maximal territorial expansion from Ecuador to Argentina, and from the
highlands to the rainforest, the different textual levels were managed under the
control of the Inka state, governed in cloth by the voice of the Inka and his great
lords, and interpreted by his scribes, accountants, and secretaries. In this sense,
woven and braided Andean texts as regional instances of writing always had—
and still have—specialists in their reading and interpretation, those from the lo-
cal textual community, versed in the communication media of the locality, who
simply transferred their abilities to the new forms of writing introduced from



“Let us,” said one of the natives whose language we could speak, but imperfectly, “build from
the trees a thing we call a ‘ship’—from the wood remaining I will show you how to make
‘paper’—on this ‘paper’ (once we set sail) I shall show you how to ‘write’ (with a charred twig
from the same tree)—and if your grandmother is with you, here’s how we suck eggs.”
Tom Raworth, Logbook

With the Spanish Conquest, the resulting contact between different texts and
forms of writing, as well as different textual, numerical, and literary practices, ini-
tiated a textual struggle of epic proportions. Since then, overlapping textual fields
have coexisted as historical strata of practices and interpretations, lapsus memoriae
in time and space between one practice or another, one interpretation or another,
amid the play of interests in reproducing one or another element of society.
The present attempt by the educational reform to disseminate alphabetic
writing in regions such as Qaqachaka “in a different way” is not entirely new,
since it logically follows a process that began in 1532. Thence the many similari-
ties, in theory and practice, discourse and ideology, between the efforts of the
colonial Church to convert Andean infidels to Christianity, and present-day at-
tempts on the part of educational reform experts and ideologues to convert il-
literate Andeans into literate citizens. It is not by chance that both attempts use
Andean languages as a means of conversion.


The other face of the coin in this historical experience is the Andean recep-
tion and interpretation of European written materials and writing practices. Even
now, the reception of present-day educational reform materials by parents, teach-
ers, and pupils alike, takes place in the context of five hundred years of reinter-
pretations of European reading and writing practices. There we witness just an-
other stage in this historical textual struggle, whose hybrid examples result from
crossing different textual practices in an ongoing play of resistance in which the
comunarios, as social actors, reinterpret foreign writing and numerical practices
according to their own criteria.

C o n t a c t a n d M u t u a l I n c o mp r e h e n s i o n

After the Conquest, many but not all of the institutional bases of former
Andean textual organization were separated from their centers of power and
brought under Spanish administrative control. Andean textual practices and lit-
eratures (textiles, tucapus or boxed weaving designs, kipus, pictography, painted
boards, oral traditions, songs, and so on) underwent European scrutiny and a
comparison with alphabetic writing. In parallel, Andean populations compared
European reading and writing with their own textual practices, adapting them in
the image of their own forms of expression. This mutual reflection between dif-
ferent textual practices resulted in a new heterogeneity in regional hybrid literary
forms and a clash around concepts of writing.1 This confrontation was of such a
nature that Antonio Cornejo Polar, in Escribir en el aire (1994, 88–89), points out
the danger of perceiving it “from one side only,” insisting that the object of study
should be the “marriage of contradictions” through contact, mutual feedback,
and interpretation.
In the modern commentaries from Qaqachaka that we examine in later chap-
ters, writing is both feared and seen as powerful. Moreover, it is common to hear
that writing did not arrive with the Spaniards, and some sources attribute its dis-
covery to the Inkas. This view was common in the colony. For example, the Span-
ish historian Montesinos ([1643] 1882) states (supposedly with the influence of
Blas Valera) that the Inkas knew writing (qillqa) but forbade its use under pain of
death; they even burnt the wise men (amawta) who practiced it, for the very prob-
lems that writing could bring about. Likewise, the mestizo chronicler Santa Cruz
Pachacuti Yamqui, whose own hybrid text, the Relación de antigüedades deste reyno
del Peru (ca. 1613) moves between visual, nonverbal, ritual, and written codes,
gives us to understand that the Inkas knew writing before the Spanish arrival; he
mentions certain “pre-Hispanic chapters” and a mysterious “large book” brought
from Cusco by a messenger.2

For many chroniclers, two distinct perspectives (the Andean and the Euro-
pean) concerning the nature and status of writing constitute the central facet of
that fatal event at Cajamarca in 1532. Cornejo Polar (1994, 26) identifies the fa-
mous dialogue between Father Valverde and the Inka Atawallpa before his execu-
tion as the “point zero” of the colonial encounter, for all the consequences that
first mutual misunderstanding would bring about. The misunderstanding stems
from Atawallpa’s stance toward the Holy Scriptures (or some breviary) that the
Dominican priest handed him, and his remark concerning the book that Friar
Vicente held in his faith (and that justified the Spanish usurpation of the Inka’s
lands), “Nor does it speak to me!”
Even so, Cornejo Polar concedes that the interpretation of the Inka’s mis-
understanding is not at all easy, and that there are clear differences of opinion
among those who commented on the same event. Notably, two of the Indian
chroniclers, Titu Cusi and Santa Cruz Pachacuti, do not make an issue of this
episode, while even Garcilaso gives more importance to the Spanish plunder that
In general, the encounter at Cajamarca is interpreted as evidence of two
worlds textualized differently: the one of the Western Holy Scriptures, the other
of Inka divination and orality.3 The same interpretation has been applied to the
dramas of the Death of Atawallpa (Atawallpap wayñuynin) that appeared a couple
of decades later in 1555 (according to evidence gathered by Bartolomé Arzans
de Orsúa y Vela in 1705). Narrating the same encounter in Cajamarca, these dra-
mas emphasize the contents of a letter to the King of Spain (or from Pizarro to
Atawallpa), giving prominence to the position of the Spanish Crown instead of
that of the Church. Such dramas are still enacted in towns of central Peru and in
Oruro Department in Bolivia, not far from Qaqachaka.
However, at the very heart of Atawallpa’s misunderstanding about Euro-
pean writing lie clues to an Andean perception of European writing in terms of
its own scriptural practices. (This would also explain the attitude of the Andean
chroniclers who tend to disregard the whole episode.) According to the version of
the drama recorded (or considerably reworked) by Jesús Lara (1957, 101–5), first
the Inkas compare the European sheets of paper in the religious book handed
to Atawallpa by Father Valverde with corn husks (chala blanca), better known to
them. Then, the Inka religious leaders interpret the written letters on this paper
as the tracks on the ground made by insects or birds, or the outlines of animals
(deer or llama) grazing on the land. Or else they are seen as sets of rounded ele-
ments and winding pathways. That is, the Quechua-speaking playwright would
have us think that, for the Inkas, paper has to do with the earth, while letters are
tracks or other familiar forms that appear there, in the same way that textiles

have plain pampa or earth background areas on which the figures of birds and
animals are foregrounded, and just as kipus have rounded knots like “eyes” and
long “pathway”like pendant threads.
The indigenous reaction in the face of other European practices suggests simi-
lar kinds of reconceptualization. We know that Inka kipus were substituted by the
European rosary. In about 1613, Guaman Poma remarked on the catechization of
kipu use by the Church by way of rosaries for the confession of sins, and how the
kipus still handled by scribes and secretaries of that time had become “most Chris-
tian” (f. 358–59, f. 615–16, f. 635–36). However, his accompanying drawings indi-
cate the vitality of this change for Andean peoples. As Dransart (2002) points out,
the regional form of rosary—its tens of beads interspersed with miniature crosses
and metal vessels called “blessings” (bendición)—reproduces under a Christian
guise vital aspects of Andean liturgy, above all the idea that liquid flow in those li-
bations results in the fertility of the flocks and produce of the land.4

The Introduction of Alphabetical Writing

and the Struggle over Meanings

These intercultural ways of interpreting the textual practices of the Other

continued throughout the colonial period as overlapping conceptual fields, dis-
tinct at the extremes but interpenetrating at their points of contact. At an official
level, in the centers of power, the imposition of alphabetic writing (and Western
numeration) gave impulse to a new form of reading in a European sense for the
bureaucratic elite. The production of an Andean colonial literature was propelled
from these centers for the initial purposes of Hispanic-Andean contact, and later
for greater penetration and domination. At a local level, the previous textual and
numerical practices continued. At intermediate levels, new combinations of hy-
brid and reinterpreted practices emerged, according to the criteria of each group
of protagonists and of a new group of mestizo social actors versed in both His-
panic and Andean practices.
The official process started in 1570 with the Toledan reforms to the state, fol-
lowed by ecclesiastical reforms initiated in the Third Lima Council (1582–1584).
After the sacking of Vilcabamba and the execution of Tupaq Amaru I in 1572,
there was a more systematic attempt to erase what remained of Andean writ-
ing. The Ordenanzas of the Viceroy Toledo for the city of La Plata (now Sucre,
Chuquisaca Department) specifically forbade figuration in stone, ceramics, and
weaving, because of its ties to the pre-Hispanic gods. The same Ordenanzas for-
bade the use of written Quechua because of its ties to the Inka and his voice, and
the persistent claims for Inka rights to land.

Even so, many underlying structures remained the same. The new state and
ecclesiastical boundaries at regional and local levels, especially in the Southern
Andes, tended to be superimposed over those of the former Aymara (and Inka)
provinces, chiefdoms, and federations (Lupaqa, Charkas, Larecajas, Karankas,
Qhara Qhara, Killakas-Asanaqi, and Pakajaqi). These conglomerates of multi-
lingual nations were still not considered as “Quechuas” or “Aymaras” but rather
as different social groups with their own histories and sociocultural and textual
practices, which continued in adapted forms.5
Under the Toledan form of indirect rule, these provinces continued to gov-
ern themselves under their own chiefs or caciques for the purpose of demanding
tribute. The main change was the massive reinstatement of indigenous popula-
tions in new centers—the reducciones, built around a central plaza, where the pax
toledana could be appreciated all the more for facilitating the collection of taxes
and the processes of indoctrination under local administration. These new reduc-
tion towns became the focal points of a reorganized religious life constructed
around the annual feast cycles of the Catholic saints instead of the older Inka
gods. There, mestizo lords (caciques) became masters of the new forms.
In the case of Qaqachaka, the social memories of being formerly a part of
Qhara Qhara federation, and later of the reduction town and head of doctrine
(cabeza de doctrina) of Condo Condo, now in Killakas-Asanaqi federation, are still
very much alive in the extensive “pathways of memory” or memorized libation
pathways of Andean feasts, whose Aymara name (thakhi) has common roots
with the Quechua (taki). Even today in Qaqachaka, as in other historical annexes
of Condo Condo, these pathways of memory, with their spondaic rhythms, are
an important structuring mechanism for the song verses they often precede.6
Under the new colonial administration, Andean languages (Aymara, Que-
chua, Pukina, and others) were converted into fiscal categories for the reckoning
of taxes and linguistic categories for the purpose of religious conversion. This
is why the first written documents in these languages are ecclesiastical works to
serve conversion and indoctrination.
It is not by chance that the Church has been the most interested party in using
Andean languages as a means of conversion, whereas the state (and Crown) pre-
ferred the direct transition toward Spanish. Even after having changed the means
of communication toward alphabetical writing on paper, the Council of Trent
(1563) and several papal bulls charged the Spanish Crown with the conversion of
the natives, stipulating that the holy sacraments should be explained to the Indi-
ans in their native language. To facilitate this process, Aymara and Quechua were
designated in the latter part of the sixteenth century as “religious languages” (len-
guas de religión), with the Church directly responsible for their territorial expan-

sion to encompass new groups of speakers, and the widening of their functions
to include written materials, rituals, and religious sermons.
From then on, these languages no longer constituted an obstacle to an evan-
gelizing project based on “the” Word of God.7 The preparation of the first mate-
rials for indoctrination in Andean languages, put into motion by the Third Lima
Council, was still under the direction of Father José de Acosta and his translators
(which in the case of Aymara included the mestizo priests Blas Valera, Bartolomé
de Santiago, and Francisco Carrasco, with Father Alonso de Barzana as collabo-
rator).8 The first catechisms were written in the two principal languages of Peru,
while a Doctrina cristiana y catecismos para la instrucción de los indios (1584), a Ter-
cero cathecismo, and a Confessionario (1585) were prepared in trilingual versions in
Spanish, Quechua, and Aymara. Popular pamphlets from this period are still used
today by local doctrinal teachers (doctrineros) in region such as Qaqachaka.9
The annotations to these works reveal the degree of standardization of An-
dean languages, a policy initiated as an ecclesiastical hold on whatever deviation
or doctrinal error might occur on the part of local priests and doctrineros. The
standardizing principles they used are not known exactly, although they must
have been based on the model for Spanish proposed in the Gramática de la lengua
castellana by Nebrija (1492).
This standardizing process was based on certain prejudices, that “the Indi-
ans did not have words for all that is spiritual, not vices, nor virtues, nor of the
other life or states of the same.”10 Two linguistic strategies were adopted: either
the clerics paraphrased Christian terms, or they adopted Andean terms with the
appearance (from the European point of view) of being the cultural equivalent
(Itier 1995, 324). From this point of view, the authors of the Doctrina cristiana
were imaginative, translating sense for sense instead of word for word, as was
the recommendation of the Third Lima Council, and the novices were then in-
structed in the new Christian sense (ibid., 322).
Even so, bearing in mind Andean peoples’ modes of resistance, there were
other criteria in play. In the first place, traditional beliefs would sink into obliv-
ion only if converts had the will to adopt Christianity. In the second, the attempt
to describe a theology presented as radically different in the same words as that
it was trying to displace ran the risk of installing heterodoxy.11 One result of
this first intercultural struggle was a reformulation of religious ideas and prac-
tices according to indigenous assimilative capacities, in which certain practices
and Christian terminology became Andeanized.12 Similarly, recent studies have
shown how the most intimate confessions wrung out by priests in their attempts
to extirpate idolatries reveal the voice of regional religious ideas underneath the
colonizing surface.13

The contact between different textual and communication media in this co-
lonial interface gave rise to influence in multiple directions. For example, Mar-
got Beyersdorff (1992) holds that what was at one moment considered the “oral”
tradition of Inka prayers (in the case of texts collected by Molina and Santacruz
Pachacuti) was really part of a more complex sequence in which Hispanic doctri-
nal materials were translated into Quechua, later adapted to the orality of local
ritual practices, and then collected in written forms as supposed examples of this
same oral tradition.14
Neither must we forget that there were differences of approach and opinion
in the very heart of the Church, some under grassroots influences. Studies by
Laurencich Minelli (1999b, based on colonial documents recently found in Na-
ples) suggest that the group of mestizo translators of the Third Lima Council
(Blas Valera, Barzana, Santiago, and others) formed part of a confraternity called
Nombre de Jesús, struggling to restore a neo-Inka religion in the face of the In-
quisition. If this was the case, then their role in the standardizing team was more
ambiguous, their interest in linguistic variants (and Andean texts) on the periph-
ery leading them to adopt a more flexible attitude.
These ambiguities at the heart of colonial evangelization have echoes in Qa-
qachaka. As we shall see in chapter 8, the church prayers offered at Easter still
follow the colonial models introduced during the Third Lima Council. However,
these colonial texts in Aymara now form the written support upon which a pos-
terior orality informs local people of the elements of writing, particularly the
letters (litras) and doctrine (lutrina) of ecclesiastical prayers, through which they
reinterpret their own form of Catholicism.
The fact is that throughout the colonial period it was nearly impossible to sep-
arate the learning of reading and writing from that of the faith. In the sixteenth
century, under clerical demands, an exogenous European form of schooling was
imposed on the former Andean schooling system, whether in rural communities
or in the Spanish landed estates called encomiendas, where the landed gentry, or
encomenderos, were likewise obliged to indoctrinate their vassals.
In the few elementary schools that existed, reading and writing practices were
learned on the basis of “prayer books, psalms and breviary.” Pupils were forbid-
den any other kind of text; “rather all books should be devout and good, and
should teach matters of Christian religion and good customs” (Schroeder 1994,
66). Children of other faiths were not admitted to the schools, only the “previ-
ously baptized” (christianos viejos). In this manner, colonial schooling came to be
one of the most important instruments of evangelization. “Education meant
forming the true Christian, it meant moral formation, it meant learning to pray.
The class in catechism came to be the principal method” (ibid., 79).

With all these demands for catechizing in the new faith, how did Andean pop-
ulations interpret the new schooling rituals and textual practices? It seems that
they simply interpreted the new practices according to their own previous experi-
ence. Initially, their points of reference were most probably the Inka schools for
the noble elites and sons of regional caciques and the aqlla wasi (schools for se-
lected women) that formed the prior system of Andean community education,
as well as the great ceremonial centers at state and local levels.
Only when the state and the Church became aware that attempts at conver-
sion were not obtaining the necessary degree of success did a struggle begin
between two positions regarding language: that of the Church, which favored
teaching in native languages for conversion purposes, and that of the state and
Crown which favored the transition of Indian education toward Spanish, the lan-
guage of empire.
This struggle between the linguistic policies of Church and Crown contin-
ued for a century more, when knowledge of Andean languages continued to be a
prerequisite for priests who worked in the region, in addition to being a form of
ecclesiastical control over their proselytizing programs.15 But finally both inter-
ests came to coincide, perhaps when later generations of priests felt less enthu-
siastic about learning native languages. In 1643, the Crown ordered the teaching
of Spanish and Christian doctrine to all Indian children. From then on, changes in
linguistic policy by both Church and Crown reflected their greater consciousness
of the failure of attempts at conversion.16
One source of this failure was the different interests at play in interpreting
textual meanings and the heterogenous texts to which they gave rise. Linguistic
studies underline such differences between the Andean meanings of certain key
religious terms and the Christian reformulation of their meanings by the clergy
of that time (yachachi- for “create,” llut'aq for “modeling in clay,” tuku- for the act
of “transforming into,” jucha for “sin,” supay for “devil”).17 Their observations
about colonial textual struggle are helpful in considering the present-day refor-
mulations of Andean words, coined according to the criteria of the experts in the
educational reform.
The Andean languages changed to accommodate the new forms of linguis-
tic and written domination. The growing pressure of daily contact, translation,
and bilingualism brought on an initial stage of lexical transference. Around the
middle of the seventeenth century, the written (and spoken) Andean languages
were transformed into what Cerrón Palomino, with reference to Quechua, con-
sider a “conquered and reduced” version, in which semantic changes reflect their
colonized status.18 Another transference at a semantic-religious level, pointed out
by Albó (1999), is the “Christianization” of some terms (Pachamama, the An-

dean Earth deity, became Santa Tira), while others were “demonized” (as supay
or saxra). Later still, Aymara incorporated Spanish forms of sentence subordi-
nation, as the former passed from the paratactic structures of orality (sentences
in succession without the use of conjunction) toward the periodic structures of
written texts.19
From another standpoint, the attempts to translate Andean languages and
cultures to Spanish (and vice versa) created new forms of “speech” and “writ-
ing” as “interlanguages” and “intertexts,” whether in the adapted Andean lan-
guages or the new “Andean” Spanish. Some hold that these hybrid forms, created
through syncretism and cultural assimilation, simply imitated the new colonial
norms; others consider them to be the creation of postcolonial racial mixing (mes-
tizaje).20 Still others emphasize the dynamics of textual practices in contact, of ex-
periments and cultural resistance, whereby transferences serve as compensatory
strategies, forms of intercommunication, and the appropriation of the Other.
This diversity of opinion reflects the fluidity between genres at the points
of contact. At one pole, Hispanic influences in spoken and musical genres took
place. Informal contact, accompanied by rumor and chatter, happened on the
landed estates (encomiendas and haciendas), in the mines, urban centers, and pil-
grimage routes. Barter and exchange would have included the interchange of
remedies and tales. So European genres arrived, were disseminated, and became
adapted to their new surroundings. The European tale The Eagle and the Vixen
of the Siglo de oro, for example, became Andeanized in the story of the Condor
and the Fox.21 At another level, the more formal pomp and ceremonies of church
and state demanded the introduction of new musical instruments (or the retun-
ing of traditional ones) and the learning of new genres to honor the new gods.22

T e x tual R eformulations and the

S tratification of M emory

For Andean populations, the first deciphering of the characters of the Holy
Scriptures (with their enigmatic forms of “white corn husks”), together with their
admiration for the Spanish ability to talk with the “white cloths” (parchments)
before them (according to commentaries by Titu Cusi and Guaman Poma), grad-
ually gave way to an Andean manner of “reading and writing” alphabetically. So
European writing was incorporated into Andean textual practices according to
regional criteria of interpretation.
This was partly due to the colonial techniques for teaching Christian doc-
trine in rural areas where alphabetic writing had less influence. In the attempts

to teach the new fixed and standardized texts for Mass and to instruct the faithful
in prayer, historical efforts to improve “lay literacy” transferred written liturgy to
other means of communication. Among these were the glyphs in Quechua and
Aymara that appeared engraved on rock and on hide and parchment, the graphic
basis for later recitation. As we shall see, such glyphic catechisms have been pop-
ular devices for the illiterate faithful up to the present day in rural areas such as
As didactic resources, these glyphs have certain similarities to the Testerian
texts of the Franciscans in Mexico at the beginning of the colonial period. The
wordplay in many instances suggests a phonetic reading, such as Pärssinen has
suggested for the Inka kipus. As in the Testerian examples, such glyphic systems
with their spiral and boustrophedon forms (figure 3; called in Aymara llawuraña)
probably had pre-Columbian roots. Ibarra Grasso (1953) calls these glyphs “indig-
enous Andean writing” in the sense that Andean peoples might have been incor-
porating written catechisms into their own forms of writing.
The gamut of textual practices and practitioners needed in the colonial pe-
riod for teaching (and learning) the new standardized writing gradually adapted
to local needs, giving rise to an era of mixed genres and performers. A generation
of mestizo bilinguals carried out official and unofficial tasks as “translators of An-
dean traditions,” whether as local teachers, interpreters of Church doctrine (doc-
trineros) for the rural faithful, or those petty lawyers (tinterillos) who translated
Spanish law orally for their rural clients. Indians who spoke Spanish served as
scribes (escribanos) and memorizers (memoriones) charged with remembering the
matters of recent community councils, although these often retained the older
name of “kipu guardians” (kipukamayuq). Bilingual Indians, preferred as leaders
and interpreters, were given official employment from 1552 onwards in the New
World courts (audiencias).24
The modern successors of these former petty lawyers are the “nonliterate”
experts in the teaching of Christian doctrine, versed in oral techniques (of sound
and gesture) or modeling in clay to communicate a standard writing they cannot
read alphabetically. In chapter 8, we shall see how their aural use of wordplay as
a didactic recourse for lay literacy created some of the characteristic features of
modern Andean Christianity in places such as Qaqachaka.25
The interpretation by Andean populations of these other forms of writing
had repercussions in the reformulations of their own textual practices. Thus
Western alphabetic writing was incorporated into regional practices of oratory,
poetics, and memorized recitation. Andean populations drew alphabetic writing
into other familiar domains having to do with ceremony and ritual and into for-

Fig. 3. Catechisms in the form

of boustrophedon and spiral
(Reza-lipichi from the Island
of the Sun, in Posnansky 1957,
plate 69 A.a.; Ave María from
San Lucas, Chuquisaca, in
Ibarra Grasso 1953, plate 21).

mer liturgical forms, a process that constituted a curiously disguised act of resis-
tance. Specific Andean literatures of resistance availed themselves of these addi-
tional resources based in their own textual practices.
This theme figures as the central plot in many versions of the Drama of the
Death of Atawallpa. If the colonial intention was to develop such theatrical specta-
cles to show Andean peoples the death of their king (and with him, their world),
in practice its performance was to have the opposite effect. The Inka’s rejection
of European writing in the play expressed this wider strategy of avoiding writing
in foreign terms, while reinterpreting and appropriating it according to the more
familiar textual practices to do with the production of local lands and flocks. As

forms of resistance, these dramas harked back to an alternative theology based

on the regenerative force stemming from the Inka’s germinal head and the conti-
nuity of his voice for many more generations.
Many scholars (Luís Millones, Juan Ossio, Mercedes López Baralt) interpret
these dramas as part of an Andean messianism, grounded in the authoritative
text of the extensive Inkarriy cycle of tales, elegies, textile designs, choral works,
and so forth, focused on the hoped-for renaissance of an Inka presence. Alberto
Flores Galindo, in Buscando un Inca (1986), recognized in the dramas aspects of
Andean “utopian thought,” while contemporary critics read in them a historical
debate concerning the status of writing.
More pertinently for our arguments here, Beyersdorff holds that the roots of
the Inkarriy cycle, as of the dramas themselves, derive from an Andean liturgical
form concerned with conflicts over lands and their resolution. In Historia y drama
ritual en los Andes bolivianos (1997), backed by portrayals of local histories (often
in disputed territories in what is now Oruro Department), she perceives in these
dramas a double-edged Andean reinterpretation that opens with a reformulation
of the historical demands of the Spanish requerimiento (a legislative order read
before a military battle, which authorized one nation to subdue another). More-
over, this reinterpretation resorts to the formulae of a more ancient Andean cere-
monial dialogue that concerned interethnic claims for land (their own textual ref-
erence point), when the fertility of the soil and the seeds of the next harvest were
endangered. For her, the ritual gestures and oratory fall within “reformulated an-
cient protocols” between groups that have fought for millennia over lands and
their resources, an equivalent to the challenging carbet oratory of the Tupi or the
ceremonial dialogues proffered when lowland groups meet.26
A related point of view is that of Cornejo Polar, who claims that these dra-
mas were based on pre-Hispanic customs between rival groups, especially the
forms of oratory and dance steps (tinku or wanka) carried out over the bones of
the dead (1994, 57n86). Their performance was a way of “dancing history” (in a
liturgical sense) instead of writing it; it was a more dynamic form, less fixed in
meaning and more open to other interpretations.
So in the first centuries of contact, not only the colonial dramas but all the ef-
forts at indoctrination in the faith by the laity through literacy constitute a field
of struggle of conflicting textual practices, each side attempting to incorporate
the textual practices of the other in a mutual cultural cannibalism. The result was
a gamut of double interpretations, textual reformulations, clashes over mean-
ings, wars over images, and mutual transferences and interferences in spoken lan-
guage, in the new hybrid texts, and in the mixed literary genres.
On the Andean side, compensatory textual strategies appropriated not only

writing but European textual practices as a whole, whether tales, prayers, dramas,
clothing, legal summons, or litigations. In the colonial dramas, Andean popula-
tions appropriated the European theater to reformulate past memories in a mil-
lenarian vision, a convenient coating on unsuspecting occasions when they could
continue with a celebration of an Inka presence. In the terms of García Canclini
(1989), these were contemporary rites of entering and leaving the written form
as a temporary convenience. However, they had an unexpected result. These re-
formulations of writing, of Christian liturgy and other European texts, left open
the possibility for Andean populations to continue appealing to the Inka’s voice
as the supreme source of power.

T h e T e x t u a l F o u n d at i o n s o f t h e N at i o n

With the founding of the Latin American republics such as Bolivia in the
nineteenth century, the process of forging unified and coherent nations from
what Aurolyn Luykx (1999, 18) calls a “diverse mélange of frequently antagonis-
tic social groups” depended on broadening state control over the means of pro-
duction through writing, as had occurred in the colonies, to extend republican
sovereignty through systems of communication and mass education.
In Imagined Communities (1991), Anderson holds that these emerging nation-
states, as elsewhere, were founded on the state control of written documents. He
likewise traces the emergence of citizenship in nation-states to the simultaneity
of production made possible by the mechanical printing of documents. Just as
paper sheets covered with printed text came out of the printing machine, so citi-
zens emerged from the new national institutions, such as schools, at once literate
and trained in civic duties in the interests of the nation-state.
Anderson’s argument seems weak vis-à-vis Andean populations with their
own textual traditions (a point to which we shall return). Nonetheless his argu-
ment does seem to hold for the new Andean republics as expressions of criollo
identity and political interests, founded on paper (the same textual medium as
had been used in the colonial period), now amply developed through mechani-
cal printing to handle what he calls the “beginning of print capitalism.” By this
means, the elites of the new republics were able to control state bureaucracy (in
daily newspapers, textbooks, publishing houses) and the gradual dissemination of
Spanish as the official language of the nations. But just as in the colonial period,
the function of printing was reappropriated by other social actors in reworked
configurations of the former textual struggles.
With the founding of Bolivia (1825), the criollos (Spanish descendants born in
South America) incorporated texts written in native languages into their own am-

bit, but only when it helped their wider project of identity and national integration
through writing. Now the performance of Andean-language versions of national
works (especially anthems of national and regional identity) became common-
place in primary and high schools. In those years, the textual incorporation of An-
dean peoples into the nation was based on the political, nationalist, and religious
requirements of criollo sovereignty, directed at their conduct as good citizens.
But by the late nineteenth century, other voices came to be included in a new
configuration of the struggle for land. While in the first centuries after 1532 the
land question had been tied to the rights of the Inka and the confrontation be-
tween different textual practices as these impinged on land matters, now in the
Republican era the same matter became disengaged from its origins to become
polarized between different class interests.
A primary wave (ca. 1890–1920) of interest in the “problem of lands” and the
emerging powers of print was expressed by the indigenistas, members of a move-
ment defending indigenous culture, mainly writers, novelists, and politicans.
These forged an alliance between an anti-oligarchic movement (against both land-
owners and clergy) and a socialist plan founded on the basis of a peasant econ-
omy (the so-called “land problem”), influenced by José Carlos Mariátegui, direc-
tor of the periodical Amauta and author of the pioneering Siete ensayos (1928).
With the indigenistas emerged a written expression of rural life and its oral tradi-
tion as part of the modernist literary agenda, evolving new genres (above all the
novel) suited to the urban bourgeoisie, an ascending literary class with provincial
roots, and the transformation of a nationalist ideology from criollo values (cen-
tered in Hispanic languages and culture) to mestizo ones (with more acceptance
of indigenous cultures and their values).27 The “problem of lands” shaped this
new search to rethink a mestizo nation through an emerging urban textuality
and accompanying written expressions in pamphlets and manifestos.
A secondary wave (ca. 1874–1930) of common interests about land ownership
and the function of printing coincided around a secularizing of the liturgy, led by
the reformed textual community of apoderados (empowered ones, title bearers)
and local doctrineros (teachers of church doctrine). Just as the ceremonial and ju-
ridical system of the Inkas and other Andean states (through kipu, weaving, and
orality) constituted the institutional control of textual practice in rural areas, now
the administration of colonial and republican law, through the documents and
notarial records of legal discourse, formed a bureaucratic substratum with which
oral history and Andean literatures had to engage.
Now, written Spanish came to be the accepted medium for legal and public
discourse, even in cases to do with rural areas.28 The need to formulate argu-
ments to support appeals at higher levels of the colonial and republican adminis-

tration led rural populations and their representatives to adopt Spanish legal con-
ventions and concepts. As a result, many regional rebellions under indigenous
leaders (such as that of Tomás Katari in the 1780s) were decidedly legalistic in
their demands for justice.
In rural communities, this need to conform to Spanish legal conventions was
met through the formation throughout the region of a cadre of title bearers (in
Aymara titul q'ipi) defined by Rappaport (1990) as the local “textual community,”
men empowered to conserve and recite from memory the contents of colonial
title documents concerning land and its produce to interested parties in any dis-
pute (whether community members, state functionaries, or provincial lawyers).
They traveled constantly between the locality and provincial and urban centers,
conducting their affairs in both native languages and Spanish, but according to
the norms of legitimate Andean textual transmission.
Toward the end of the nineteenth century, their skills were engaged in re-
gional efforts to counteract a renewed attempt to usurp ayllu lands (under Presi-
dent Tomás Frias’s Land Expropriation Law of 1874) in another wave of large
landholding (hacienda) expansion that sought to undermine the economic basis
of indigenous communities and their collective forms of organization. In this his-
torical juncture, the title bearers (apoderados) were convened to form regional,
national, and pan-Andean alliances to defend their territory and cultural survival.

New Junctures in Continuing Textual Struggles

These same men came to guide the movement in rural areas for indigenous
schools.29 In these grassroots struggles to defend ayllu lands, the title bearers of
local textual communities worked with community members to wrest control
over reading and writing from mestizo and criollo hands and into indigenous ed-
The continuing demands of the 1874 Land Expropriation Law meant that the
struggle to reclaim ayllu lands depended on the successful handling of colonial
title documents, written mainly in Spanish. (In Qaqachaka, the primary reference
document is the Composición de tierras of 1646 under the direction of José de la
Vega Alvarado.)30 Local apoderados took advantage of the liberal discourses of
the period to incorporate rural communities into the state educational system, so
that another generation might learn reading and writing in Spanish and so better
defend their land interests. Above all, they sought liberation from the harsh tribu-
tary obligations of the period. To their way of thinking, education would con-
vert Indians into mestizos and so liberate them from the indigenous tribute that
formed 50 percent of state income until Melgarejo’s 1864–1871 government.

In the 1930s, the apoderado movement flourished under such leaders as Santos
Marka T'ula and the great title bearer of Qaqachaka, Feliciano Inka Maraza, com-
bining their struggles over land with an emerging educational struggle. Although
the Warisata ayllu school is the best known of the educational experiments of
this period, dozens of other indigenous schools emerged due to the efforts of
these men. The case of the title bearer Santos Marka T'ula is typical, since he re-
quested the establishment of rural schools together with military training for the
community men as part of the struggle to defend ayllu lands. Likewise, Feliciano
Inka Maraza founded a private school in Qaqachaka.31 In this context, the notion
of the ayllu school did not emerge suddenly out of the liberal ideas of the mes-
tizo Elizardo Pérez; it came rather from the long-standing concepts managed by
the title bearers whereby the new indigenous schools were relocated in preestab-
lished communal settings.
The results of their efforts brought about another wave of reactions on the
part of the landowners (hacendados) and conservatives of the time. Arze Aguirre
(1987) describes the harsh repression of indigenous educational leaders, teachers,
and pupils accused of communism and subversion.32 The formation in 1930 of
the communitarian organization Sociedad República del Kollasuyo by the educa-
tor Eduardo Nina Quispe (then director of indigenous schools) irritated further
the authorities, who feared his intention to reestablish a more indigenous repub-
lic within Bolivian territory (ibid., 95). Nina Quispe himself, from 1930 until his
imprisonment in 1933, also advocated before the president of the republic for an
agrarian reform that would relegitimize the original titles of communal property
that had been disregarded or discredited (ibid., 31).33
Quispe’s vital link between education and land was now taken up by other so-
cial classes. During the Chaco War, the linguistic barriers between Aymara, Que-
chua, and Guaraní troops and their monolingual criollo officers were constant
obstacles to the accomplishment of military duties, causing the unnecessary
deaths of thousands of combatants. Territorial defeat and the immense slaughter
of Bolivian troops drove politicians and popular subalterns alike to contemplate a
Hispanicizing campaign for the whole indigenous population. New political par-
ties and trade unions were founded for whom cultural and linguistic homogeni-
zation was the solution to the former fragmentation of the country. In particular,
the MNR (Movimiento Nacionalista Revolucionario or National Revolutionary
Movement) and POR (Partido Obrero Revolucionario or Revolutionary Work-
ers’ Party) debated Bolivian educational policy.34
In the political alliance of the 1940s and ’50s between peasants and the MNR,
both groups sought counterarguments to the dominant political discourse of rac-
ism and social Darwinism that underlined hacendado interests. The MNR saw in

the education of the masses a civilizing process for a backward race, which could
transform it into a more useful labor force while levying votes; this idea was
neatly implemented in their euphemism, “peasants,” for those who were previ-
ously called “Indians.”
The emerging cult of schooling embodied the concerns of the young MNR
militants, who saw in Bolivian nationalism the overcoming of existing ethnic dif-
ferences, and in education a key to personal and social advancement in the new
integrated nation. That is why, as Luykx points out (1999, chap. 1), the educa-
tional discourse of that time centered on ethnic and cultural differences in terms
of “nationality”: Andean rural people were not considered sufficiently “Bolivian.”
Now both rural people and the young MNR militants resorted to learning Span-
ish and forgetting native Andean languages. While rural people, in the first mas-
sive waves of migration from rural areas to the cities, sought in Hispanicization a
way to power, MNR party members sought a way to national integration.
The changes unleashed by these militants with the MNR-led Bolivian revo-
lution (1952), the agrarian reform (1953), and above all the first educational re-
form (1953) with its Code for Bolivian Education, introduced onto the national
political scene a third wave of social actors: a labor force of rural teachers with
strong union and class interests and a vested interest in modernity. By the 1960s,
thousands of speakers of Andean languages had entered as teachers into the state
educational system. The political climate put an end to the employment in rural
areas of the former literate mestizo teachers in favor of these rural teachers with
closer community ties and the backing of the teacher’s unions.
Now, a struggle for education replaced the former struggle for land. The title
bearers were forgotten and the new generation of rural teachers insisted that their
educational struggle coincide with the priorities of modernization worldwide. In
this political climate, they remodeled their practices on the basis of criollo mod-
ernist rhetoric and educational ideologies that reinforced their new occupations,
identities, and aspirations, in what Platt has called their socioeconomic “whiten-
ing” and their submission “to ‘the one and only future’ of progress preached from
Europe.”35 There began in Bolivia what Baptista Gumucio (1973) called “national
suicide” and Illich (1971) called “suicide by an overdose of schooling.”
The problem lay with the new core of teachers. Raised in a rural environ-
ment and bilingual in Spanish and Andean languages, they were not really literate
(in either the written or spoken registers) in Spanish or in the Andean languages.
Referring to this problem, Luykx (1991) distinguishes between “native speakers”
and “native readers” to exemplify the failure among students of teacher training
colleges to command the technical aspects of alphabetical reading and writing.36
A ready (yet short-sighted) solution to this problem was to adopt didactic

techniques that disguised this limited capacity in reading and writing on the part
of students and teachers alike. As a result, a new wave of orality that drew on
previous Andean techniques of memorization reinforced the division between
spoken and written languages. Students in the classroom simply received dicta-
tion, and then repeated the same words in their own expositions, without ever re-
ally making an effort to understand their meaning.37
Where the Warisata experiment had found ways to modify schoolroom ped-
agogy with modernization in mind, the new generation of teachers, especially
those from rural backgrounds, would favor the more familiar practices of recita-
tion and memorization. Until the 1994 educational reform, schoolroom practice
was directed at a one-way learning process, with emphasis on dictation, memo-
rization, and copying (the form rather than the contents) instead of comprehen-
sion, and the oral transmission of them all, with a traditional schoolroom layout
reinforcing this sense of regimentation and conformity.
Teacher training colleges fomented key ideas of liberal debate, playing on the
same double discourse (of traditionalism in teaching techniques yet in a class-
based modernity) to reinforce the pretensions of their student teachers to switch
social class by integrating indigenous with educational struggles. They adopted
as founding myths the new euphemism of campesino (rural farmer) for Indian,
the post–agrarian reform relationship of rural farmers with their lands (in the re-
gions of former haciendas), and the importance of the ayllu school in the founding
of indigenous education. Gradually, the whole educational system turned back to
archaic textual practices (both state and ecclesiastical), including the specifically
Andean interpretations of alphabetical reading and writing and those recitation
and memorization techniques deriving from the secularizing processes of ecclesi-
astical literacy adopted by rural populations over centuries.
As they became unionized, rural teachers entered as staff and later as direc-
tors of regional teacher training colleges, and later still as university teachers. In
this way, Andean reading and writing practices spread to every level of the for-
mal academic system. The territorial expansion of these struggles gave rise to
other changes in the educational system. Previously, oral presentations had de-
rived from the written register; now a new form of academic writing, based on
oral forms, disguised teacher’s limitations in this register.
The efficient functioning of the state educational bureaucracy could only be
maintained by developing a system of “cover-ups” to disguise the failings in read-
ing and writing by both students and teachers alike. There began the gradual insti-
tutionalization in the educational system of “presents” and “bribes” at all levels.

N e w I nstitutional M eans of P atronage : P opular E ducation

and the “ E nabling of the M asses ”

In the face of the popular tendencies of rural schoolteachers and Aymara

speakers themselves to “Andeanize” school-based literacy practices by favoring tra-
ditional practices of recitation and memorization, another exogenous and elitist
trend emerged from the 1960s onward. While rural teachers adopted Hispaniciza-
tion and the retrieval of didactic practices forged in the colonial contact between
Andean and European textual practices, this other group sought political alliances
to modernize the educational system through the inclusion of Andean languages
as a means of change. Instead of leaving writing in the hands of an elite of priests
and teachers of church doctrine and resorting to orality for teaching rural popula-
tions (as during the colonial period), they were to apply massively the teaching of
writing in Andean languages, with the support of more global interests.
In this juncture between the ecclesiastical struggles of previous centuries
and the modernizing drive of states worldwide, educational reforms flourished
throughout the continent. Andean languages were now reformulated on the ba-
sis of European texts, and ways of teaching Andean languages were adapted ac-
cording to the norms and practices of Western writing, which sought to change
notions of personhood and knowledge (as writing practices had done formerly
in Europe). Although this process had its predecessors in the first decades of the
colony, its territorial limits now encompassed state schooling. A new textual con-
sciousness, already incorporated in the everyday practices of the technicians and
pedagogical advisers of the reform, was transferred to rural areas in many of the
literacy programs under the aegis of the reform.
These two currents, the popular Andean and the European elitist, with their
fundamental differences in approach, ideology, and class interests, have gone on
transforming a struggle of centuries until the present day. The teaching profes-
sion’s criticisms of the present educational reform should be understood in the
light of these emerging conflicts between distinct textual practices at every level
of the educational system, and their echoes in the original confrontation in Caja-
From the 1980s on, international aid began to circulate in the Andean coun-
tries for cultural and linguistic rescue and revitalization, and above all for massive
literacy programs in both Spanish and in the Andean languages under the de-
mocratizing rubric of bilingual education. A new flotilla of experts (who in the
main did not speak Andean languages) came to oversee the application of those
programs and supervise the production of materials in these languages. At the
upper end of this market, the international organizations UNICEF and UNESCO

financed collections of Andean literatures. In another example, the Peruvian-

German Project in Puno “rescued” popular literature and sponsored publication
of an Aymara vocabulary in this regional variety. Meanwhile, the production of
Spanish language textbooks (Semilla, Alborada, Primeras luces, Surco) continued as
before, a result of the efforts of mestizo teachers with a great deal of experience
in the classroom.
In this context of cultural rescue, the institutional approaches to the collec-
tion of Andean literatures followed common tendencies: the disconnecting of
Andean languages from their territorial roots, the “depersonalizing” of narratives
as regards gender and age groups, and above all the “reduction” of everything
to writing. In practice, many of the collections of oral traditions were just frag-
ments, torn from their original social and cultural contexts, and hence from their
original territories.38 The techniques used in their compilation, due to the lack of
linguistic expertise of their collectors, were in many cases improvised, even when
speakers themselves transcribed versions of a tale on the basis of originals that
they attempted to “improve” by modifying (or omitting) what they considered
superfluous details. With few exceptions,39 the educational auspices of interna-
tional institutions such as UNICEF and UNESCO allowed oral traditions (story,
myth, song) to be gathered and transcribed in written registers (as had been done
in Europe) with the purpose of converting them into materials for child literacy,
as if they were nothing more than that. The present efforts of bilingual inter-
cultural education typify this global attempt to transform regional texts toward
written criteria in a narrow sense, getting them down on paper while ignoring re-
gional practices of prosody, rhythm, versification, and so on.
A particular risk with the current reform’s focus is the undermining of An-
dean languages through its approach to standardization, which does not take into
account the vernacular speech used by speakers in their communities, that is, the
regional varieties that express their cultural and linguistic identity, including those
aspects that identify the moral economy and social ecology of their world. More-
over, when educational programs are centralized, the standardized variety they
adopt is often not recognized by speakers as part of their vernacular tongue.40
The undermining of local languages and textual practices also occurs, as Muhl-
hauser (1990) argues, when “everything is reduced to writing.”

The Reform Curriculum and the Invention

of Andean Childhood

The dangers of undermining local languages, first evident in the design and
execution of the reform textbooks, or “modules,” directed at rural children in the

1990s, have continued to the present in the child-centered activities of reform in-
cursions into multimedia presentations, in ayllus such as Qaqachaka.
The reform curriculum follows the international recommendations for sit-
uated learning. However, a critical analysis reveals aspects of an elitist stance:
on the one hand, a quasi-religious attempt to depersonalize Andean criteria of
speech, personhood, knowledge, gender, and age groups, and relocate them into
the book; on the other, the wider criteria of modernization through which this
goal is accomplished. The proposals put forward under the aegis of bilingual in-
tercultural education are still molded by a preestablished criollo framework that
seeks national uniformity through alphabetic writing on paper.
This tendency works at an ideological level in the attempt by bilingual in-
tercultural education to impose one textual formation (alphabetical reading and
writing), with its associated textual polity, over another (traditional Andean re-
gional practices). In practice, the reform curriculum insists on incorporating An-
dean textual practices (weaving, song, riddles, ritual, and so on) into the criteria
of the dominant textual formation of writing on paper. (See figure 4.) In exam-
ples in the materials of communal activities such as weaving, playing, or marking
the animals, even attitudes toward body parts or human creation are molded ac-
cording to the demands of alphabetic writing and its textual basis in paper.
The primary didactic order traces a sequence in which the child is guided step
by step toward the new goal of writing the world on paper. To achieve this, the
regional practice of modeling things in clay is uprooted from its Andean context
and relocated as a preliminary stage in an evolutionary sequence that works its
way relentlessly toward alphabetic writing. In the sequence of actions prior to
writing, small children who do not yet hold a pencil must learn to model things
in clay or plasticine. The striking resemblance between this schoolroom appro-
priation of creating in clay and the creative acts of Andean gods and heroes of
modeling the world in clay, encountered in ancient Andean myths, is drawn into
the service of reading and writing.41
As a second step, the children must learn to draw the same things they mod-
eled in the first units. Let us consider three examples. Figure 5 shows a spiral
made in paper in which two children are creating “paper people.” Note the simi-
larity here between this paper spiral and the way in which both colonial and mod-
ern doctrineros would teach prayers on the basis of clay images, laid out in the
same spiral form (see again figure 3). In the original example of the prayer shown
in figure 3, the writing of the Doctrina cristiana was incorporated into the Andean
textual form of a spiral (llawrantaña) and recited orally. In the module, this logic
is inverted: the paper figures, already depersonalized, are incorporated into the
same spiral form, but at the service of reading and writing on paper. Similarly,
Fig. 4. Drawings from Jakhüwi 4 (p. 29).
ig. 5. The cover of Jakhüwi 1 with people of paper and ink.

Fig. 6. Drawings from Aru 1 (p. 26). A child on the floor, while a group of children traces
his body with pencil on paper.

figure 6 shows a boy on the floor and a group of children tracing the outline of
his body with pencil on paper. Another illustration shows a girl tracing her hand
on paper.
Each example—persons in a group, the boy’s body, or the girl’s hand—comes
to have validity in the modules only when the images are transferred to paper
by molding, tracing, drawing, or coloring them. The reform lexicon for paper-
centered classroom activities echoes the same concern with translating reality

into the new textual mediums, by modeling it in clay (lluch'iña), tracing it (ri-
xuña), transferring it from one textual medium to another ( jamuqaña), drawing it
(riwujaña), coloring it (samaña), and finally, writing it down in words (qillqaña).
In this way, the new textual ways of recreating the world on paper (instead
of expressing it in fleece) reiterate the missionary techniques from the early co-
lonial period, centered similarly in children.42 These two periods are related, too,
in the generalized undermining of Andean languages in favor of Spanish as the
dominant language. This occurs with the invention of lexicon centered in the
new writing practices, and in the standardization of Andean languages (inspired
by Nebrija) to coordinate the new reading and writing practices at a pan-Andean
Concerning lexicon, the same prejudices against the supposed “limitations”
of the Andean languages are heard as were expressed at the beginning of coloni-
zation. Where previously Andean languages “lacked” the words to express spiritu-
ality and the life hereafter of the new Christian faith, now they are seen to “lack”
words to express the demands of the new faith in writing. An inherent complicity
here between language and race interprets a language and its uses as emblematic
of race, so that languages, like races, can be considered to be in different stages
of development. These prejudices are widely held by some planners of teaching
materials, who point out the “lexical limitations,” “disorder,” and “confusion” of
original languages, which warrants greater codification. The result is a depiction
of Aymara as archaic, curiously free from any “contamination” from Spanish in
the last five hundred years.
In inventing an Aymara lexicon according to the demands of writing, and
with the paternalistic attitude of expanding the vocabulary of rural people, many
Andean concepts and meanings were overlooked in favor of reproducing in an
indigenous language the concepts of the dominant national society. The most
outstanding case of semantic reformulation is the use of panka for “book,” which
causes confusion for teachers, parents, and pupils alike.43 Panka has its own mean-
ings in both Quechua and Aymara. In Ecuadorian Quechua, the term denotes a
wrapping of corn husks, whereas in Bolivian Quechua the equivalent is p'ankha.
To add to the confusion, in certain regions of Bolivia (Qaqachaka), the same
term p'ankha applies both to corn husks and female sexual parts. Faced with this
term in classroom pragmatics, a usual retort in Qaqachaka is “Jusi Mariya pankha-
taya,” “José María, the beetle,” which anchors this new term back into the do-
main of well-known folktales. Why then did the reform experts choose the term
panka for “book,” it being most improbable that Andean people had ever used
corn husks (in the way of papyrus) for making their own texts, and when in prac-
tice the usual term is liwru, borrowed from the Spanish?

This choice seems to derive from the same colonial encounter in Cajamarca
between Father Valverde and the Inka Atawallpa, and the dialogue centered in
the Andean misunderstanding of writing as presented in the Drama of the Death
of Atawallpa (in the version collected by Lara). If this is so, then the forced adop-
tion of the term panka for “book,” instead of being based in any regional criteria,
seems to reiterate the confusion of the Inkas in the face of alphabetical writing,
and their comparison of this with “white corn husks.” The practical result of
this lexical invention and redefinition is the general consternation on the part of
Aymara-speaking rural teachers, parents, and pupils alike: “This is not our Ay-
mara, this is Reform Aymara!”
The modernist alliance of an international elite with ecclesiastical and state
support was now able to consciously mold the necessary changes in the imagery
of a more inclusive nation on paper, identifying in turn the last frontiers of for-
mer notions of personhood (as well as gender and age groups) to be crossed so
as to incorporate rural communities into the mercantile economy. Instead of the
former subsistence texts based on fleece, and subsistence production organized
according to local gendered norms, the dream of modernization (conceived from
afar) converts them into texts of consumption, to be consumed by a recently in-
vented Andean childhood.
The greater insertion of rural areas into this mercantile regime would seem
to warrant the transformation of language itself. At a grammatical level, the
modules simplify Aymara syntax, and at a lexical level, the former verbal dynam-
ics of Andean languages are effaced in favor of nominalization by adding the suf-
fix –wi. For example, the verb tantachasiñäni was used formerly to mean “Let’s
get together to talk things over.” Now reform Aymara has institutionalized tan-
tachawi as “general assembly.” The modernizing criteria underlying this process
of nominalization impose a series of transferences of features belonging to one
group of languages and textual practices (the Hispanic) upon another (the An-
dean). One result is that spoken Aymara (or Quechua) is now influenced by the
norms of the newly written register.44 This imposed transposition from everyday
speech to a more formally controlled “arche-writing within speech” is similar to
that identified by Derrida in On Grammatology precisely in relation to the genera-
tion of proper names. It would imply a link between the fragmentation of the
local means of production and checking the flow of verbal dynamics, perhaps to
allow market penetration into a more fragmented Aymara territory.
These modern processes of language change have their historical precedents
in the introduction of catechisms in Andean languages in the first decades of the
colony. Both periods give rise to distinct facets of change in textual practices. One,
at a quasi-ecclesiastical level, attempts to depersonalize and undermine regional

criteria of being and knowing, ways of voicing reality, and notions of person-
hood, gender, and age grades, transforming them according to the demands of
alphabetic writing. Another, at a more technical level, implements these transfor-
mations in the written register according to the standardizing norms currently in
vogue: script, orthography, grapholect, punctuation, syntax, nominalization, and
so on. Both facets are conjured up in the centers of power in an alliance between
an international elite (Nebrija in the past, or the experts of the present), the state
(with its teams of technocrats) and the Church, all seeking the common goal of
reconfiguring and modernizing the empire, or nowadays the nation, on paper.
The very latest stage of modernization in the region helps clarify some earlier
tendencies we have possibly underplayed. This is a phase of multimedia presenta-
tions that accompanies child-centered writing “raids” on ayllus such as Qaqachaka
led by teachers and their international allies. Where at one time children learned
cultural practice from adults in a living environment, and later they were to serve
as experimental fodder for pending societal transformations, first into Christian-
ity and then into modernity, now a decade of modernity has transformed ayllu
life enough to demand school-centered raids of oral tradition by teachers “be-

Fig. 7. Children sacrificing a sheep in a multimedia presentation (2004). Photo by Elvira Espejo Ayka.

fore the culture is destroyed forever,” with children cast in a new role of “cultural
guardians.” In some recent public performances for international consumption,
children are encouraged to enact culture alone, in child-centered dramas of ev-
eryday life—rituals, wedding ceremonies, reading the coca leaves, dancing and
singing their history, even sacrificing their own animals (see figure 7)—all decon-
textualized from their original cultural purposes.
One reaction to all this is that just as the result of the first intercultural reli-
gious encounter was the reformulation of Andean religious practices and ideas
by incorporating some Christian terms and practices, so in the present educa-
tional reform, Andean populations have incorporated the reform writing prac-
tices into their own textual practices, just as they did centuries ago. Let us turn
now to see how this was done.

T h e R o st r u m o f H e a d s

La n d, S e e ds, and L etters

The Cycles of Production and Reproduction

The indigenous problem can be identified with the problem of land.

José Carlos Mariátegui, Siete ensayos......

Despite the political and social transformations Bolivia has undergone since first emerging
as a nation, Bolivian schooling displays strong continuities with the past, both in its surface
methodologies and in its underlying aims.

Aurolyn Luykx, The Citizen Factory

Part 1 showed how the gradual textual domination of alphabetic writing oc-
curred within a series of alliances between state and ecclesiastical policies, a so-
cial order and a discursive formation. Beginning with the Spanish Conquest (and
the encounter at Cajamarca), this domination was strengthened by the forma-
tion of the nation-state and the republic, and since then been introduced mas-
sively into the school precincts. In parallel, this same process was interpreted by
Andean populations according to their own textual practices, and incorporated
into their own struggles for land. However, the textual basis of this means of in-
terpreting and incorporating schooling on the part of Andean populations needs
to be clarified. If communities did not simply absorb alphabetic writing, how did
they convert it into their own terms?
Part 2 turns for an answer to a local ethnography of Andean schooling cen-


tered in the herding hamlet of Livichuco, in the upper reaches of the major ayllu
of Qaqachaka (Abaroa Province, Oruro Department). The history of schooling
is now viewed through the more colloquial local discourse of the different social
actors found there (rural teachers, developers, and the comunarios themselves),
within the wider historical context of the regional textual polity of which Livi-
chuco forms a part.

Texts and Lands

Certain clues to the reception of writing by local communities such as Livi-

chuco can be found in the second edition of Anderson’s Imagined Communities
(1991), which includes a chapter dedicated to the textual formation of the Latin
American nations and mentions the Andes in passing. For Anderson, the insti-
tutionalization of the nation-state in Latin America was already present at the
beginning of the nineteenth century, with the semi-industrial publication of the
documents of state bureaucracy (newspapers, textbooks of the official versions
of national history, ecclesiastical texts), and the gradual development of a com-
mon educational curriculum for the nation and a common pedagogy for the
armed forces. In the case of Bolivia, however, these publications did not reach the
majority of the rural population. More than access to written texts in themselves,
it was education that constituted (in Myrdal’s terms) the fundamental element of
monopoly that generated the social and economic stratification of inequality.
We should seek something, then, in the textual imagination of Andean ru-
ral populations such as Qaqachaka that contributes toward national textual unity.
We encounter this possibility for textual unity in local notions of writing. An-
derson senses one characterization of the nation-state as an imaginary model of
space and time, bounded by the reach of vernacular writing. However, in the An-
des, it is not the presence of writing that comes to constitute the nation (in fact,
vernacular writing has hardly penetrated all the highlands even today), rather the
manner of perceiving it. This perception, more corporeal, almost totemic, consti-
tutes an expression of the more “primitive” social body of the nation. Anderson
intuits this possibility in just two lines: “Official nationalisms can best be under-
stood as a means.......for stretching the short, tight, skin of the nation over the gi-
gantic body of the empire.”1
This image of the nation as a gigantic stretched skin concords precisely with
the expression of territorial unity in the Inka empire, when weavings from the
four quarters of Tawantinsuyu were received as gifts of alliance (pelts from con-
quered peoples) in Cusco, later to be redistributed to new allies in a territorial
extension of the custom of the victorious Inka to dress in the skin of the flayed

enemy.2 So just as communities became absorbed into the Inka empire by being
enveloped in a skin of woven cloth, so they would become incorporated into the
modern nation-state by being enveloped in a skin of paper. Moreover, as we shall
see, Andean interpretations of paper as a textual element animated by ancestral
spirits permits it to replace the former woven expression of land.
In The Calligraphic State (1993), Brinkley Messick examines the same process
in Middle Eastern Yemen, centering his attention on the way in which different
textual practices, state and provincial, operate in the same territory. He draws a
distinction between local “subsistence” texts and the “mercantile” texts at state
level. The former were developed and circulated in more “closed circuits” based
on family genealogies and personal contacts with scribes, while the latter were
developed in more “open networks” of commercial interchange, and through
the professional scribes of the marketplace.
In the Andean case, the Inka state and its regional satellites were once envel-
oped in the same symbol of continuity, namely cloth. However, modern school-
ing, as the mediating space between the state center and a regional periphery,
was to become the site of overlapping textual struggles. This is why rural teach-
ers have conflicting goals, the result of two distinct historical currents: as descen-
dants of the provincial title bearers in their struggle for land through written
documents, and as functionaries of the nation-state and its innumerable sociopo-
litical institutions, backed by fiscal and textual powers and bureaucratic-patrimo-
nial authority.
For their part, community members were now required to manage written
documents and classroom textual practices in terms of their struggle for lands
and the produce from them. Communal obligations to the state concerning
schooling are molded by the same historical criteria. The Popular Participation
Law (1994) promoted decentralization throughout the country and the greater
distribution of the country’s resources based on a head tax. However, its posture,
which makes out that communities are participating for the first time in com-
munity-state relations, does not take into account either the historical or juridi-
cal background of these relations. At the same time, the law has dramatically
changed the textual basis of communities regarding their rights to land. Nowa-
days, communities are obliged to form parent-teacher committees (juntas esco-
lares), although these have existed for centuries, and to understand and handle a
new wave of legislative codes and interpretative texts, without which they might
lose their access to land. Concerning education, the comunarios are duty-bound
to understand and take part in the new educational discourses, which reject the
traditional methods of teaching and learning—recitation and memorization—in
favor of other, yet untried methods centered on reading and writing. Addition-

ally, they are obliged to accept the new core curriculum, planned far from their
boundaries and reality.
How do they fulfill their new role? The answer lies in the textual history of
community-state relations. At the basis of the textual authority proposed by both
laws (Educational Reform and Popular Participation) are state interests in the
handling of certain modes of production. Since the colonial period, state author-
ity has attempted to manipulate the farming and herding systems of rural com-
munities while integrating them through taxation into more urbanized systems
linked to national and transnational markets.
Rural teachers and many local nongovernmental organizations implement
these policies without further ado. For example, the most ambitious educational
project to date in Qaqachaka gives precedence to changing local textual prac-
tices (in favor of reading and writing), clothing (in favor of charity-donated West-
ern clothes), and local modes of production (through mini-projects that favor a
greater integration into the market).
But these modernizing tendencies—especially those related to schooling—
must be located within the longer textual history of rights to land. In passing,
both Anderson and Messick call our attention to the historical relationship be-
tween provincial schools and the state, and the nexus between the yearly entry
of children into schooling to learn the annual (or quarterly) curriculum and the
tributary system of a population head count or agricultural income handled by
the state. In the Andes, state interest from colonial times was rooted in the tithes
of agrarian production and the payments of tribute as sources of state income,
backed by a body of jurisprudence and state laws. These, in turn, defined the
limits of Hispanic domination, mediated by state laws and their textual basis in
The textual implications of the European history of tributary circulation be-
tween community schools and the state illustrate Andean variants on this theme,
whether tribute to the state in produce, labor, or a population head count. In The
Order of Things (1991), Foucault analyzes the linkage between the state’s inter-
est in the annual circulation of paperwork (and paper money) and an exchange
for rights to land to the value expressed thereon. He cites the Scottish econo-
mist John Law on the use of paper currency as a token of territorial property:
“banknotes mortgaged against lands and due to be redeemed by annual pay-
ments.......these notes will be exchanged, like minted coin, for the value printed
on them” (ibid., 182).3 Similarly the cycles of economic circulation between local-
ity and state were ordered by annual harvests and quantity of production from
the land. As Foucault points out, to make a reckoning of this kind you start with
production from the land and then calculate the wealth derived from it, directly

or indirectly. The agricultural income is then paid annually or “as is customary”

at the end of every quarter (ibid., 186).
In the Andes, as in Europe, similarities between state discourse concerning
the circulation of money and paper between localities and the center, and that
concerning school production, are expressed in terms of cycles: curriculum, tri-
mesters, and so on, all based on paperwork. Even the cyclical terminology of the
curriculum with its loaded terms (classes, forms, levels, standards) brings to mind
the state icons in play. So, just as the Inka state obtained territorial unification
(while preserving local usufruct rights) through the flow of weavings in tribute
(in the way of pelts), so the modern state achieved this through the flow of paper
(or value) in return for land.

T h e S c h o o l o f t h e B o l i v i a n S tat e

At one pole of these overlapping territories, texts, and administrative tex-

tual practices, teachers as state functionaries came to adopt the official history
of the nation deriving from the new mestizo-criollo political juncture of 1952,
and they were charged with implementing this new kind of tribute centered in
paper. From then onward, they viewed the liberation of the newly called “peas-
ants” with the Agrarian Reform Law and the first two governments of Víctor Paz
Estenssoro as an advance toward civilization. On a par with the criollo discourse
concerning the nation, they omit mention to any former period. Teachers were
now challenged by a new enemy: the “local ignorance” that could thwart the
impending wave of modernization. As “engines” of progress, their goal was the
teaching of Spanish, the language of the nation, and skills in reading and writing,
the textual practice of the nation.
But despite the teachers’ attempts to mold the people of Qaqachaka into a
technical labor force to better cultivate the local lands, a continuing demand for
cheap labor in the national economy ensured that the majority of rural youth
continued to be employed as porters and housemaids for the middle classes of
the urban centers. Justifying their failure, teachers argue that they are still outside
the political structures that might achieve these advances. They also complain
about the lack of will on the part of parents to send their children to school and
about the general lack of intelligence, which they blame on the malnutrition chil-
dren receive at home.4
The problem, from the teachers’ point of view, is the local priority of “feast-
ing and working the land” rather than sending the children to school. For their
part, the comunarios don’t complain much about scarcities in the formal state sys-
tem, and they complete their education by other means. It is a norm for boys

to complete their formal education in the barracks, and for girls to go as house-
maids to the urban centers, where some might enroll in evening institutes to
learn a trade (for example, sewing or pastry making). Both options echo what
the ayllus have always done under previous Andean states, when the young men
served as the Inka’s warriors or in the mit'a (a system of corvée labor) and the
young women served in sirwiñakuy (as concubines) or entered the houses for se-
lect women called aqlla wasi.5 In this way, comunarios absorbed the new ideas con-
cerning the official history of educational struggle, and appropriated them into
their own textual history.
Other aspects of the teachers’ attitudes toward schooling can be seen in
school rituals, whether in the civic activities, the daily parades centered on raising
the national flag, the singing of military marches and national anthems, or the
rituals held on 2 and 6 August. Until now, studies of these rituals have underlined
their function as the expression of the state apparatus. Reading and writing, the
disciplinary modes of power, and above all a generalized militarization were all
interpreted as rites of the nation, focused on the transformation of pupils into
Bolivian citizens.6
As these studies show, however, the nationalism expressed in rural communi-
ties in the context of civic activities and national days is conflictive, above all re-
garding the intermediary position of the teachers, who identify themselves with
“brother peasant” at the same time as they differentiate themselves from them on
questions of class and calling. This question of teacher identity can be heard in
civic speeches, which concisely and intelligibly express the history of the school
as a part of the progress of the nation.
Livichuco’s celebration of 2 August illustrates this selective history. The act
opened with profesor Samuel Gisbert underscoring the three “fundamental pil-
lars” of the nation: the Día del Campesino (Peasant’s Day), the Agrarian Reform
Law, and the creation of the indigenous school at Warisata, whose anniversaries
are all celebrated on 2 August. Anthems allusive to this date were then sung, fol-
lowed by a speech by profesora Florencia de Gisbert emphasizing three more key
anniversaries, all of them linked to 2 August. The first was the death, during the
Wars of Independence, of the Quechua poet Juan Wallparrimachi in 1815 (giv-
ing Independence an indigenous tinge), the second, the foundation in 1931 of
the Warisata Indigenous School (and its influence in the naming of Indian Day,
now called Peasant’s Day, by President Germán Busch), and the third, in 1953, the
Agrarian Reform decree in Cochabamba, with its famous refrain to mark the end
of huge landholdings (latifundio): “The land belongs to those who work it.”
These speeches express the social reality of Bolivian schooling, a result of

conflicting sociocultural ideologies of identity and difference, class and nation.

More interesting in both cases is the coincidence among these anniversaries (of
the Wars of Independence and Wallparrimachi’s death, peasant liberation, and
education) and the specific date of 2 August when the ayllu lands are traditionally
plowed for sowing in the new annual cycle. In the presence of local parents and
children, the teachers seem to mold the new educational struggle into the same
time frame as the former indigenous struggle for land, but in their case according
to the tributary demands of the nation.
The details of another civic event, the Anniversary of the Fatherland (El Ani-
versario de la Patria) on 6 August, confirm this, but in an unexpected manner
which veers toward a more communal interpretation. The preparations for the
anniversary that we witnessed began on the freezing night of 5 August. At seven
o’clock, we went to parade along with the children, teachers, and comunarios at
the entrance to Ventilla hamlet, in a place called Uma Jalsu, “Where the Water
Comes Out.” The cutting wind made us all tremble. As we waited at that spot,
the authority in charge went to light two sticks of dynamite in the meadow. Head-
ing the parade was a consort of panpipes (sikuri) led by a huge drum, followed by
profesor Samuel and various national flags and local banners. Then, waiting in line
with their teachers were the schoolchildren of three different age groups. Each
group carried colored paper lanterns, first green, then yellow, and finally red, ex-
pressing in the colors of the Bolivian flag the social body of the nation. Behind
this colored human serpentine of lanterns came the women with their babies on
their backs, and finally us. As I took some photos, a comunario offered me a tot of
alcohol for a toast, an excellent remedy against the cold. (See figure 8.)
Next we started marching in time to the panpipes along the old tambo route
toward the school. The air was animated every now and then by triumphant
shouts of liberty and fraternal prayerlike echoes: “Glory to Bolívar, Glory to Su-
cre, Glory to Murillo, Glory to Túpac Katari and Glory to Bartolina Sisa!” We
all sensed the emotion in what seemed much more than a simple civic duty. Af-
ter about five minutes we entered the main door of the school and began to cir-
cle toward the right (toward the sunrise, although it was still the middle of the
night). First we passed by the flagpole, and then the school rostrum, to return
once again to the main gate.
Once the parade was over, and after a short pause, the teachers and pupils
took turns, as always, with a speech by profesor Samuel about Bolivian history,
followed by the national anthem and the “Hymn to Sucre,” to the accompani-
ment of scratchy taped music. Then they offered us the place of honor on a
bench in front of the rostrum, and there began skits by the children on various
ig. 8. The parade which began at Uma Jalsu (Where the Water Comes Out), Livichuco, 5 August 1998.
ig. 9. The parade which began at Uma Jalanta (Where the Water Goes In), Livichuco, 6 August 1998.

themes—corrupt politicians, an awki awki dance, a burlesque on the bohemian

life of university students, a devil dance or diablada, Inkas, Ekeko (the Andean
god of plenty), and so on—all according to the program painstakingly drawn up
by the teachers.
The following morning, the turns commemorating the patriotic feasts con-
tinued with another parade in full daylight, this time beginning from the oppo-
site direction, in a place called Uma Jalanta, “Where the Water Goes In.” From
the name, we realized that this place was cosmologically opposed to that of the
night before, and that in the parades we were somehow helping the cosmologi-
cal cycling of water in a kind of Pachakuti or world renewal, perhaps to open the
new agricultural cycle. (See figure 9.)
A second parade was headed again by flags and banners to the accompani-
ment of panpipe music, in an Andean military jula jula. We marched along the
main tambo route again and then entered the school premises, passing the flag-
pole in the middle of the precinct. We made a pause, giving time for the different
school authorities (feast sponsors and others) to take up their positions on the of-
ficial rostrum (a platform beside the national flag and coat of arms). Once the au-
thorities were in place, we paraded past them at a very slow and ceremonial pace,
the men goose-stepping and the children in rows of three, backs straight, heads

Fig. 10. The march in front of the school rostrum, Livichuco, 6 August 1998.

high, putting all their attention in raising the left foot exaggeratedly, all in time to
the martial music. (See figure 10.)
After the parades there was a more relaxed atmosphere, with corn beer (chi-
cha), fresh bread, and a banquet of meat for all, in which the teachers and their
families took the place of honor around a table in the middle of the school pre-
cincts, served by the school mayors (alcaldes). The feast continued all night long.
While the rest of the comunarios played music and danced, the school authorities
were charged with making toasts at a certain spot in front of the rostrum.
While previous studies insist that this civic act has all the pomp of a national
ritual, they fail to explain the details of these rituals: the curious conjunction of
opposed toponyms for the places where the parades begin each day, the mean-
ing of the honored ritual sites in the school precincts (flagpole and rostrum), the
nature of the toasts and, above all, the ambiguous position of the teachers who
seem to operate not just as functionaries of the state bureaucracy, but as a vi-
tal part of the production of communal lands. We shall examine these details in
chapter 6, but first let us explore the communal school as a mediating institution
in the historical relationship between texts and lands.

T h e C o mm u n i t y S c h o o l

In contrast to the official history of the nation perpetrated by the teachers,

we saw how Andean jurisprudence, drawing on its own historical memories of
textual authority, reconstructed the school in the community to serve as the bul-
wark for the defense of lands and communal values, planned by its leaders in con-
junction with the men empowered with the task of document keeping—the title
bearers or apoderados. To this end, the comunarios adapted their communal con-
cepts and practices to take into account the new school-centered organizations,
which became part of the community’s own patrimony. Communities controlled
the schools, and now they demanded even more participation in their supervi-
sion and fiscal matters, whether of the school premises or the annual program
of administering teaching staff. Besides this, the school played a vital part in com-
munal ritual life.7 The former territorial struggles were now transformed into an
educational struggle centered on the school, in a new alliance of convenience be-
tween community members, title bearers, and rural teachers.
In this context, the community school came to mediate the textual discourses
of comunarios concerning their own modes of production (farming and herding)
within the framework of a former woven polity. There, the accumulated strata of
texts and textual practices echo the depth of memory on which comunarios draw
to act out in practice (or comment theoretically on) their relations to specific An-

dean states. Ask the comunarios about their obligations toward state schooling and
they will answer with characteristic appeals to memory, considering their histori-
cal obligations to schooling and the authoritative bases that underlie their present-
day school duties.
This stratified interpretive practice echoes the dual history of community-
state relations. Official textual history in Bolivia locates the origins of its civic acts
in the formation of the republic, and the origins of the present-day Educational
Reform and Popular Participation laws in the context of modernization (under
its liberal guise), promoted by the policies of the MNR party (Movimiento Nacio-
nalista Revolucionario) in 1952 and 1994. But in comparison, communal textual
history locates the origins of community-state relations (whether through tribute
due to the state, communal obligations, attitudes toward school, or reading and
writing) in the much longer trajectory of former Andean states. In their enact-
ment of many extant textual practices linked to the perpetual struggle for lands
and tempered by these memories, comunarios perceive the school as “communal
patrimony,” not simply that of the present-day liberal state.8 From the perspective
of communal textual history, the obligations of local school authorities are rooted
in and authorized by communal obligations to former Andean states, so they still
appeal to woven polity. Here, schooling is integrated into the former administra-
tive apparatus of tribute, acted out by duty holders in the communal obligations
to the state that each original landholder (originario) must accomplish in turn.
The historical roots of the hierarchical community-state relation date back
to Inka expansion, its dominant mode being that of victor to vanquished. Given
the Qaqachaka reputation as “warriors of the Inka” (from the upper moiety), it is
not surprising that they locate schooling in the hierarchy of warfare and territo-
rial conquest that accompanied their coming under the yoke of the Inka state ap-
paratus (when they formed part of the Qharaqhara federation).9 In practice, their
immediate charge was tilling the lands of the Inka and tending his flocks, as rec-
ompense for their incorporation into the superior political domain.
In the following centuries, the colonial system of tax tribute to the King of
Spain replicated this former system. As Platt (1987a) points out, this recognized
the rights of possession of ethnic groups to their lands as an “ancient socializa-
tion of the enemy,” through their political incorporation into the new state appa-
ratus under the charge of the victorious ethnic lord or mallku:
For the Aymaras, the services they offered the King had to be reflected in a generous flow
but in the inverse sense: in receiving tribute (tasa), the King—just as the Inka before him—
recognized the rights of possession of Aymara ethnic groups to the lands cultivated by
their ancestors. The primary position of cultivated land in the concept of tasa reminds
us of the terms of the ancient “socialization” of the enemy, through his political “incor-

poration” under the charge of the victorious mallku: in fact, the word tasa was employed
to mean not only monetary tribute but also the land whose possession was guaranteed
through the payment of tribute. (1987a, 114)

For the colonial state apparatus (as formerly for the Inka state), tribute was
a “duty” (jucha) toward the functioning of the wider system, on a par with the
communal obligations of sponsoring feasts, standing their turn as authorities, or
fulfilling an act of vengeance in war.10 In the present-day language of the region,
these obligations are still experienced as “bearing a sin” (jucha q'ipiña) or “assum-
ing a debt” (juch yanaña). So a comunario might complain “I’ve been put into this
duty” (jucharuw wayuntxituxa), since he must forcibly fulfill his obligation and
“any demand whatever might befall him.”
For the comunarios, the present system of duty holders (cargos), with its shades
of tribute, emerged more immediately during the colonial period, with the intro-
duction of Western writing, along with ecclesiastical teachings with regard to sin
and Christian doctrine. More specifically, the comunarios of Qaqachaka, like those
of northern Potosí, relate the historical introduction of “passing turns” (turno
pasaña) with the Spaniards who compelled them “to carry large bundles to the
point of making them bend over.”
Apart from the origins of such burdensome duties in the colonial period, the
comunarios have another explanation for the origins of the duties they performed
more willingly under former Andean states. Here, the duty holders (husband and
wife) represent the victorious Inka and his Coya, exercising sovereignty over their
subjects in a common realm, according to differently constituted laws and sym-
bols of authority. As Platt points out, passing turns has to do with inter-Andean
struggles in which the killing of an enemy gave the victorious mallku (or Inka)
the right to appropriate the lands of the defeated. This regional logic also links
victory in war with the burden of sponsoring a feast, as a celebration centered on
lands and their produce.
In the past, feast sponsors in the region held in their possession a “head” (as a
seed that would germinate), but nowadays they carry nothing more than a deco-
rated woven coca bag (wallqipu) hung with tassels. It is customary to carry this
small coca bag slung from the shoulder, and it is compared to the trophy heads of
former times. In ritual language, the expression jucha p'iqi (head carrier) is used
for the duty, or cargo, of passing a feast.11 Moreover, these head bags command
the authority of directing the duty holders that carry them. According to Don
Domingo Jiménez, these head bags “usually talk” to their owners on Epiphany
(Día de Reyes, 6 January), the very day when the duty holders are changed annu-
ally in the ayllus.
The equivalent for a female duty holder is her coca cloth (istalla), which has

the power to “make everything speak” through its textile mouth. Don Domingo
explained how the link between these weavings and the enemies killed in battle
bestows upon them their power of speech, and of instructing the entering ayllu
authorities each year on the duties that will guarantee their ongoing rights to
land. Instead of appealing to written authority and documentation as the prece-
dence for justifying ongoing and customary duties, the same function is attrib-
uted to weaving.
Just as the comunarios differ from the modern state in their interpretations
of the origins and tasks of duty holders, so they have other interpretations of
the origins of schooling. Whereas the teachers, as state functionaries, appeal to a
new “enlightenment” brought about by modernization (post-1952), the comuna-
rios understand schooling within regional ideas concerning land and production,
interpreted in a myth about Inka origins centered on the birth of the sun and its
expression through the textual practices of the place. This imagery emerges in co-
munario descriptions of school origins, siting, and construction, and of the school
premises as the locus for fulfilling ongoing Inkaic rites as recompense for land.
Sometimes when we asked the rural teachers and comunarios about the sit-
ing of the school, they forgot the supernationalist speeches about the freedom
unleashed by the Agrarian Reform, and appealed to Andean narratives. Based
on local oral history, these locate the founding of school sites long before the
governments of Paz Estenssoro, in the liminal moment between the time of the
Chullpas and the new Inka time. This kind of origin myth is very widespread,
as shown by the example recorded near Cusco by Ortíz Rescaniere (1973). It is
equally common to hear that the construction of the outer walls of the school
precincts, even of the classrooms themselves, were made with the “Inka’s stones.”
Such is the case in Livichuco, in the pueblo of Qaqachaka, and in some commu-
nities of Carangas (all in Oruro Department).12
When the sun rises, so the school rises. In a variant of the well-known tale
of the great Inka boundary marker Juana Doña Ana and her circuit of the Qa-
qachaka boundaries when the ayllu was born, the storyteller Elvira Espejo Ayka
told us how Doña Ana, at the moment when the sun rose, looked over the exten-
sive ayllu lands under her charge and pointed with her hand, saying: “This will be
such and such a school.......and that will be such another school.”13 This local oral
history of schooling is not unique to Qaqachaka. In other regions too, the siting
of schools took place in enchanted and sometimes bewitched places. Doña Felici-
dad Lara, a retired teacher, tells of her experience in the Uruchi school district of
Chakarani (Salinas), where “the school was next to the church” and so had been
built over the graves of that place: “It was said that in that spot two or three rural
teachers had died.......because that place was malignant........It may be, but you

could take holy water or something of the sort there with the priest. I don’t know
what would happen, but they say it’s dangerous to go to that little school.”
In other cases, the land on which the school is sited is considered to be ex-
tremely productive; as Don Domingo Jiménez comments, “the school is forever”:
“We say: ‘Let’s seek publications, the school, and irukasyuna.’ We must seek well
for this, that’s why you must determine a good site for the school. Publication,
education, seek the right site.” By translating “education” as “irukasyuna,” Don
Domingo makes a deliberate pun, as iru is a kind of coarse wild grass on which
the flocks feed, especially llamas. For him, the historic nature of these places as
“lands of the Inka” makes them produce well whether it is with hay for the lla-
mas, or publications for the pupils.
As a part of communal obligations, each comunario must also hand over a
plot of his own land to the school, if called upon to do so. Don Domingo nar-
rates: “There was a plot for the school and they’d sown wheat.......two elderly
people.......and they’d cultivated it in spite of having been notified. Later, the di-
rector hit them until he gave them blooded noses. ‘Damn it! They should have let
it be, because it was notified!’ Then the wheat was left enclosed by the pupils, and
later they used it........The children ate it after it had been turned into bread in Au-
gust. Now that land belongs to the school.”
The question of land, education, and produce from the land go together. In
the past, if a comunario did not honor his obligations regarding school lands (called
uraqita: “land obligations”), the threat came from the whites (q'ara) in town who
had the power to imprison him. Nowadays, the obligations come from the local
school board.
Aside from their origins in the lands and stones of the Inka, in ancestral
graves, the school site has ties with other archaic places. These serve as historical
points of reference that ground present-day land obligations in the ritual obei-
sance of former Andean states, as the significant institutional basis for ongoing
textual authority.
One of these is the ritual site called ch'isiraya, the place of entry and leave-
taking for the ritual matters of each ayllu, where the ancestors would bury their
own. The central school of Qaqachaka is built near the cemetery (and an Inkaic
burial place), a short distance from the ch'isiraya. With the ritual obligations of
the place transferred to the school, the ch'isiraya now serves as the primary focus
for the rites of entering the ayllu, whether for the school parades or the arrival of
a new teacher. According to Elvira Espejo, the ch'isiraya has to be remembered in
the toasts for the school, by women as well as men.
Whether teacher or comunario, you must always enter or leave by way of the
ch'isiraya to perform whatever activity, and then you are considered a “juridical

person” (jaqi) with ties to the land. That is why the toasts to this place must grad-
ually inscribe in words any teacher’s route into the ayllu territory, including the
resting places, the animal that brought him (or her), and so on. After tracing his
entry into the territory, only then are toasts made “for his hand, his eyes,” and fi-
nally “for the children.”
Other ongoing institutional ties between the community and the state in Livi-
chuco hamlet were focused on the building of the former branch school over a
tambo (way station) premises, a stopping place with kitchens and stores of emer-
gency supplies that formed part of a system of ancient trading and communica-
tion routes, with Inka foundations. We examine in detail comunario memories of
the tambo in River of Fleece, River of Song (2001, chap. 11). What interests us here is
the connection comunarios make between the history of the tambo and ayllu edu-
cational history, mediated by the Inka roots they have in common.
Elvira Espejo heard stories from her grandmother, Doña Gregoria Mamani
(of Taqawa hamlet), telling how tambo history and school history “went to-
gether.” In the past, the ayllu authorities rode into the tambo on mules, horses, or
tame donkeys, “just as the children would arrive at school in later years.” As such,
the tambo was the place where they decided to found the school.
While the authorities labored there, they thought out how schools could be
introduced into the ayllu, and when the postal service arrived, they announced
that they could read and wanted a teacher, and went with the postal carriage to
the old Bolivian capital, Sucre, to obtain one. Just as the main ayllu had been con-
structed with a central pueblo and plaza, followed by more distant hamlets, so
they started to build the central school and then later the branch schools in the
surrounding ayllus and hamlets.
Elvira heard tell how all young people from ten years onward, men and
women, single and married, gradually learned to read. “They say that they paid
the teacher each month with a sheep, an arroba (25 lbs.) of freeze-dried potatoes
called chuño, and the same quantity of fresh potatoes. They also paid him money,
or whatever else they had.” Once schools were established in all the ayllus, other
facets of life changed. The patronal feasts of August were introduced into the
hamlets, and there were new norms for the school uniform, an overall made of
white homespun: “They say that pupils who entered without their woolen caps
were considered ‘dunces.’ And the girls entered wearing hats and carrying man-
tles. If they carried a folder, it was badly looked upon, and they’d say ‘Is she a
tomboy? Is that why she’s carrying a knitted folder?’” After the classrooms were
built, they were furnished, initially with chairs and tables made of local materials.
If children did not go to school, they had to pay a fine of one sheep.
Older people make a connection between what they formerly learned in the

tambos and what is now learned at school. For instance, Don Donato Inka of Livi-
chuco recalls his turn as postilion (mule driver) in the tambo there, and how the
ritual practice of making toasts to the tambo route was a way of teaching peo-
ple—“Even the mule used to study in the tambo.” According to Don Donato, the
various ritual sites were recorded in these toasts, above all the stones with special
names. In this way, the local tambo itself formed a nexus of memory for recording
all the elements of the tambo system, and its part in another important colonial
obligation to land, namely to serve the mit'a (system of corvée labor) in Potosí:
There was a special place prepared where you toasted and offered things to Santa Moni-
quita. It was a lengthy toast, with corn beer, alcohol, and anything else. There, sheep heads
and other things were offered. There were those kinds of places for making libations.
It was also to make toasts to the mit'ayo of Potosí. All the resting places were remem-
bered, and there were rounds of names and you had to toast them. The road was toasted
particularly well. As you make the toasts, so you arrived at the destination without any
mishap. But if it was not toasted well, then anything might happen, like falling off the
mule, or else the mule could die.

Even though the tambo fell into disuse after 1952, the use of its ritual sites con-
tinued in the new context of the branch school, for example those of the Guard-
ian Mountain. We shall see in the following chapters how these archaic tambo
rituals and ritual sites have been transferred to the school premises.
These historical memories about community-state relations shape the way
comunarios implement the new laws today. On the first day of a workshop held in
Livichuco (on 2 May 1998) we touched on the matter of community-state obliga-
tions and the changes wrought since the introduction of the Popular Participa-
tion and Educational Reform laws. While all of the comunarios agreed that com-
munal obligations were based on land rights, they feared any impending changes,
and there were marked differences about how to deal with such changes. Those
of the pueblo and the lower ayllus, more conservative in their opinions, still recog-
nize their obligations to the state in recompense for their rights to land. Whereas
those of ayllu Arriba (of which Livichuco forms a part), under the influence of
the Inka Maraza family with their long trajectory as title bearers in understand-
ing and debating ayllu matters in a national context, criticize past obligations in
favor of more equal relations.
Livichuco, as the former tambo of Qaqachaka, plays a key role in this de-
bate. Present obligations of comunarios toward teachers and pupils hark back to
their tambo obligations. In the past, each of the six minor ayllus of Qaqachaka
was obliged to serve turns as postilions (mule drivers) in the Livichuco tambo.
Since the 1950s, these turns have been transferred to the central school in the

pueblo, where comunarios go to serve the teachers in their daily chores, and serve
up food and drink in the national and school feasts. In addition they must prepare
breakfasts of bread and hot sweet drinks each day for the pupils, duties that paral-
lel the turns of the postilions in the former tambo to provide for the ayllu, when
they traveled first for sugar, then for flour, then to the Yungas to bring coca, “and
when you arrived in the village, you had to share all this out.”14
Tambo obligations also shaped the relations of gender and age groups in the
place, just as the school does now. According to Don Donato, more men (the city
upstarts or mozos) used to arrive at the tambo, and now the majority of the teach-
ers are mozos (in the popular parlance of the place). In the meantime, the women
were in the kitchen, called “the house of fire,” “cooking for the whites (q'aras).”
The introduction of the educational reform has compelled comunarios to con-
sider if their duties serving as postilions are still pertinent. In this context, Don
Santiago commented on his rights to land in the past and now. As the son of Don
Donato, he has the critical perspective of a family of intellectuals and title bear-
ers. He admits that the previous obligations were different, but that the Spanish
abused them in the “time of slavery,” biasing them toward working just for the
tenancy of the land. For Don Santiago, fulfilling these duties now has more to do
with “the right to speak,” and its political and juridical implications for change,
than rights to land.
His father agrees with this point of view and rejects the opinion of those
from the lower ayllus who are “too traditional, allowing slavery (pongeaje) to go
on in the school, even now.” Both mentioned how teachers had taken advantage,
making the school authorities work “like servants” carrying their bundles and
children when they arrived at the school. For them, “people don’t know their
rights (derechos).” Don Donato’s hope is that, with the educational reform, the
old servile system will vanish in favor of fairer relations. Don Santiago and his fa-
ther, aware of their own past and of the processes of change in other places, are
at the forefront of the current demands for change in the relations among com-
munity, school, and state.
Although the changes in 1952 were first taken up in the new educational
struggles of rural teachers, over time other protagonists, namely the title bearers
and the comunarios themselves, are gradually coming to accept the terms of this
struggle, even though it is still interpreted in terms of their previous struggle for
land. They even recognize the textual changes between these struggles, in which
alphabetic schooling for their children now substitutes for the tributary obliga-
tions that guaranteed their previous rights to land.
The consequences of land redistribution after the Agrarian Reform of 1953,
compounded by the accelerating demographic growth in recent decades in the
free ayllus of Oruro and northern Potosí that provides cannon fodder for the

constant land wars, have demanded this change in attitude.15 So, paradoxically,
although the origins of indigenous education in the twentieth century were
founded on the struggle to reclaim communal lands, in recent decades the re-
duction in the size of land holdings (minifundio) with the agrarian reform, and
the general land shortage with demographic growth, have compelled many rural
parents to opt for educating their children to give them an alternative to “scratch-
ing a living from the soil,” as they did. Comments by Don Domingo Jiménez,
from a valley community in northern Potosí, illustrate how the comunarios inter-
nalized this new attitude, particularly in the case of boys:
Those without much land are the ones who want their children to study. There’s not
enough land for everyone. When there are many children, the land is parceled off in small
plots. If they go off to study well, then they’ll not want this inheritance of land, or cows,
or agricultural tools. They’ll die out, even the cows’ll die off and be eaten, they’ll simply
vanish. But writing, this is an everlasting inheritance. By knowing this, they’ll live well
wherever they are, with good food and drink.

Schooling encourages the process of deterritorialization and the ever-in-

creasing emigration from communities toward the cities, in spite of the fact that
“some of us don’t want them to go off elsewhere.” For girls it is the same. For-
merly they had their dowry in animals, chattels, and weavings, but now many
prefer to sell off everything in favor of studying. This happened with Don Do-
mingo’s granddaughter: “‘I don’t want my inheritance,’ she says. She’s gone to
her father’s place. He told her he’ll give her a cow. ‘You can sell it for me.......I
don’t want it, you butcher it and eat it, I don’t want it. That’s why you’ve made
me study. You’ve already given me an inheritance. I can live wherever I go with
that,’ she told him.”
For those of us who study and write, the ties between writing practice and
territory have perhaps gone unnoticed, as has their influence in the massive out-
migration from rural communities (and now the immense diaspora worldwide).
In the recent past of communities like Livichuco this link was better understood;
it provided the reference point for the imagery of textile designs and the iconog-
raphy on stones, glyphs, pottery, or wood. Now, when local land rights are being
wrested away with even greater speed, many comunarios experience this process
in terms of the conversion of their former lands into paper artifacts (laws, certifi-
cates, degrees). What were once seeds become letters to be strewn around, and
their very children are cast like seeds toward the urban centers, all according to
the criteria of the new writing practices that served the formation of the nation-
states of past centuries. The new electronic textual practices are accelerating this
diaspora, and we shall resort to reworked agricultural metaphors to explain it.

C yc l e s o f M etamor ph os is
The Children as Enemies

..... and war songs are often used as lullabies.

Eduardo Viveiros de Castro, From the Enemy’s Point of View

In that year, the potatoes were wonderful, just like people’s heads!

Elvira Espejo, 15 January 1999

Within the nexus of community-state relations, Andean childhood hinges on the

ambiguous role of children in the reproduction of two textual polities: that of the
state, through the textual practices of schooling, and that of the ayllu, through
the textual practices of the community. Once this ambiguity is understood, then
we can approach an Andean view of “interculturality.”
In part, the differential role of children in both textual polities derives from
their function in overlapping tributary systems. In this tributary context, official
state history gives emphasis to the institution of schooling as a place to forge
the homogenization of the nation and mold its citizens. We put forward the ad-
ditional possibility that schoolchildren, in their role as pending citizens, are a
form of communal tribute, part of a pact in which parents, as original landown-
ers (originarios) of the ayllu, contribute to the state an annual “sacrifice” of their
children in exchange for their communal rights to land. This would explain why


communal and state control over childhood has always been so widespread and
It goes without saying that childhood plays a vital part in the eternal cycle
of reproducing the social processes of production, both in the tributary demand
to produce a new generation of taxpayers, and also in providing at home and in
the ayllu the future labor force needed to contribute the harvests of farming and
herding. Under state control, the colonizing models of biopower, which sought
to control production through bodily labor, formalized these relations of repro-
duction. As a result, the periodization of women’s labor pangs, in the reproduc-
tion of the labor force, coincides in the region of study with that of tax payments
on the part of the community. Both were measured in “thirds” (tercios) as part of
a system of payments thought to coincide with the gestation period.1
The textual basis of both tributary systems is inscribed in comunario memory.
History tells us how under the Inka state each male contributor between eighteen
and fifty years old was obliged to work the lands of the Inka, tend his herds, and,
as head of the household, give textiles in tribute. Qaqachaka comunarios record
the additional obligation of serving in the Inka army. Under the modern state, the
original landowners (originarios) in Qaqachaka still pay tribute even though the
need for this was formally abolished in the Agrarian Reform Law of 1953, an indi-
cation of its continuing significance for them as a guarantee of their rights to land
in perpetuity according to previous tributary systems.2
The textual expression of this payment is iconic. Once a year the necessary
tax is gathered, and then the leader of each of the six lesser ayllus delivers the
amount under his charge to the fiscal authorities in Oruro. The receipt (phini-
kitu) for payment is written on a small piece of paper that the leader carries like
a feather in his hat,3 aware of the similarity between this icon of tributary pay-
ment and the symbol of victory in the Inka army, of wearing a feather in the
The tributary system of the modern nation-state must also ensure that its
children are incorporated into the dominant economic system, of which the ayllu
is on the periphery. To assure this, the school serves as the key institution where
children are transformed from “internal” persons from the community into “ex-
ternal” persons with relations outside the community, with different goals and in-
terests. This transformation has a textual dimension as it involves important tex-
tual changes, the most pertinent of which is learning reading and writing.
Even so, there are different ways of interpreting this process of transforma-
tion, depending on where you are within it. Within the nexus of community-
state relations under the Inka, a continuity of functions between “one’s own” and
“the other’s” was mediated through cloth, in what we come to call an Andean
C Y C L E S O F M E TA M O R P H O S I S 89

form of interculturality. But within the nexus of relations between rural commu-
nities and the mestizo-criollo nation-state, functional and textual ruptures occur
between the one and the other.
In order to understand the full working of community-state relations,
we must examine regional concepts of childhood within the wider context of
schooling as the primary mediating institution between distinct textual polities,
their textual bases, and the different historical memories they draw upon. Bear-
ing in mind the historical configurations of the school-community relation, we
should also examine the part played by children in the mediation of local ideas
about identity and difference, of Self and the Other, as a vital part of Andean in-
Centering ourselves in the experience of tribute from the perspective of
Livichuco (Qaqachaka), we examine the articulation between the “productive”
and “reproductive” levels of the tributary system from a regional equivalent to
Bourdieu and Passeron’s notion of “reproduction” (1995). This permits us to un-
derstand the function of the school in the intergenerational transmission of ide-
ologies, knowledge, and textual practices, in which political and ritual aspects
predominate. In this intergenerational transmission, we can appreciate that each
year, children enter into the school premises not only as potential contributors
to the tributary system (on a par with the produce of the land and the offspring
of the flocks), but also as constant replacements for the contributors who can no
longer pay, those of the place who have passed away.

T h e C h i l d r e n ’ s F u n c t i o n i n t h e C o mm u n a l S c h o o l

In spite of the interest of colonial chroniclers (among them Cieza 1550, 1553;
and Cobo 1653) in the part played by children in Inka rites, few contemporary
ethnographic studies have examined the role of children in present-day rituals.4
Apart from Romero (1994), they focus even less on the articulation between the
ritual skill of children and day-to-day educational matters. The crux of the mat-
ter is this: given that the school provides an intermediate space between the com-
munity and the state, in which state tributary demands and associated textual
practices are articulated, how does the community interpret this phenomenon
according to its own interests? In other words, how do comunarios interpret the
processes of transformation that occur in the school premises, through which
their children are incorporated into the state tributary apparatus, in large mea-
sure through paperwork?
There are many aspects to the comunarios’ interpretation of day-to-day school
writing practices. In the commentaries we heard in a Livichuco workshop, those
laq'a awil
awila/ achachi
achila thakhi titul
uywiri sayxata q'ipi
tira mamitas
Tata authorities spiritual
wirjina of weaving
Santiago kin

qullu samiri of rutucha
parti midwives awicha achila
kin kin
yatiri compadres
tatala tiyula
spiritual tiyala mamala tatala
influences jisk'a tayka awki jisk'a
mamala tatala
qulliri qaquri
elder elder
sisters brothers

inspirational influences
influences p'axcha teachers at school

sirina school

Fig. 11. The model of communal education developed by the comunarios.

C Y C L E S O F M E TA M O R P H O S I S 91

present regarded the school as a “formal” state educational institution that func-
tioned beside their own “formal” educational institutions (see chapter 5). In these
parallel educational systems, they held that the didactic classroom practices of
recitation and memorization complemented (up to a point) their own didactic
practices of libation making, that school writing practices complemented their
own textual practices of weaving and braiding, and that school numerical prac-
tices complemented their own numerical practices (all points we shall examine in
later chapters).
In the same workshop, the comunarios developed their own scheme of instruc-
tion based on a series of concentric rings (rodeos) centered around the child at
home. During the two days of the workshop, first the men and then the women,
in groups, developed the same model with a few differences. First they incorpo-
rated the child within family and community spheres of influence (parents, aunts
and uncles, grandparents), including those of spiritual kin (godparents), to later
include state influences (teachers) when the child enters school. The same model
gradually expanded to include the principal experts of the place, the older people
and elected authorities (jach'a jaqi) in charge of guiding the children (thakhichiri,
“road guides,”), the wise ones (yatiri) in charge of ritual matters, and midwives
(usuyiri) for matters of childbirth. Finally, the model was extended further to in-
clude the elements that bestow power and spiritual authority to the persons in
their charge: “lightning” (q'ixu q'ixu), the outstanding hills (mallku), the main rit-
ual sites, and the sources of inspiration in music and song, namely the warrior
spirits of the dead (jira mayku and jira t'alla) and the sirens. (See figure 11.)
In essence, this communal educational model draws upon an institutional hi-
erarchy that includes the home, at one pole, and the political and ritual system
of governmental administration, at the other. The higher levels of regional An-
dean government, now absent, are represented here by the outstanding moun-
tains (mallku) of the region.5 The model also draws on a variety of educational
influences: economic, political, ritual, and spiritual. There is no predominating
influence; all coincide in the concentric rounds of libations and recitations, and
crisscrossing pathways of the ayllu territory.
In the dialogue that accompanied the design of the model, the comunarios
treated the school premises as an important ritual site, incorporated into a hier-
archy of such sites under the dominion of the highest mountains. In the ritual
activities that take place there, older people and those charged with ritual du-
ties (authorities, yatiri) fulfill their roles by drawing on the memory of former
Andean states, above all the Inka state. Thus the communal educational model
grounds its authority in historical precedent, and its institutional support in re-
gional discursive and textual traditions.

In this sense, the fulfillment of ritual duties in the school milieu not only con-
stitutes a series of communal didactic techniques, in the hands of the older peo-
ple and experts of the place, for the instruction of children in their own ritual
obligations to the community. It also transmits to them a greater historical con-
sciousness of the former Inka state as the primary institutional site of textual au-
thority and precedence.6 Ritual activities within the school premises are driven by
the memory of an Inka presence in the region and with the learning by comunar-
ios and children alike within an inherited Inka pedagogy.
These rituals draw particularly on Inka memories and symbols in the con-
text of practices related to communal modes of production and reproduction.
Regarding tribute, the key part played by the children in this economic, political,
and ritual constellation is that of assuring the continuity of this fertilizing cycle
through a community-state interchange centering on farming and herding pro-
duction, and of fulfilling their part in the reproductive cycles of the unending
provision of contributors, all in exchange for land. For these reasons, we hold
that communal ideas about the schooling complex, especially in the ritual sense,
function as a vital part of an educational strategy developed by the community in
a specifically regional “intercultural” context.

Andean Childhood and the Fertility of

t h e C o mm u n a l L a n d s a n d F l o c k s

What special quality do children have that grants them the power and au-
thority to fulfill these ritual functions? From the community’s point of view, the
schoolchildren act as vital intermediaries between the community and state by
assuring the fertility of the communal lands and flocks through the new textual
practices they are learning.
In commentaries by comunarios, the children sitting two by two at their
benches are compared with a team of oxen, so that the children’s role in learning
the foreign textual practice of reading and writing in the classroom is incorpo-
rated into communal interests in land. In a further play of analogies, they com-
pare the form and direction of the written lines on the page with the boustrophe-
don (to-and-fro) movement of the plow team, and the letters of writing with the
seed sown on the land.
Doña Antonia Espejo of Livichuco made this remark: “That’s why they’re
called a plow team. That’s why, for sitting two by two. Some are able, and they
teach one another.” For her, “it’s as if they were tilling the land” (yapuchañjamas-
paw); the letters they write are “like seeds” and the lines of writing “like furrows.”
That’s why parents insist that their children write in straight and not twisted lines
C Y C L E S O F M E TA M O R P H O S I S 93

in their exercise books. Otherwise, “people will laugh at us,” just as they do when
a field’s furrows are all crooked.
Other comparisons concerning the part children play in the school are ex-
pressed in the language of shepherding. Learning to read and write is compared
with grazing flocks on the local pastures. In a logical extension of this compari-
son, communal authorities and parents alike view their contribution of sheep for
the monthly upkeep of the teachers as the obeying of immemorial laws that de-
manded their own produce as part of the communal responsibility toward the
ordered running of the school.7
Shepherding is a mainstay in rural communities like Livichuco, so it is com-
mon to hear that “the children seem like animals in a flock.” Likewise, the tradi-
tional authority (or mayor) charged with looking after the school is considered
to be a “shepherd” (awatiri) and “the owner of the children.” Comunarios use de-
rivates of the Aymara verb “to drive animals” (anakiña) to refer to the way the
school mayor pastures the children like lambs. At the beginning of the school
year, when the parents leave their children with the school mayor for the first
time, they apply the term anxaruyaña, meaning to “hand over like livestock.”8
Then, as the mayor takes care of them during the school year, they use annaqa-
niña, which means “to drive animals from one place to another,” more specifically
in the direction of the final exams.9 Finally, with their liberation after completing
the final exams, they use anarpayaña (or anarpawiyaña), meaning “to release them
like livestock” and leave them once more on their parents’ pastures.10
The application of farming and herding language to school matters is not
just figurative. The whole ritual complex of the school forms part of a configura-
tion of ritual sites that concern the well-being of the communal lands and flocks.
The main ritual site in Livichuco, a pointed summit that commands the whole
community and functions as its guardian mountain, is called Anxata: “Put to pas-
ture.” Similarly, the main ritual at the beginning of the school year, which con-
cerns learning reading and writing, is called iskuyl anxata: “Put [the children] onto
the school pasture.”
Another key nexus between the ritual languages of farming and herding and
the schooling of children is the Aymara notion of wawa—“baby,” or “child.”
Wawa interrelates school scriptural practices with the textual practices of the lo-
cality through a series of associations made by community members. First we
should make it clear that wawa refers not only to human children but all kinds
of other offspring: produce from the land, from animals, and other things that
would in Western terms be considered inanimate, such as saintly images, stones,
hills, and so on.11
Some key meanings of wawa refer to states of transition. For example, school-

children as wawa mediate between the annual production of the ayllu as a form
of tribute to the state, and the constant reproduction of contributors, this by con-
stantly replacing elders and those who have passed away. In the cycle of potato
cultivation, too, the use of wawa in ritual language refers to the initial growth of
the new plant, tender and unripe, as well as the new tubers that culminate the
cycle with the next harvest.12 In this case, the semantic domain of wawa includes
the start and culmination of the cycle: seedling and tuber. Similarly among ani-
mals, wawa describes the initial fetus, as well as the baby that results from gesta-
tion. This wider meaning of wawa gives us clues to the function of children in
Both children and wawas are considered to be “flowers” (phaqara), and so they
are called in ritual language. In fact, their function in communal rites is said to de-
pend on the fact that “they are flowers.” This is not simply a figurative trope, but
part of a ritual language that links wawas with “flowers” and the spiritual domain
animated by “breath” (sami), thought to raise flowers and also the voice, imply-
ing that the ritual ability of children is driven in part by their vocal power. An as-
sociated idea proposes that flowers, as the new “hair” of plants and animals (as
corn silk or camelid fleece), embody the incipient fertility of the annual cycle.
In its widest sense, wawas articulate the liminal spaces between opposing do-
mains: masculine and feminine, the defense of what is and what is not one’s own
at the ayllu limits, the realm of the dead and the living.
In Livichuco, the rituals most associated with children are those to request
rains, a matter of life and death in the highlands.13 These rituals assure the an-
nual transition between the dry and rainy seasons and the greening of communal
lands with produce and the flocks with offspring.14 So how do comunarios explain
the ritual ability of children to beseech for rains so successfully?
The main reason given is their innocence: that children are still “without sin”
(jan juchani), since they “still don’t know about sex.......nor do they know about
killing or talking ill of anyone.” For the same reason, there is a norm against
adults talking about sex in the presence of children, and parents forbid their chil-
dren from playing marriage, “in case it becomes true.”
The specific nexus between children and the rains derives from the idea that
children are God’s “angels,” in the sense that they “have still not suffered any kind
of misfortune.” According to the wise man Don Domingo, this generalized inno-
cence of small children gives them the capacity to appeal directly to the gods of
rain, the Lord and Lady of Waters (qarpa mayku, qarpa t'alla) without the media-
tion of a priest or yatiri. In the gendered division of ritual labor, the girls request
rains from the Lady of the Waters while the boys beseech the Lord of the Waters
C Y C L E S O F M E TA M O R P H O S I S 95

to irrigate the community (qarpañataki), both under the guidance of elder peo-
ple. The innocence of children compels the rain gods to listen. Later, “the heavy
rain will come,” as will the irrigation waters. Inversely, the rain gods do not listen
to the requests of elders because of their burden of sin.
Children with this ritual ability tend to be no more than seven years old; older
ones can be withdrawn from a ritual at any moment if “they are no longer in-
nocent” for having “spoken” with the opposite sex. Although this ritual ability is
most apparent in the rites to request rain, it is also evident when small children
act as “intermediaries” between the community and their gods in the rites con-
cerning limphu, or “limbo.” Limphu occurs when a baby dies and is buried unbap-
tized, especially when incestuous kin have a baby without naming it and bury it
in a col (apacheta), acts that attract hailstorms that castigate the community and
bring misfortune upon the parents.15 To save the situation, the smaller children
must climb a mountain to pray to God (on a Thursday, considered a “good day”).
The communal authorities have the power to sanction people who have com-
mitted these misdeeds with a fine, the money from which is used for the school,
as the matter has to do with the schoolchildren. In Don Domingo’s community,
“for a boy it’s 100 bolivianos [then about $15] and for a girl it’s 50 bolivianos.”
Other ethnographic descriptions of the ritual capacity of children to commu-
nicate directly with God in communal matters (in the rites to request rain, those
to save a situation of limphu, or of prayer at the Feast of the Dead), mention their
energies (ch'ama), their ability to play (anataña), and their power to attract by
their beauty, a quality called ayri. In another comparative commentary, recorded
by Andrew Orta in the region of San Andrés de Machaqa (La Paz Department),
the first rains to bring fertility after a long period of drought are considered the
“tears of the little children” that have died on the way home.16
Further evidence concerning the nexus between the innocence of children,
their ritual ability and schooling, can be found in the struggle over the meanings
of school uniforms, especially the white smocks worn as overgarments. For their
part, teachers view the smocks as markers of the transition experienced by chil-
dren from the world of homespun handled in the community to that of learn-
ing reading and writing on paper at school. Moreover, this transition turns them
into new citizens and active participants of the nation, roles reinforced by the
ideas about progress and development that accompany this transformation. In
this sense, the white wrappings of the smocks, stretched like hides upon their
little bodies, envelop the children as tribute, part of a communal sacrifice toward
the social body of the nation-state.
Comunarios such as Don Juan Maraza, of Livichuco, while recognizing this

facet of the smocks as a wrapping of the nation, also admit other aspects. He
noted the requisite use of white smocks as “a matter of the director’s office,” in
orders that come from the director of the central school, and even farther up,
“from the district directors” at the apex of the educational hierarchy. But he held
that older people also related the whiteness of the uniform with the “innocence”
of the pupils, who have “clean hearts” (ch'uwa chuyma) and an absence of sin. As
“white flowers,” they have strong ties to the spiritual world, ample reasons for
them to wear white within the school premises. (We should not overlook the
strong colonial ecclesiastical influences behind these ideas.)
For his part, the wise one Don Domingo Jiménez affirmed the association
between the white color of the smocks and paper, the medium of reading and
writing, but he went one step further, incorporating this association into his own
views of the spiritual world. For him, the white color of the smocks as an inter-
mediate element in the cycling of spirit justifies the opinion of the teachers “who
say we must use the smocks because we learn to write on white paper.” At the
same time, he insisted that it is only through learning to write on white paper
that children “can learn about the spiritual world” and have “the spirit fly high”
in glory.
Aside from their innocence and strong ties to the spiritual world, children
have a special ability to communicate in the ritual sites common to the school
and the community with the very medium of rain, in the clouds and mists found
in the hills.17 As Don Domingo says: “They are stand-ins for the clouds.” In com-
munal thinking, these are places where the dead dwell and where the little angels
return if they die while still very small.
This constellation of relations implies that the world of children not only con-
cerns fertility in general, in the productive and reproductive cycles of the ayllu as
part of the state tributary system, but also the impulses that initiate these cycles.
Just as children communicate with the elements that produce the rains (clouds
and mists), so (by a causal series) they can release the rains themselves. In this
sense, the cloud white school uniform also symbolizes the children’s contribution
to the annual cycle of rains.
This association between the “little angels” and the clouds of the heights
has been known by the comunarios for centuries, through its ample illustration
in colonial art. The church of Qaqachaka, as in other colonial reduction towns,
housed paintings of this kind up to the last decade. Don Domingo recognizes
this source of this imagery (“I’ve seen the little children with their tiny shoes on
their feet, sticking out of the clouds”) as part of a wider association between the
winged little angels, the spiritual domain of which they are a part (ispiritu), and
the worlds of birds and paper.
C Y C L E S O F M E TA M O R P H O S I S 97

In practice, when a baby dies it is buried dressed with small white paper wings
that “help its spirit to fly.” As they make the paper wings, parents say they are
“making doves,” and when the spirit of the little angel finally flies off, it is identi-
fied as a “dove” that people watch fly up “above the sky,” toward God. The dove-
spirits of the little angels are compared to the “pair of doves” of a couple at mar-
However, this colonial Hispanic world of ecclesiastical teaching on canvas
about little angels, white doves, the paper of the Holy Scriptures, and the spirit of
God, was interpreted by rural communities according to their own experiences.
For them, another aspect of little angels is their association with the warrior spir-
its of the dead in the netherworld. This is the case in northern Potosí as well as in
Oruro, where these spirits of the dead (called jira mayku and jira t'alla) are said to
“come out of the middle of the clouds,” like God’s little angels.
This tie between God’s little angels and the warrior spirits has other reso-
nances in community definitions of what makes a “little angel.” In fact, many peo-
ple that we interviewed distinguished between the little angels who are “younger
and innocent” and the bigger children who “already know how to chew meat.”
(This unexpected difference may allude to certain rituals of Andean states in the
past, participation in which was prohibited for smaller children, as well as the re-
gional interpretation of the Eucharist, that equally restricts the participation of
older children until they have confessed to cleanse themselves of sin.)18
The transition for male children from innocence to adolescence through the
maturing of the teeth, to “eat meat” for the first time and so come to defeat the
enemies of the world, is a vital part of Andean textuality. It forms the plot of the
common tale of the Bear Man (Jukumari), the Aymara warrior par excellence.19
According to research by López G. in Carangas, parents and grandparents tradi-
tionally told this story to children in the rainy season, the seasonal context reiter-
ated throughout the tale, to illustrate the key nexus between the ritual ability of
children and the incipient fertility of the ayllu through rainfall (1998, 38–39).
Many versions of this tale link children with fertility and rains in a transi-
tional stage of growth that takes place in the school precincts. The heroic Bear
Cub, fully grown, with the appetite and physical strength to defeat any adversary,
classically has three chores to fulfill in proving his transition from apprentice to
warrior-contributor, after having defeated and killed his own father. Each task al-
ludes, in some way or another, to the cycle of activities we have examined in the
present chapter: the initiation of the rainy cycle and the development of male
physical energy, with its ties to anthropophagy.
In one version of the Bear Man tale, told to us by Milán Espejo of Qaqa-

chaka, the first task of the Bear Cub is “to ring the church bell.” But because of
his excessive strength he not only rings the bell but brings the belfry tumbling
down. In many rural pueblos, the sound of the bell is a call to the faithful to “pray
for rains.” The Bear Cub, then, not only participates in the matter of praying for
rains, but overdoes it. In the second task, the Bear Cub helps on the school prem-
ises, but in an uncontrollable way, convincing the teacher that he no longer needs
schooling. The Bear Cub has developed from apprentice to full-fledged warrior.
In the third task, the Bear Cub defeats the anthropophagous Damned One (Con-
denado) ravaging a village. They join battle to the onomatopoeic sound “ch'atx”
of rotting body parts being hurled from side to side, the implication being that
the Bear Cub is even more anthropophagous than the defeated Damned One.
Only after proving his skills in the productive chores of the ayllu, and its
protection against enemies, can the Bear Cub, now mature in body and mind,
become normal. The didactic function of such tales is to teach boys, at the ap-
propriate stage of their development, how to achieve maturity on attaining ado-
lescence. This happens in the social, cultural, and seasonal context of their im-
minent incorporation into the productive forces of the ayllu, to substitute for the
dead fathers.

Tribute in Children and Their

C yc l e s o f T r a n s f o r m at i o n

Regional ideas about childhood form part of the nexus of intercultural rela-
tions in the context of giving up children in tribute to the state each year through
schooling, in recompense for rights in perpetuity to lands and flocks. The state
school, on receiving this annual offering of children in tribute, has the task of
transforming them from persons of their own milieu to those with an outside
perspective. As a symbol of this transition, the children wear white smocks, like
paper wrappings, to mark their incipient acquisition of reading and writing. The
comunarios interpret this present transformation in terms of the past. From their
standpoint, the activities of the school precincts have to do with a fertilizing cycle
centered in children’s ritual powers. The dynamics of these differing perceptions
of conflicting dominions articulate the regional view of “interculturality.”
Any definition of interculturality, according to Howard-Malverde (1996, 116),
depends on a particular point of view, whether these are the horizontal and more
egalitarian relations between similar rural communities, or the more hierarchi-
cal vertical relations of a rural community inserted into a nation-state. In Qaqa-
chaka, there are horizontal relations of peaceful alliance between neighboring
communities, and more vertical and conflictive relations of domination, whether
C Y C L E S O F M E TA M O R P H O S I S 99

between Qaqachaka as a whole and the Bolivian state, or between communi-

ties at war with one another, in the ongoing boundary fights between ayllus that
characterize the region. The school mediates both sets of intercultural relations;
the activities that take place at the school form a microcosm of conflicting inter-
ethnic relations, as well as those between the community and the state.
A vital part of social and ritual memory of the community dwells on these
bellicose relationships between ayllus, and between the community and the state.
At a national level, warfare has to do with struggles for territory (in the Chaco
War or in the constant incursions from Chile on the Chile-Bolivia border) and,
at the local level, with the contemporary community’s fights for their own lands,
when they draw on memories of the Inka’s lands with their incorporation centu-
ries ago into Tawantinsuyu.
At the same time, state demands that local communities reproduce their
means of production (in labor and productive apparatus) and so fulfill their tribu-
tary obligations, have intruded on the community from the outside throughout
the colonial period and into the modern age. In this sense, the reproduction of
Andean communities has always depended in practice on their relation with the
outside world, and with varying notions of the Other, that is to say certain kinds
of “interculturality.” The regional symbolic economy, too, would seem to articu-
late this relation between communities and their Others (the state or neighboring
ayllus). Even so, at an intercultural level, the elements of this struggle (of war and
domination, expansion and control) are perceived differently by the protagonists,
each group wanting to reinforce what is theirs and diminish what is alien. As a
result, Andean intercultural relations have a characteristic approach toward the
ontology of being.
In Qaqachaka, the initial process of “defamiliarizing” children on the school
premises has many similarities to the domination of a captured being in the con-
text of warfare between rival groups. During schooling, the forces on the periph-
ery are absorbed to reinforce state tributary demands in the center of power. As
a corollary, a part of a child’s being ceases to form an integral part of the com-
munity and becomes incorporated into the state apparatus. Wolcott’s (1993)
comparative observation is pertinent here, that Kwakiutl children in British Co-
lumbia, Canada, are like “prisoners of war” in the educational milieu of the dom-
inant society, and the teacher represents “the enemy.”
Although the community-state relationship is less developed (as are tributary
relations), many of the same characteristics can be found in lowland Amerindian
societies (Tupi Guaraní, Araweté, Tupinamba, Wari). Here the primary mode of
interaction with the outside world takes the form of “ontological depredation.”
This term was coined by Eduardo Viveiros de Castro (1992) to describe the com-

plex of Self-Other relations linked to the activities of previous interethnic wars,

and incipient relations between ethnic groups and the state. Such relations have
depredatory and “destructive” aspects (incorporating the forces of the Other) as
well as “constructive” ones (absorbing the forces of the Other to reinforce the
Self ). The articulation of these relations constitutes an ontology of being.
Carlos Fausto (1997, 1999), developing Viveiros de Castro’s model, perceives in
this constellation of practices among some Amazonian groups the key moment
that initiates a whole cycle of activities related to interethnic struggle. Fausto calls
the first part of this cycle that begins with the capture of the enemy “productive
consumption,” to express its destructive as well as constructive aspects. This is
followed by the stage of “familiarization” of the captured enemy so as to incor-
porate him into the victorious group, which occurs through the conversion of a
“murderer-victim” relationship into that of “owner and child,” which he argues
is the principal scheme of symbolic control in the Amazon basin. He calls the
moment that articulates these two modes of relationship “familiarizing depreda-
tion,” describing the productive process of a symbolic economy based on the ap-
propriation of alien subjectivity from the outside world.
Depending on the different points of view (“perspectivism” to use another
of Viveiros de Castro’s [1998] terms) regarding these conflicting relations, two
aspects of this cycle occur simultaneously within the school precincts in Qaqa-
chaka. In the interests of the state, children in the school environment undergo
a process of “familiarization” molded through the teacher-pupil relationship.
However, from the community’s point of view, this same process is one of “de-
familiarization”: teachers, as state functionaries, are charged with transforming
“captured” children into beings of their own group, in order to appropriate alien
subjectivity and convert it into their own. The community reacts by generating its
own mechanism of reappropriation, drawing on the past to reinterpret the pres-
ent, in order to incorporate the alien nature of schooling into their own cycles of
production and reproduction. In this instance, the captured and defamiliarized
forces of the children are transformed into the ayllu’s own fertilizing forces. As
we shall see, these conversions also have their textual aspects.
The “defamiliarization” of Aymara-speaking schoolchildren in Qaqachaka
classrooms begins when they don’t understand anything, since all is in Spanish
and mutual communication is made difficult. In this alienating environment, ad-
ditional incentives such as international donations of food encourage the children
to learn another system of values. Elvira Espejo remembers her added motiva-
tion to learn foreign letters “for donated milk,” as if the exotic school milk was a
reward for the process of swallowing the alien letters of reading and writing.
C Y C L E S O F M E TA M O R P H O S I S 101

Fig. 12. The tongue and throne motif (Posnansky 1957, plate v, Chimbote Urn, and Imschoot 1990).

Such experiences lead comunarios from other regions, such as Don Domingo,
to interpret the alienating school environment as a place that “absorbs” children
to “turn their heads” and that, in the worse case, symbolically “eats up the chil-
dren.” Other parents confirm that schooling makes children “stupid” (sunsurata)
as it “heats the head” (p'iq junt'utatxijay) or “cleans the head out” (q'umachasjipï
p'iqix), in the sense of erasing their local knowledge and “lessening their commu-
nal strength.”
In his own interpretation of schooling, Don Domingo referred to the ancient
Andean symbol of “tongue and throne” that is widespread in stone, ceramics, and
cloth. One day, we were talking with him about some Tiwanaku iconography il-
lustrated in the book Tihuanacu: La cuna del hombre americano by Arthur Posnan-
sky (1957), especially a photograph of the Chimbote Urn showing the image of
“tongue and throne,” with that of a child with a flattened head passing over the
tongue. (See figure 12; the child is shown at the left.) Don Domingo immediately
related this image to what traditional schooling does, when the goal of teaching
is to dominate the voice of the children by that of the teacher. “Well, the school
has called the children and, after gathering them all together, it eats them. Then
at another moment, they say, when it’s time, it lets them go too........In school,

it gathers their tongues, even though they study there.” Here he related the en-
trance of a child into school, from the standpoint of the state, with the flattening
of its head, and its incorporation into an alien dominion, voice and all.
At a textual level, the historical image of “tongue and throne” seems to ex-
press the authority of the person sitting on the throne (the teacher today) to sub-
jugate the tongues of all his subjects (the pupils) through vocal power and the
territorial extension of his verbal reach. In the visual expression of this dominion,
the tongue as an integral part of the throne has the capacity to wrap and even
swallow the subject with its vocal demands.
Pitted against the tendency of the traditional school to silence rural children
and convert them into disoriented beings, the present educational reform makes
this initial transition of values easier, according to the new constructivist psy-
chopedagogical methods, less traumatic than previous ones. A characteristic di-
dactic technique of the reform is encouraging “play” by making clay miniatures
in the classroom, especially when children’s Spanish (as their second language) is
still not fluent, in activities that are molded in the units of the reform modules
(see again figures 4–7).
Nevertheless, there is a fundamental difference between play as one of the
goals of the reform’s pedagogical practices, and play’s function in the community.
Sillar (1994), for example, relates the didactic function of child’s play in nearby
Macha with learning to “play with God” and so communicate with the deities, as
in the ritual requests for rains. For their part, Isbell and Roncalla Fernández (1977)
emphasize the importance of play in the socialization of Quechua-speaking chil-
dren in Peru, especially in the years of adolescent sexual and intellectual explora-
tion, when a competitive game (pukllay) is practiced related to herding and the
learning of warlike activities through duels and competitive dances.20
Sillar proposes that Andean play, instead of fostering the dependency of chil-
dren on their teachers and parents, rather encourages the dependency of adults
on their children, in the sense of valuing their contribution to communal well-
being (1994, 49). From the perspective of community needs, child’s play stimu-
lates the fulfillment of their ritual obligations in ayllu life by initiating the cycles
of production and reproduction through prayers for rains, and so create a new
vegetative covering and an assortment of new “flowers.”
This is why many comunarios reject the teaching through play as done under
the educational reform, as the children who play in that way “no longer feel a part
of the household chores.” They sense that the present approach to play is cre-
ating a rupture with communal modes of production and textual practices, en-
couraging a textuality centered in the acquisition of reading and writing, points
C Y C L E S O F M E TA M O R P H O S I S 103

we develop in chapter 5. In this sense, the new approach to play fostered by the
reform has much in common with similar changes that happened in Europe, no-
tably the “invention of childhood” in a specific historical moment when the mer-
cantile goal was the greater inclusion of children in the world market.21
In contrast to what is taught in the reform curriculum, playing with miniatures
in the Andes has ties to that other constellation of ideas based in regional textual
practices and linking children with rains, life, death, production, and reproduction.
In Qaqachaka, this is illustrated by the miniatures of daily implements made by
children (ploughs made by boys, and looms made by girls, for example), which are
buried with the placenta of a newborn baby or with an adult corpse.
In many Inka sites, such miniatures of children, animals, agricultural pro-
duce, and objects made of stone, ceramics, wood, shells (especially Spondylus),
even gold and silver, have been found on the highest summits. The miniature
male figures don’t wear breechcloths or earplugs, suggesting that they repre-
sented small boys before the onset of adolescence and the associated rituals of
becoming orejones (“long ears”) or warriors.22
Apart from the miniatures, remains of sacrificed children have been found in
such places. As Sillar observes (1994, 56), Inka human sacrifices were rare, but just
as the Aztecs and Moches offered prisoners of war, so the Inkas offered children.
Previously we suggested that the annual offering of children to the school on the
part of the community is a sacrifice-like tribute in recompense for communal
land rights. The study of Andean war and sacrifice, although horrific in its impli-
cations, provides important clues for understanding the regional history and in-
terpretation of schooling. It also suggests ways of understanding these phenom-
ena even beyond the Andes.
Why did the Inkas offer children in sacrifice, rather than prisoners of war? A
partial answer can be found by scrutinizing some present contexts in which both
statuses, child and war prisoner, are almost interchangeable.
Our recent study of childbirth in Qaqachaka and other Andean regions
shows how certain practices of childbirth, deeply rooted in culture and difficult
to change, draw this comparison.23 One is the custom of not breastfeeding im-
mediately. In Qaqachaka, more attention is given to the immediate needs of the
mother, and after having washed the newborn, it is simply put to one side. Only
if it cries a great deal is it breastfed; if not, then the mother waits some hours be-
fore feeding it. When we asked the reasons for this attitude, we were told that the
baby “must be strong,” “it must be a good warrior,” and “it must resist the diffi-
cult highland environment.” Moreover it must learn to live modestly, and “not be
spoiled,” for “it must not take advantage of all the community’s resources.”

Consider as well the custom of securing a newborn with swaddling and other
weavings, and then wrapping it with a thick sash in a “nest” of cloth that covers
it during the first three months of life. Such treatment goes against the modern
tendencies to “liberate” the baby in every sense from the start. Again, the reasons
given by mothers for this treatment of the baby are “to give it strength,” “so that
it is hard,” and “for it to grow straight.”24
These customs recall similar practices in other parts of the world (among
the Aztecs of Mesoamerica and the Pano, Cashibo, and Matses of contempo-
rary Amazonia), where the newborn is treated using a vocabulary of warfare
and compared explicitly to a “captured being.”25 These variants on the same per-
vasive symbolic economy, in both highlands and lowlands, seem to concern the
same nexus of sacrifice and tribute in conflictive relations between groups. An-
dean childhood has ever been controlled and measured by state interests to ob-
tain a constant provision of labor force. What remains to be understood is the
relation between the new entry of pupils, the domain of war prisoners, and the
generation of tax-paying males in their role as warriors.
If a child becomes a prisoner of war with a flattened head (speaking symboli-
cally) on entering a state school as a tributary sacrifice from the community to
the state, then how does the community achieve the reintegration of a battered
head as a child of its own? The answer can be found in ideas concerning children
as intermediaries (between the world of the dead and the living, the alien and
one’s own, the bad and the good) leading to the other pole of the cycle of meta-
morphosis in which schoolchildren are immersed.
In “La papa, el amor y la violencia” (Arnold and Yapita 1996b), we examine
this constellation of ideas in the context of the present-day wars between the
ayllus of the region, although its history is much longer. Here, the gradual trans-
formation of what is alien into one’s own occurs when the head of a dead en-
emy is converted into a child of one’s own group, through an extensive ritual
process.26 In the context of state schooling, something similar happens. From the
standpoint of the state, the child as a captured being (a prisoner of war) is con-
verted through education into an enemy trophy head (always in the figurative
sense). But from the community’s point of view the same complex is reversed—
and the enemy trophy heads converted back into children.
These ideas echo the “trophy head complex” found in other Amerindian
cultures, the best-known being that of the shrunken tsantsa heads of the Shuar
(Jívaro) of Amazonia, as described by Descola (1993a). There, through a series
of rituals (also called tsantsa) lasting more than a year, with their own choreog-
raphy, narratives, songs, and prayers, the tsantsa head “assumes all the roles of
C Y C L E S O F M E TA M O R P H O S I S 105

a symbolic procreation: of non-kin, wife-giver, wife-taker, husband, and finally

embrion” (1993a, 183) Through the rites of the tsantsa head, the Shuar come to
dominate the power of the enemy trophy head, converting its strength into the
ancestral power of their own group. In the meantime, the head/baby is said to
weep uncontrollably and the women of the place must console it. At the conclu-
sion of this long ritual process, the trophy head is converted into an embryonic
child, reborn as blood kin of the same group. As Descola comments, it is a rite
charged with “esoteric allusions to death and rebirth, fecundity and birth, the sav-
agery of cannibalism and the immemorial rules of social harmony”27
In Qaqachaka, a similar process occurred in the recent past. The head, seat
of the powerful spirit (now called ispiritu, but whose Aymara name ajayu is still
heard), was highly desired as a war trophy, especially that of an enemy leader.
Captured in battles over boundaries, such trophy heads were exhibited first in
public on top of a pile of stones called taqawa (the same place we find later on
the school premises). The warrior who captured the head took it home later,
where it was kept wrapped in a black cloth, in the charge of the woman of the
house “to make her fertile.”28 The head would weep like a baby and the woman
had to console it. After a period of three years, and the enacting of various ritu-
als in which libations and incense were offered to the malignant enemy head, it
was transformed into a benevolent baby. By way of this process, the spirit of the
baby/head was transferred to the belly of the wife of the conqueror of the en-
emy spirit, to be reborn into the warrior’s home.
Similar ideas can be found among other Amerindian warriors, for example
in Amazonia among the Wari29 and the Araweté, a group of Tupi-Guaraní. In
the latter case, Viveiros de Castro (1992, 9) observes a textual facet of this cycle
when “war songs are often used as lullabies.” Like the present-day rites of child-
birth and the former rites of warfare in Qaqachaka, these examples reiterate the
similarity between the recently born and a captured being, and imply that the
newborn results from the metamorphosis of this captured being, or its head as a
trophy of war.
This suggests that in the wider context of Andean intercultural notions of
struggle and reappropriation, two tendencies centered on the school and the
yearly intake of pupils happen simultaneously. Through the annual incorpora-
tion of children, their conversion into captured beings and their subsequent reap-
propriation into community life, one process is directed at the reproduction of the
state in the community, and the other at the reproduction of the community in the state.
In addition, comunarios reappropriate the power of the enemy and convert it into
their own productive cycles through the communal textual practices of a woven

polity. During this process, the captured heads become the templates of a gener-
alized ontological space from which all the practical activities of ayllu production
and reproduction are generated.
In Qaqachaka, the upbringing of children (like the reappropriation of trophy
heads) articulates gender relations, both in the gendered division of labor (ritual
and everyday) and in the masculine and feminine forces that contribute to child
development. Men as warriors must capture trophy heads as an alien force to be
appropriated. Like seeds that sprout, the alien forces in trophy heads are gradu-
ally domesticated and transformed into the multiple babies (wawa) of the house-
hold under the care of women. In this way, the community depends on mascu-
line physical strength and forces of production to attain a new harvest of human
babes in the home, vegetable offspring in the fields, and animal offspring in the
flocks, and then to sustain and nourish them through physical work.
In terms of reproduction, male access to the potentiality of the seedlike tro-
phy heads helps generate the fecundity of women and in turn other reproductive
and productive cycles. In terms of the textuality of their practices, male physical
force is indispensable in obtaining the basic seeds (trophy heads) and then open-
ing the furrows in the fields, perceived as the warp of the earth, in the ensuing
farming cycles. In their accompanying activities, men call the rains with their mu-
sical instruments, and stimulate production through their sweat (jump'i), energies
(ch'ama), and spiritual strength (ispiritu). Masculine authority also draws on the
ritual power of speech of the trophy heads in the public sphere, to command civil
obedience through language.30
Feminine authority is based on household and ayllu production and the up-
bringing of offspring (human, animal, and vegetable). A woman has the produc-
tive task of raising human babies at home, where the initial care of the babies,
as transformed enemies, concerns a domain in which the foreign and uncontrol-
lable forces found at any boundary (the ayllu ancestors, or enemies beyond the
ayllu bounds) must be domesticated within human society. Later on, she releases
them to the more masculine domain of the school. Another productive female
task is sowing the baby/seeds in the fields and then tending them in the weftlike
to-and-fro routine of agricultural work, under the tutelary spirit of the Virgin
Earth (Tira Wirjina or Pachamama). Women imitate the maternity of the Virgin
Earth in tending the fields, where the plants are not inert things but have their
own spirits or essences: the “virgins” of each house plot. Thus, agricultural pro-
duction also depends on the ability of women (in their maternal and fecund as-
pects) and the fertilizing spirits of the place. To this end, the women sing wayñus
and other verses directed at the plants in the rainy season, while the men play
their pinkillu flutes.31
C Y C L E S O F M E TA M O R P H O S I S 107

The Virgin Earth in her role as tutelary spirit of the orchards and house plots
is the creator and mother of the cultivated plants that are her babies, while at
the same time cultivation is always under the threat of her curse. According to
Elvira Espejo, the Virgin Earth and the other virgins under her charge call into
existence not only cultivated plants but also the flesh of the animals and the mak-
ing of weavings, so she bestows upon women the domestic arts of horticulture,
cooking, and weaving. In this way, the Virgin Earth mediates between the textual
practices of the place and the forces of production.32
Like written ramifications upon the primary symbolic template of the trophy
head, the patterns of plant growth and development, the fulfilling of human pro-
ductive tasks, and the work of weaving, all follow a series of recurring homolo-
gies. For Elvira Espejo, the degree of success in agricultural production depends
on the growth of the roots of the plants in the fields. Just as the plant roots must
grow long and intertwined, so the pathways of daily activity of the cultivators,
men or women, must be long and crisscrossing. She explained: “Our pathway is
the same as that of the plant entanglements, the same as the roots.” Similarly, she
compared the lines of furrows, and the interlacing of pathways traveled in daily
tasks, with the to-and-fro movement of the textile weft upon the warp. In this
case, the warp is constituted by the plants themselves, and the interlacing of weft
and warp with the growing entanglements of the baby plants.
The same logic applies to the role of children in school, in the production
and reproduction of the ayllu elements. The pupils, in their daily tasks of learning
reading and writing, are furrowing the ayllu plots like plough teams and grazing
like sheep on the ayllu pastures. Through these processes, the children are plant-
ing letters like seeds in miniature, and then transforming them to weave a new
vegetative covering. According to communal thought, the elements of school
reading and writing are being reassimilated in a woven template under their own

To wa r d a n A n d e a n T h e o ry o f T e x t ua l R e p r o d u c t i o n

This Andean sequence of ideas concerning production and reproduction chal-

lenges the theoretical postulations, uprooted from their context, by Bourdieu and
Passerón (1995) in their proposals concerning “reproduction” in the French edu-
cational system. Our alternative approach is to relocate the theories of Bourdieu
and Passerón in a historical context, that of community-state relations. In this
spirit, we put forward an Andean theory of reproduction centered in schooling,
in which the elements to be reproduced are grounded in farming and herding
production in their eco-biological sense. This regional theory of reproduction

derives from the textual dynamics of the warlike yet fertilizing metamorphosis
from baby to head and back to baby again. Given the regional dynamics of onto-
logical depredation, the textual expression of this process draws on the produc-
tion of new persons from the elements of their destruction.
Textual practices express this metamorphosis in cloth, iconography, ceram-
ics, and narrative. One well-known example is that of Inkarriy, with its messianic
stance, which draws on the transformation of hair to roots to express the dy-
namics of refertilization. In the myth of Inkarriy (the Inka king), the Inka’s head
is buried like a seed in the earth where his hair will sprout again like roots. On
sprouting from the seed/head, the baby hair/roots will stimulate the eventual
union of the head with the body, and so reconstitute the Inka’s authority at some
moment in the future.33
The fertility of hair (whether the hair of the dead, the hair of the new plants,
pastures, or fleece) is essential in the reconstitution of a new being. As they say, it
is hair rather than the other body parts that incarnates its powers of fertility in life
and after death. Don Domingo Jiménez compares hair specifically with the rains
(the domain of the children) that irrigate the earth, making it fertile: “Hairs are
the rain, that’s what waters.” Likewise he compares the sweat that drips from the
hair onto the brow with fertilizing irrigation, particularly when men sweat dur-
ing agricultural labor or in warfare: “Sweat irrigates us, it falls from here, from
the brow; the sweat empties out from here, and then we start sweating.”
In Qaqachaka, however, the founding myth of all productive and reproduc-
tive domains concerns the way that textual practice transforms a trophy head
into a new being. As a regional counterpart to an alien Holy Scripture, this myth,
in a quasi-religious sense, proposes that the dead can be resurrected on looms
into new beings (a point we develop in chapters 5 and 10). Weaving discourse,
drawing on this myth, proposes that the origin of Andean textiles derived from
the same primordial element of a trophy head, transformed through the activ-
ity of weaving into a living being (a baby of the same group). Much supporting
archaeological evidence and textile iconography suggests that this was so. In this
case, the primordial warps were the very hair of the trophy head stretched out
over the loom cage.34
Contemporary weavers imply that this is the case when they compare warp-
ing the loom with tending the long hair of a head. As Don Domingo observes:
“Whether it’s a mantle or any other weaving, it’s as if we were to loom up with
hair.” The young weaver Elvira Espejo confirms the assertion that warping the
loom is like tending hair of different colors. Even more pertinently: “In warping
the loom to weave, it’s as if we were transforming a baby into a person.” In warp-
C Y C L E S O F M E TA M O R P H O S I S 109

ing the loom, then, the alien forces of an enemy head/baby are reappropriated
and made into a new person of the same group (wawa jaqichayasiña).
This myth explains the conflicting layers of textual appropriation centered
on the bodies of communal schoolchildren when they are taken as trophy heads
in tribute to the state. On the state side, the children must reproduce local farm-
ing and herding production, not with the sweat of their brows in the fields or
pastures, but on paper. Like sheep in an immense new pen, they must bleat in
chorus according to the verbal directions of the teacher. Within these same goals,
the pupils must envelop their former clothes woven with fleece (the hair of the
flocks) with the new wrapping of white smocks as the textual medium that gives
them the necessary spiritual energies.

Wa r rio r s and Weav ers

The Pathways of Learning in the Community

I believe that there is box-thought, the thought we call rigorous, like rigid, inflexible boxes,
and sack-thought, like systems of fabric.

Michel Serres, Rome: The Book of Foundations

With the Spanish Conquest and five hundred years of colonization, republican-
ism, modernization, and now globalization, many formal Andean educational in-
stitutions—the aqlla wasi, houses of learning for young weavers, and the yachay
wasi, schools for the sons of caciques—were wiped out or seriously undermined.
But other aspects of the regional educational apparatus, with textual roots in the
previous period, continued in changed and adapted forms. It is to these we now
The systematic communal educational practices found in Livichuco and other
rural communities of the region challenge the point of view that holds there is
no “formal” education outside the institutional context of modern schooling.1
Rather these communal practices have formal and informal aspects, parallel to
those known as “popular education,” “adult education,” or “alternative educa-
tion” in Bolivian society at large.2 In the hands of local experts, specialists, and au-
thorities, and organized through an institutional structure of “pathways,” these
practices have moral, ethical, and ontological dimensions that concern knowl-
edge and its transmission in the community, along with the appropriate commu-
nal and corporeal sites where this takes place.

WA R R I O R S A N D W E AV E R S 111

T h e A n d e a n N o t i o n s o f “ Pat h way ” a n d
“Becoming a Person”

The communal educational practices found in this Andean community share

the common emphasis in Amerindian thought on the production of persons
rather than material things. This has textual ramifications in the instructional
modes of teaching and learning within the community.
In Livichuco, two constellations of related ideas apply to the development of
persons from childhood until the moment of “becoming a person” at marriage.
One concerns the ontological notions of being, knowing, and personhood, when
attention is given to the socialization of children through the interventions of
others in child development, and the maturing of the body rather than the in-
tellect. The other concerns the structured pathways of learning as instrumental
institutions. Here ontological depredation gives importance to appropriating the
subjectivity of Others through head-taking and incorporating it into one’s own
from the “pathway of marriage” onward.
Both constellations of ideas share a common space of knowledge, a frame of
spatial and conceptual limits reached during one stage and then ruptured before
proceeding to another more extensive space. Because of this, the prior stage is
more internal and the second, more external. This frame of reference is homolo-
gous to those that apply in weaving.
Where the educational reform curriculum enforces notions of personhood
in reading and writing and a new paradigm of being and knowing, communal
education locates the notions of personhood, being, and knowing in the body
and the plastic expression of this corporeality in cloth. This emphasis on corpo-
reality as a vehicle for the acquisition of knowledge is widespread in Amerindian
thought. For example, the epistemology of the Cashinahua (hini Kuin) of the
Brazilian and Peruvian Amazon basin, as described by McCallum (1996), empha-
sizes the functions of the body in the social and cultural construction of being.
Among the Cashinahua, as in Qaqachaka, the internal development of the fetus
depends on the contribution of corporeal substances from its parents, whence
it absorbs medicinal plants in infusions and the special foods of pregnancy, tem-
pered through the specific actions of internal processes.3 Later, at birth, there be-
gins a series of more external processes aimed at hardening the small child’s body
through parental action. In short, the construction of a person occurs in relation
to other persons, in the context of an ecosystem where identity is attained gradu-
ally in a defined territory.
McCallum differentiates between the initial construction of the “natural per-

son” and then that of the “social person,” and something similar happens in the
educational scheme of the comunarios of Livichuco (see again figure 11). The first
stage occurs in the family setting when the child is modeled by interventions on
the part of its family members and its education is oriented toward the space
of the home. During this period, knowledge is internalized. Becoming a social
person occurs later, when knowledge is externalized though physical action, dis-
course, and song. At this stage, the “persons who know” transmit their knowl-
edge to others through the textual practices of the place.
In this context, Qaqachaka notions of the body are not singular but are con-
stituted by a series of distinct spirits or souls developed in different stages and tra-
jectories. One of the spirits resides in the skin, where the natural phenomena of
the ecosystem are molded. In Qaqachaka, the term janchi (skin) is interchange-
able with kurpu (body);4 there is no Aymara term for the “body” as such, because
people are not thought to be separate from their surroundings as in modern Eu-
ropean thinking. Typically, Elvira Espejo observes how the skin is a mediator be-
tween the internal organs and the external world, a kind of “sieve” (suysuña) with
open pores (sit'artata) that must be protected. The condition of the skin (and of
any additional skinlike wrappings of woven cloth) defines the condition of being
a person and even that of the social body.
As a result, the various spirits or souls of the body (ajayu, janayu, animu, suti)
are perceived as wrappings or layers with a textile-like nature, one over the other.5
This bundling of souls is not static. From childhood until old age, the inner lay-
ers gradually become externalized in the dynamic transformations between lay-
ers of spirits, the same dynamics that produces corporeal development, growth,
and maturity, and then a final decadence.6 Qaqachaka ideas about the body, being
and knowing, and growth and development all have textual homologies in cloth
rather than paper, though these same ideas are transferred to paper in the context
of schooling or the state bureaucracy.
Qaqachaka notions of the “juridical person” (jaqi) are similarly grounded in
weaving homologies. One “becomes a person” (jaqichaña) at marriage,7 a trans-
formation marked through weaving. In chapter 4, the weaver Elvira Espejo com-
pared the action of warping the loom with “making a person out of a baby”
(waws jaqichayaña). Cipriana Apaza, another weaver, but from the region of San-
tiago de Huata (La Paz Department), compares the action of weaving with “con-
verting something into a person” in the following sayings about weaving based
on the same verb (jaqiptayaña):8
WA R R I O R S A N D W E AV E R S 113

Janiw jank'as jaqiptarapkituti. ”I am not converted so rapidly into a person,”

lit. “It is not converted into a person so quickly
[for my benefit].”
“Saw jaqiptayxä” sasaw, “yanaskta.” “I am trying to convert the weaving (sawu) into a
person,” saying.
Isi jaqiptayañax ch'amawa. “Converting clothes into people is difficult.”

Cipriana Apaza applies this same verb of “making a person” both when the
loom is warped and also when the textile borders are finished.9 Then, by logical
extension, this weaving action is compared with the farming action of “finish-
ing well the furrows of a field” (suka jaqichthapiña). In another farming activity,
that of “removing stones and weeds from the surface” and so “selecting well” the
crops to be cultivated, the related verb jaqichsuña is used.10
For Elvira Espejo these verbal derivations based around “making a person”
have metaphorical dimensions to do with the action of filling a defined space in
all directions with the things that belong there, and which the owner desires in or-
der to achieve personhood. For her, filling the loom is making a person, as is fill-
ing the home with belongings gradually after marriage, when one says “let’s fill
the home with belongings, little by little” (Sumat sumat jaqichashñani). Elements
in the field that do not belong, such as weeds or stones, must be removed, leaving
only what does belong, that which “makes it a person.” Likewise, in the case of a
spoiled child, the wrong attitudes in its character must be removed before it can
“reach personhood.”
In this sense, the idea of making a person, while linked more immediately
to the domains of marriage and weaving, extends into a wider semantic domain
concerning desire as a motivating force, on the one hand, and on the other, filling
a defined space in all directions with a multitude of elements and then finishing
its borders.

A ndean I deas of “ K no w ing ”

More generalized ideas about knowing, whether based on the Quechua verb
yachay or the Aymara yatiña, refer to a similar morphology in the organic sense of
development to occupy a certain space. The more metaphorical sense is equally
important here, given that Andean concepts of knowing have become the sub-
ject of debate in the context of the educational reform, especially in the design of
the new textbooks with their textual framing of knowledge on paper. Although
there are dangers in comparing the present lexicon with a more distant historical
context, nevertheless present debates are foreshadowed in the evangelizing pro-

grams at the beginning of the colonial period, when similar attempts were made
to skew Andean concepts of knowing, for example in translations into native lan-
guages of the idea of a Christian creator god.
The French linguist César Itier (1992), in a summary of the debates surround-
ing the pre-Hispanic meanings of “knowledge” in a missionary context, concludes
that the derivations of the Quechua verb yachay can be reduced, not to “create”
in the Christian sense, but to a specific unit of meaning: the move to “accom-
modate itself within limits” or of “occupying a whole space” (101). For Itier, this
notion has a botanical homology, namely the growth of a seed or plant from its
embryonic state to its full development—“multiplying like sown land” (multipli-
car como sementera), as yachacuni is defined in G. Holguín’s vocabulary of 1560—
that extends to the way other things are developed: corn, coca, food, springs, ir-
rigation ditches, walls, and so on (ibid., 102, 105). Itier argues that glosses such as
these led to yachay being identified as the Quechua concept closest to the “creat-
ing” of the Christian God.
These historical Andean ideas of “knowing” and accomodating knowledge
“within limits” concern a similar semantic domain to that of the contemporary
Aymara notion of “being” (jaqi), and its allusion to a constellation of corporeal
and spiritual aspects, as well as the very seat or space of knowledge, and its tex-
tual expression. As we have seen, contemporary Andean ideas of “knowing” also
tend to emphasize the dynamics of seminal, multiplying, and fertilizing processes
linked to weaving activities.
A possible conceptual bridge between both these domains “being” and
“knowing” can be found in the term yä (or yäna), the probable common root of
Quechua yachay and Aymara yatiña, “to know.”11 Yäna is still used in the Aymara-
speaking region around Lake Titicaca to mean “goods, possessions, or things,” in
the specific sense of a man’s or woman’s assets (chachan yänanakapa, warmin yäna-
nanakapa) that have accumulated at marriage, or the “parents’ things” already ac-
cumulated, that on their deaths will be divided among the descendants (Maman
yänakapasti: And mother’s things?). Similarly yänasiña means “negotiate or sell to
gain money,” with the same implication of multiplying things.
In his attempt to identify differences between Western urban and Andean in-
digenous thought, the Argentinian philosopher Rodolfo Kusch, in El pensamiento
indígena y popular en las Américas (1973), notes the expansive modality of Andean
concepts of knowing from entries in colonial vocabularies. He identifies an op-
position between two tendencies: one toward the fragmentation of reality (in
Western thought) and the other toward its unity and multiplication (in indige-
nous thought). For Kusch, the first tendency is characterized by causal thought,
in which each subject “sees” the world, delimiting its details in order to confront
WA R R I O R S A N D W E AV E R S 115

it efficiently. In order to understand it you ask “why?” in an analysis of reality that

seeks articulated divisions and perceives these divisions separately. The second
tendency, by contrast, is characterized by connective thinking whereby each sub-
ject “feels” the world, and instead of asking “why?” asks “how?” For Kusch (1973,
83), the “how” refers to modality and deals with an organic vision of reality to do
with feelings expressed through the organic and seminal semantic domain of the
Quechua verb yachay, “to know.”12
He gives as examples the Quechua definitions of yachacuni: “to increase” (acre-
centar according to Molina, ca. 1575)13 and “multiplying as a seed-bed” (multipli-
car como sementera according to Santo Tomás, 1560), and the Aymara equivalent,
yatayaña (or yatiyaña): “Make, prepare, or compose something, and raise some-
thing that is typical of God (Hazer o adereçar, o componer alguna cosa, y criar que es
propio de Dios according to Bertonio, 1612: vol. 1, p. 304). Like the Andean notions
of personhood discussed above, the semantic domains of “knowing” similarly al-
lude to the generative spaces in which knowledge can proliferate.

H eart and H ead as S eats of K no w ledge

In the modern Aymara of Qaqachaka the dynamic transmission and recep-

tion of knowledge between persons is expressed by adding suffixes to the basic
verb “to know,” yatiña. This dynamic operates within the corporeal processing
of knowing and its interaction with memory in the two principal seats of know-
ing in the body, the heart (chuyma) and the head (p'iqi). The heart is held to be the
principal seat for women, the head for men. In the initial construction of gender
during childhood, these differences are reinforced through food: as a norm, the
head and hard parts are given to males, the heart and soft parts to females.
Many studies of chuyma, “heart,” underline its metaphysical meaning bet-
ter translated as “conscience,” or more idiomatically as “pluck” or “guts,” rather
than as the physical heart, which is lluqu.14 Chuyma is a generalized seat of one
of the vital souls. Among the Cashinahua, a principal soul resides in the con-
science, memory, thought, and feeling, whereas in Qaqachaka the equivalent is
called ajayu (to do with the state of the blood and skin), combined with the flow
of ispiritu (more related to breath and the voice). Chuyma, as the locus of circula-
tion of these elements, describes character, aptitude, thought, feelings, and intel-
These metaphysical bases structure educational practices and determine
their modes of textuality. Given the sociability of communal knowledge, the idea
of teaching someone has little relevance, neither has the definition of a school-
teacher as “a person who knows” (as used in the materials of the educational

reform). But it is generally recognized that the knowledge of “someone who

knows” (yatiri) or “who makes someone else know” (yatichiri) is deposited in
the heart or head, and a good teacher is considered to be “someone with a good
head” (jupax suma p'iqiniwa).
The reception of this knowledge by a pupil is received (“grasped”) in the
heart (chuymar katuqaña), and children who grasp knowledge quickly are called
amta chuyma, “knowledgeable or remembering heart.” In a weaving analogy, this
knowledge is held to be “tied” into the heart (chuymar chinuntaña), as if it were
just another thread tied to the ball of wool that constitutes this organ. For the
weaver Elvira Espejo, the heart is simultaneously a “bundle of threads” (ch'ipa
chinu) and “a mass of veins in constant movement,” like scissors (kachi winarara).
Once knowledge is incorporated into this ball of blood-threads, it passes from
the heart to the blood system, from there to the head and finally to the voice, the
oral medium through which it is transmitted to someone else. In this dynam-
ics of orality, the genealogy of textual transmission is achieved within the same
framework of reaching the corporeal limits, and then surpassing them through
speech, in principles of textual transmission that equally find their homologies in
Given the vital function of the blood system in the transmission of knowl-
edge, communal theories and practices of teaching and learning emphasize the
circulation and state of the blood in different moments of life. A young person is
thought to have a great deal of strong, thick, hot blood and a large heart full of
such boiling blood. With age (and especially in the case of women) the blood di-
minishes and becomes thinner; the heart also dries and gets smaller.15 As a result,
knowledge becomes more concentrated, so an older person is considered to be
“big hearted” and “knowledgeable” (chuymani). This way of thinking about older
people and their concentrated knowledge is very different from the biopsycholog-
ical focus on which the present state educational system is founded, which views
advanced age as leading to the diminution of the faculties. Communal education
grants authority to elder people as experts in the transmission of knowledge to
younger people, whereas the state educational system discounts such authority.
An older person’s knowledge is incorporated through a process of social in-
terchange, for example the pupil “takes down” a portion of the teacher’s knowl-
edge via the voice. In Aymara, this sense of receiving or “taking down knowl-
edge” from someone else is expressed by the verb yatiqaña, in which the suffix
-qa indicates the “taking down of a part of something.” In the other direction,
the transmission of knowledge from the elder person’s body to “let someone else
know” is expressed by the verb yatichaña, in which the suffix -cha indicates this
causative action. This use of the verbs yatichaña and yatiqaña indicates that the
WA R R I O R S A N D W E AV E R S 117

personal relationship between teacher and pupil is a mode of textual transmis-

sion of authority via the voice as the oral medium.16 It is common to say that the
genealogy of this transmission dates back to the Inka ancestors.
Only when a boy or girl already contemplates things and begins to practice a
task independently from others is the borrowing isturyaña, “to study,” used. The
Aymara equivalent would be yatintaña, in which the directional suffix -nta indi-
cates that the knowledge has already entered “inside” the person’s body.
The human experience of transmitting and receiving knowledge in Livi-
chuco, expressed through verbs based on yatiña, suggests how knowledge pro-
liferates through the generations, “making people” of those who receive it. Simi-
larly, evidence for oral transmission and reception does not depend on paper or a
system of documents or formal permits as in state education, but rather on the
fertilizing recitation of that same knowledge, including the names and ties be-
tween teacher and pupil, or parents and children, through the generations.
This implies that skin, blood, and voice, as the principal corporeal elements
in the dynamics of receiving knowledge and its transmission, like heart and head,
the principal corporeal seats in the accumulation of knowledge and organization
of memory, all function within a common physiological theory of practice with
textual dimensions. The skin acts as the permeable boundary through which
knowledge flows between a person and the ecosystem, blood as the fluid me-
dium of transmission of this knowledge, and voice as its medium of dissemina-
Woven homologies of this oral textuality are found in weaving practice,
where the weaving surface is considered skinlike, threads are likened to blood,
the heart and head of the cloth become the deposits of its particular knowledge
(we shall consider their precise weaving equivalents later), and the weaving spirit
that resides in the cloth directs its voice in discourse. At a more generalized level,
the operative corporeal-oral relation to the local ecosystem is homologous with
the operative textile relation to the lands and flocks of the place, permitting tex-
tual continuity from an oral manifestation to its textile one. This implies that the
textuality of these regional ethno-physiological ideas operates not only in a figu-
rative sense but also in the constitution of an “Andean field of knowledge” (in the
sense that Bourdieu uses this term).
Other textual functions of the heart and head reinforce ideas about gendered
intentionality. For Doña Sabina Mamani of Livichuco, ideas taken down from
another person are “grasped” in the heart like grains in a storeroom, or the let-
ters of writing. As she says, “We receive these in the heart as if we were writing
there.” In a storeroom of farm produce, the foodstuffs (like ideas) are taken out
and then passed on to someone else, usually as cooked food for someone in need.

In her experience, an older woman with more thoughts deposited in her heart
(like a storeroom) is more valued as a teacher in the community.
As an older man and title bearer, Don Domingo Jiménez perceives the heart
as a storeroom not of raw foodstuffs or cooked food under the charge of women,
but rather of written documents, wills or testaments (tistamintu), as part of the
state juridical apparatus that title bearers like him have had to handle since the
colonial period. As he says, ‘The heart is the testament and the testament grasps
well; a person who doesn’t have a good heart cannot think at all.” Similarly he
interprets the head as a “customs house” (aransila). As an older man, he would
have a storeroom of juridical knowledge, just as Doña Sabina would have one of
farming produce.

L earning through the S enses

Knowledge comes to be deposited in the heart and head through the senses
in the constant dynamics of internalizing and then externalizing thought. A great
deal of importance is given to a child’s learning through sight, simply by watch-
ing (iñaqaña), especially in weaving. As Don Juan Maraza commented in the Livi-
chuco workshop: “Girls weave like their mothers, they copy just by looking at
the designs, they look from one side and the other, and then they just weave;
they don’t intellectualize like teachers do when they teach, rather they learn eas-
ily, watching every day.” First you watch, then you internalize this knowledge,
and then you externalize it again in the act of weaving. Hearing is another of
the senses that facilitates the transmission of knowledge. In order to “become
a real person” it is imperative to know how to listen (isapt'aña) to the advice of
older people given in formal contexts, for example at the wedding feast, when
they reinforce hearing with some additional tactile gesture, touching the head
and shoulders of the couple. Alvarado (1998) suggests that such advice-giving
sessions are transmitted like external wrappings that the couple then has to inter-
nalize and assimilate.
Doña Sabina emphasized the reception of knowledge through hearing, the
heart, and the head, distinguishing this way of learning from reading, where
you don’t receive it well. The perspective of older people like Doña Sabina fa-
voring oral transmission over writing for its greater degree of corporeality and
spirituality (to do with ajayu and ispiritu) is comparable to the position taken by
Derrida among present-day philosophers, or by Socrates in his famous dialogue
with Plato in the Phaedrus. Socrates, who favored orality, argued with Plato (the
writer) about the inferiority of writing as an artifice of memory and learning. For
WA R R I O R S A N D W E AV E R S 119

Socrates, writing could only awaken reminiscences without replacing the true
discourse found in the psyche of a wise person, which could only be transmitted
through oral interactions.
As seats of the reception and transmission of knowledge, heart and head or-
ganize not only personal intentionality but also that at a communal level, in the
ayllu body politic. Here analogous somatic functions are applied to the ayllu com-
munal storerooms of knowledge, to be transmitted through the communal path-
ways in the fulfilling of duties (cargos).17
The masculine domain of politics tends to be concerned more with “talk of
the head,” from whose seminal condition unfolds all the rest. Hence an intelli-
gent person who represents the community is considered “good-headed” (wali
p'iqiwa). Don Donato Inka of Livichuco considers his grandfather, Qaqachaka’s
great apoderado Don Feliciano Inka Maraza, to have been “one good head” (mä
suma p'iqi) who raised the whole community on high with his knowledge and rea-
soning. These “great heads” of the community are recorded, in turn, in the liba-
tions directed by the ayllu heads (original authorities called jilanqu in Qaqachaka
but in other places p'iqi, “head”).18 The jilanqus, as chiefs of each of Qaqachaka’s
six minor ayllus, also serve in the annual cycle of duties (called cargos), when the
constituent minor ayllus take turns as “head,” “belly,” “hands,” and “feet,” in the
somatic divisions of the body politic.
Within this political hierarchy, gendered differences organize a parallel divi-
sion of ritual labor, the gender allocated to ritual sites, and the practices associ-
ated with them. In practice, the ayllu heads are couples: man and wife. In ritual
labor, the gendered hierarchy of ritual sites paired as guardian mountain (uywiri)
and corner shrine (iskina), at the levels of household, community, and ayllu,
obliges men to tend the former and women the latter. The guardian mountain
as the masculine caregiver over ayllu territory and its flocks receives the heads
(p'iqi) and hard bones of sacrifices, while the corner shrine feminine counterpart
receives the soft parts.

T h e Pat h way s i n P r a c t i c e

The ayllu body politic is organized into “pathways” (thakhi), formal institu-
tions that structure the relations of teaching and learning and of gender and age
groups and facilitate the socialization and integration of individuals into society.
The present-day pathways of Qaqachaka, as educational institutions, have his-
torical precedence in the Inkaic pathways (calles) or “visits” (visitas) described by
Guaman Poma in his Nueva corónica y buen gobierno (ca. 1613). Each instance re-

veals analogous divisions by sex and age, a sense of community obligations, a

female attention toward fine weaving, and a male one toward the expression of
physical strength.19
Javier Medina’s study of Inka pedagogy (1993) perceives each of Guaman Po-
ma’s pathways as a “subaltern caste” with its own educational practices derived
from the Inka productive system, but he does not relate this Inka model with
present-day communal practices. The historian Roberto Choque (1996), drawing
on the same source, interprets the Inka pathways as a form of “original educa-
tion” centered in the sociocultural knowledge of everyday life, at family and com-
munity levels. But he too fails to link these Inka pathways with their contempo-
rary counterparts.

E ntering the “ P ath w ay to w ard B ecoming a P erson ”

The body of knowledge presented by the older people in the Livichuco work-
shop gave particular importance to the early stages of a child’s development on
its “pathway toward becoming a person” (jaqiptañ thakhi), where physiological
and psychological aspects overlap. The maturing of the baby’s blood system is
vital for the transmission of knowledge, as is the condition of its skin, the corpo-
real seat of the soul called ajayu. In this period, the initial bodily limits of a baby
are a bag of skin, to be filled by the blood system, and then surpassed by the
voice, so that accumulated knowledge becomes externalized.
Attention is given to the initial separation of the blood between mother and
neonate, and then to the systemic circulation of the baby’s own blood (and the
flow in the direction of the heart). Special attention is given to the pulse, the
quantity of blood, its temperature, color, and density (if it is thick or thin). Once
separated from the mother, the baby must disconnect itself from the maternal
blood. Similarly its skin will close up as the “veins” of its own blood system are
A newborn babe, considered well-fed on the maternal blood in utero, is rarely
fed breast milk at birth; first, the maternal blood must be expelled in the first ex-
creta and in the discharges from the navel and the surface of the skin. At birth, a
baby’s skin is thick, soft, and tender (llullu janchi), but when fully developed the
skin becomes thin and hard (turu janchi) and can be lifted easily. “The blood now
fills the entire skin/body” and no more blood should escape.20
The blood that fills a newborn baby’s veins is thought to come from maternal
milk, and after weaning, a baby is considered to have a lifetime’s blood. This mo-
ment of “filling the body with blood” initiates other processes in a child’s devel-
opment, differentiated by gender. After menarche, young women lose a portion
WA R R I O R S A N D W E AV E R S 121

of blood with each menstrual period, so girls are thought to have less physical
strength than men, who tend to retain all their blood throughout life. It is as if to
lose blood is to lose the knowledge it is transmitting.

T he “ P ath w ay of I nfancy ”

The following stage, the “pathway of infancy” (wawanakan thakhipa), con-

cerns bodily growth and the associated development of intentionality through
corporeal comportment, in speech, conscience, and desire. At a year, a baby be-
gins to crawl about and speak and at two already understands instructions and
helps out at home. At three to four years the child learns to speak, completing
words by imitating its parents. During this period, there is the same attention to
growth within certain corporeal limits to assure that knowledge functions within
them. But sometimes development transgresses these predetermined limits and
the child “becomes stretched” (jilatatxi) exaggeratedly from one day to another,
worrying the parents since such children “usually die.”
With the hardening of the skin, and the independent functioning of the
blood system centered in the heart, the child begins to accumulate a sense of
“conscience” (chuymani), a measure of its independence outside the family com-
pound, with the rupture of the early spatial confines around mother and home.
Small children gradually “grasp” a good conscience (suma amuyunixa) at ap-
proximately twelve years old, although some might have a lesser conscience (pisi
amta) until twenty-five years or more. Again, this process involves internalizing
instructions, and then externalizing them independently, so a child is considered
to have a conscience from the moment when, at nine or ten years, “it carries the
things it has to carry and doesn’t touch things at whim.” As in Quechua, the term
ch'iki describes the vivacity of having a good conscience.21
Importance is given at this stage to the way a child expresses its intentionality
through desire and motivation, expressed by the verb munaña, “to wish, or de-
sire,” and the ways of making it manifest. A mother feels proud when her child
shows her what it has done independently, without further instruction. Similarly,
parents indulge the desire of a child who stares at something, and then stretches
its hands toward it, or when it says directly “I want this.” They comment how
“its heart really wants something” (chuymapapiniw muni, with the emphatic –pini),
relating this desire to its corporeal seat in the heart. Parents tend not to impose
their own desires on a child’s activities as is done in school, but rather wait for the
moment when the initiative comes from the child itself.
In addition, parents respect personal knowledge, recognizing that “it is not
possible to know what a child thinks in its heart or head,” and they leave it to go

at its own pace. This is why they are often unwilling to intervene in classroom ac-
tivities or a particular teacher’s techniques, thinking that the child will go forward
at its own pace as it would at home. In general, they reject “teaching” a child, in
favor of “its learning just by watching older people doing things.”
With the gradual expansion outward of the child’s sphere of activity, other
informal modes of teaching and learning begin. In the Livichuco workshop, the
mutual learning among different age groups was mentioned, as when older chil-
dren teach the younger ones at a daily level, for instance how to offer a greeting.
They learn in this way until ten years old or so and then “it is held in the heart,
and children now become embarrassed” if they are told to do things. Here the
heart is perceived as the seat of judgment.
Even so, parents recognize that the most lively children learn by themselves
just by watching, for example in weaving, while others have to be taught and
shown how to do things at each step (yatichataw yatiqi). In these moments, chil-
dren should be taught with patience, kindness, and without scolding, since as
Doña Antonia Espejo affirmed, they “lose their hearts” (juk'ant chuym chhaqta-
yasi) and “never learn after that.” There are, of course, moments in which you
should discipline a child, say a girl’s hand as she is learning to weave with a little
hit from the bone used to tighten the cloth, but this “is done only once in a life-
time.” Because of this, many parents, upon seeing the backwardness of their chil-
dren at school, blame the brusque ways of teaching in the classroom.

C hildren ’ s P ath w ay of L earning

Male and female maturity draw on the same seats of knowledge and analo-
gous ideas about closing off one wrapping of bodily limits before initiating an-
other. Here, cloth enters the learning process as the primary mode of textual
expression in both a girl’s (imilla wawa) and a boy’s (chacha wawa) pathway of
The nexus between skin and blood gives rise to the idea that a girl reaches
menarche only when she has a “hard (in the sense of sealed) skin” (turu janchi),
and only then, in later years, can she have a baby. If she is still “soft” or “tender”
(llullu), then her body might “come undone” (ch'iyanuqaña) like frayed cloth as
she has the baby.
Similarly, a girl must finish growing (jil tukxi) before her menstrual blood ap-
pears and “she purges herself.” The precise moment of her first menstrual period
“depends on her type” (in the sense of her size and disposition), be it at eleven,
twelve, or sixteen years.22 Her mother has to teach her about it and explain
why the blood has come. A girl finally stops growing completely at her “rest-
WA R R I O R S A N D W E AV E R S 123

ing point,” another corporeal limit considered to be a measure of her maturity

for marriage, ideally from twenty years on. If she marries beforehand “she gets
twisted” (q'iwjasiw).
A male’s maturity is announced when his voice breaks (literally “thickens,”
mallq'pa lankpji), externalizing his breath (ispiritu). Characteristically, this vocal
change marks his sexual maturity, and the young girls taunt him for it: “Now he
knows how to go where the young girls are! The cows do it, the dogs do it, the
cocks do it; even you do it!”

Y oung P eople ’ s P ath w ay of I nspiration

The subsequent pathway of inspiration for young adolescents is called jaynu

(or jayñu) after the weaving pick used to texturize cloth with complex three-
dimensional figures. This pathway textualizes young lives with the teaching and
learning of weaving and song for girls, and braiding and playing musical instru-
ments for boys. The practical, conceptual, and numerical steps in learning to
weave have their equivalent stages in other textual practices (music, dance steps),
as they do in alphabetical reading and writing and Western arithmetic.23
This particular pathway also relates ancestral ways with the dynamics of gen-
erational change and individual inspiration and interpretation, concerned yet
again with learning normative limits (in weaving and musical, literary, and dis-
cursive genres) and where there is space for expanding individual creativity and
expression. Once again, knowledge is first defined within certain limits, and then
these limits are broken and externalized.
Doña Antonia Espejo of Livichuco noted how a girl from three to four years
of age observes her mother weaving, attempting to make small steps with the
bone pick in a stage of watching, playing, and imitating, when she goes away for
a moment to cook. With their manual dexterity, small girls are considered to be
the best spinners; only elderly women are their equals.
At four or five years, a girl proceeds to a series of garments of lesser to greater
complexity, and with a mixture of designs, both “plain” (inaki) and “selected”
(apsu), which is more complex. At six, she advances to the complicated hatbands,
at seven to her first small food bag, and at eight to her first coca cloth. From ten
to twelve, she weaves a gamut of warp-faced designs using two and more warp
threads and, at around thirteen, advances to figures with four warps and more.
Her learning of homologies between sets of threads and established garments
constitutes another stage of internalizing the different units and their variants,
and the characteristics of different weaving genres (on a par with literary genres),
before striking off into her own inspired creations. At the same time, she must

learn the norms of combining colors (must'aña), and the textile configurations
that distinguish the weavings of Qaqachaka from those of other ayllus (and of
Livichuco within Qaqachaka), according to its territorial characteristics and the
production of food and flocks, all inspired by the colors of the sun and rainbow.
In parallel with the first flow of menstrual blood, a girl enters adolescence
capable of externalizing this physiological change in her weavings by handling
a textile esthetics based on the visual and tactile expression of the color red, in
the most difficult and spectacular figures that express her intelligence at visual,
expressive, somatic, and social levels.24 Although the great majority of girls man-
age to weave figures of ten, fifteen, and twenty warps, only the most vivacious
learn to weave figures of sixty, one hundred, two hundred, and more, in didactic
sequences comparable to the parallel steps in writing, say, a B.A. thesis. These are
the “leading, or go-ahead women” (tilantir warmi) who guide the rest. Those of
lesser intelligence “usually follow behind them,” like the animals of a flock fol-
lowing the leading animal.
A weaver increases her repertory and design complexity through inspired ex-
periments, and by copying small samples (waraqa) that she “takes down like wa-
ter” (waraqaña) from the works of other weavers, after establishing the neces-
sary agreements between them. At the same time, a young weaver develops the
complexity of her own discourse about textiles and her ability in song, narratives,
and other oral genres. Like weaving designs, verbal art is developed by learning
formula-like units and then developing the connective sections, strophes, and so
on, in a certain order, according to the norms of prosody and rhythm. At the
same time, each young woman develops her own characteristic style, founding
her song in the territory and ritual sites she knows from her herding routines.
The specific techniques of weaving and its associated practical activities
(herding, shearing, and so on) give a daily impulse to traditional life very differ-
ent from that found at school. Doña Antonia records this experience of wanting
to weave “in her heart,” another illustration of the importance of desire in com-
munal learning: “I used to get up early, I was always spinning, and had to look for
wool to spin and weave. It was a pain not being able to finish quickly; I wanted
mantles, and I’d say ‘I’d like to weave,’ and I’d always be with this [thought] in my
Learning to weave takes priority in other key moments of communal life, for
example in the patronal feasts, when unified ayllu textile corpus, with its reper-
toire of structures and designs, is on view for everyone to share their commentar-
ies, reflections, and learning. Neither the resources of a local museum, nor those
of school library, can compete with this range of possibilities for observation and
cultural reflection.
WA R R I O R S A N D W E AV E R S 125

While girls learn to weave, boys learn to braid, externalizing their physical
strength in herding and farming tasks in a woven expression of the corporeal ele-
ments (muscle, tendons) that “grasp” bone, the seat of male strength. Boys begin
to braid small garments such as belt ties. They go on to weave slings (q'urawa)
of four strands (misminta) for herding sheep. Then they advance to others of six,
eight, ten, twelve, even twenty-four strands in the most solid and attractive slings
used for dances and other ritual matters. A sling of six strands is still “plain” (llanu),
but that of eight or more is “selected” (apsu) like the most complex cloth woven
by women. Characteristically, some strands disappear into the braiding structure
to form the “eyes” of these snakelike slings. They braid thin three-strand cords
and the thicker and stronger ropes of six or seven cords called allqhamari (the
mountain caracara bird) after their contrasting natural colors. And they make the
four-stranded lassos used to tie up mules, and the prestigious lassos made from a
whole cattle hide, braided from twelve strands each a fathom long (then doubled
over to form just six strands).
Young men used to learn to make the rough homespun called bayeta de la
tierra, and the thicker cordillate, vehicles for learning another counting technique
by portada in which a typical cloth with a doubling method reached thirty-two
portadas in width. But older people complain how “now they go to school, they
no longer know how to do it well.”
For men, the equivalent to a woman’s mantle is the knotted cap (ch'ulu) to
keep the head warm, where they come to handle hundreds of threads in apsu
structures, almost equal in complexity to the weavings of young women. The
textile logic learned through these weavings and braidings provides men with the
knowledge of elementary units and their modes of expansion to other textual
practices, in the structuring and counting of musical practice, choreography, and
At a bodily level, men externalize their physical strength in adolescence
through making the kinds of sling that “will give them strength,” as each strand
“has ispiritu” and so “gives ispiritu to the whole body.” Don Domingo compares
each strand with the cartilaginous tendons called “nerves” (anku) that grasp bone.
He compares a sling as a whole to a rib with a cartilaginous section at its ends. As
he expresses it, “If it weren’t for the tendons, we’d come undone, wouldn’t we?”

C onflicts bet w een the T w o P ath w ays

When boys and girls enter school, they perceive the conflict over different
textual practices, those of the community and the school, in terms of a textual
struggle between “two pathways” (pä thakhi). This is even more intense for girls,

since in the classroom weaving is relegated to a secondary activity and reading

and writing given utmost priority. This leads them to complain how, just as they
are at their very peak in weaving, they barely learn to read and write with cur-
rent teaching methods, and feel unable to comprehend meanings or achieve self-
Don Feliciano Maraza explained how the conflict begins when children go to
school for the first time while still learning the ways of the community. It intensi-
fies when they reach twelve years, as they “must now know everything,” spinning
and weaving mantles, hatbands, belts, and bags, as well as reading, and it intensi-
fies even more between twelve and sixteen, as girls must go on learning both tex-
tual practices, “so that our customs aren’t forgotten.” For Don Feliciano, weaving
is a part of the “grandfathers’ and grandmothers’ pathway” (awil achach thakhi)
of laws and custom that are passed on orally: “We are following what we heard
from the grandparents.”
Like other comunarios, Don Feliciano denominates weaving as “our path-
way” (thakhsa) and reading and writing as “school knowledge” (iskuyl yatiña). At
school, “children forget the other pathway we call ‘culture’ and is our own.” This
is cause for concern in a place like Livichuco, where the APSU (Artesanías Para
Seguir Unidos) project initiated in the 1990s emphasizes the cultural importance
of local weaving and has brought in a substantial income for more than fifty fami-
lies in the last five years. If young people do not continue these practices when
they are adults, developing only what they learn at school, then the tendency “to
go far away from the community” will continue.
The hope in Livichuco is that the positive aspects of intercultural bilingual
education will reconcile the different pathways of textual practices, as they will
languages. In the main school of Qaqachaka, a set of vertical looms for mak-
ing carpets has been introduced. And in some branch schools, the more practical
among the teachers are teaching pupils to weave on traditional looms. This kind
of intercultural and “intertextual” dialogue is helping to bridge the present differ-

” B ecoming a P erson ” at M arriage

“Becoming a person” (jaqichasiña) at marriage marks the transition between

the pathway of young people (jaynu) and that of a full adult (thakhi). This entails
the same definition of limits to internal development and then the externalization
of accumulated knowledge in another stage. During this transition, the “natural
person,” constructed socially through the action of others, comes to externalize
the accumulated knowledge for the benefit of others. In a weaving idiom, you
WA R R I O R S A N D W E AV E R S 127

learn to “unravel yourself ” (llawuraña) in order to become a “social person” in an

active sense, at home, in the community, and to a certain extent in the state.
This unraveling of a married person demands not only the externalizing of
knowledge, but also the blooming of their family life through the subsequent ac-
cumulation of goods and the multiplication of the various wawas (humans, ani-
mals, and food produce) that they share with others.
We would argue that, in the past, this stage of life was founded on a key mo-
ment in the cycle of ontological depredation, when the energies of a defeated en-
emy were appropriated and converted into one’s own. In its unraveling, marriage
rituals even now allude to an ideal couple of warrior and weaver, experts in the
creation of a new harvest of babies from the spoils of the dead.
A similar idea occurs in the series of drawings with accompanying text con-
cerning the Inka pathways or channels of learning in the Nueva corónica attributed
to Guaman Poma (1613, f. 193–234), which highlight the ideal couple of warrior
and weaver and their characteristic tasks—acquiring a trophy head and weaving
(figure 13). Guaman Poma’s reading of the information about the pathways (as in
his whole chronicle) from a kipu (or various kipus), takes this ideal couple of the
first path as its point of reference for reading all the others.25
Just as the husband of Guaman Poma’s first path is an Inka warrior, so a mod-

Fig. 13. The ideal couple of warrior and weaver (Guaman Poma [1613] 1989, f. 194, 215).

ern husband in Qaqachaka, in his role as son-in-law, must show his bravery. He
is considered to be a “seated condor” (kuntur chuku), since “the condor as mallku,
or lord, has to defeat all the rest.” The value of a son-in-law, though, lies not only
in his strength and bravery but also in his manual labor, and sayings about the
ideal qualities of a husband refer to all these characteristics. Elvira Espejo’s com-
ment “If your husband will be a good worker, well, this will be fine” (chachma yän
lurirïni, ukay walinixa) uses the term yän(a) to refer to the husband’s ability to ac-
cumulate goods and chattels in all directions.
His comparison with a seated condor draws on the common tale of the “Con-
dor and the Fox,” when the condor’s greater bravery and endurance permits him
to defeat the fox. As in the tale, a real son-in-law “warrior condor” has to prove
himself, “as in warfare,” before the foxes (the menfolk called larita on the female
side of the family), as well as intercepting with sticks and darts the live foxes that
pester sheep. As Don Domingo affirms: “That’s where he conquers the fox, in
war, then he wins over the pumas, even the pumas.......and only then is he called a
warrior. The warrior son-in-law, he’s a good son-in-law, he’s really a person, they
Likewise, a weaver makes a good daughter-in-law, as she is a real “woman”
(warmi), whereas a nonweaver is both “useless” and “unproductive” (ch'ilima).26
As Don Domingo says, “She might be well endowed, but she doesn’t have good
designs.” During courting, young men comment on who will make a good wife.
Musical genres such as the wayñu give the maximum opportunity to express
these abilities in verses that comment on the quantity of woven garments, espe-
cially the mantles she wears.
After the wedding feast, the couple must carry out a customary period of
bride service of two weeks for their parents-in-law, although in the past it lasted
much longer. In this period, the son-in-law still has to work the land of his parents-
in-law, carry kindling, build walls, and show his capacity as a man in the rapidity
with which he can cut off the heads of sacrificed animals in the so-called “cutting
of heads” (p'iq kachiyaña). Meanwhile, the daughter-in-law must show her ability
in weaving the finest and densest cloth “that even water cannot pass through” and
that can double up a large needle, the standard of weavings used in warfare. Be-
sides this, she must show her skills in preparing food, in cooking, bringing water,
and grazing animals.
At a textual level, these social demands on a young couple are transmitted
through sessions of advice giving, called ixwa, from older people at the wed-
ding feast, and only on receiving this advice “do you become people” (ixwa jaqi-
chasiñana). Again the nature of this transformation depends initially on a period
when the limits of knowledge are defined and internalized, during which the
WA R R I O R S A N D W E AV E R S 129

couple is identified with the spirit of a head/baby (the “pair of doves” or par de
palomas). Then the couple externalize this accumulated knowledge by expanding
the domain of their matrimonial home.
Each older couple (parents, godparents, and ayllu authorities) develops a rep-
ertory of recommendations for this occasion that ideally reaches a total of twelve,
performed in an ordered sequence. They should begin with the ideal qualities of
the young son- or daughter-in-law, the tasks that each should accomplish, the ma-
terial and spiritual changes that they should experience on achieving the status of
a married couple (and becoming a pair of doves), and finally the new obligations
to their parents-in-law, ayllu, and gods. These performances echo the Inka path-
ways or channels of learning of four hundred years ago, by following an Andean
sociopolitical hierarchy.
Marital advice also concerns the textual basis of these changes. The woman
is referred to as a weft thread and the man a warp, as at marriage the ayllu textile
repertory is finally completed, when the warp is combined with the weft. This
happens not only in weaving but also figuratively in all dimensions of social life.
In another variant of the same idea, a single person is considered to be a “loose
thread” (ch'ulla ch'ankha) while a married person is “plied” (k'anthita).27
The process of becoming plied has other textual aspects, since after a short
time, the couple become “two braided words” (khiwi ch'ankha) and after a little
longer, “half a ball of wool.” After marriage, the couple also become a “plying
spindle,” wound tight and hard, because “it has strength,” whereas a single per-
son’s thread is still considered loose and weak. You can know a couple that is sim-
ply “living together” (irpt'apita) without marrying, because the thread “knows”
that it is “just so” and not well plied. The woman will ply her threads in vain as
they will “just begin to unwind again.”
The transition of two young single persons to “become people” (jaqichasiña)
at marriage also entails a “spiritual” birth, when they become “like babies” (waw-
jamaw) leaving the flesh and bone of their corporeal parents to accept the author-
ity and “other life” (mä wira) of their spiritual godparents. This transition per-
haps derived, in the past, from the metamorphosis of the trophy head (which the
son-in-law captured to prove his bravery and strength as a warrior) into a baby
(wawa), whose pathway the couple were obliged to follow. When the godparents
of marriage present gifts of garments to the “pair of new babies,” their wrap-
pings define the limits of their new condition. An analogy to butterflies marks
this incipient spiritual transformation, when the godparents’ gifts of wings (chiqa)
to the new nymphs facilitate their transition into butterflies that fly.28
On internalizing the spiritual energies (ispiritu) of the trophy head, the new
couple leaves its condition as single “natural persons” of flesh and blood to be-

come “social persons” and more spiritual beings. By unraveling themselves as

persons in later years of married life, they gradually externalize the energies of
the head/baby under their charge, above all on entering the new social pathway
that entails their communal obligation to “carry heads.” At the same time, the
spiritual power of the trophy head/baby generates a fertile period that stimulates
their unraveling in all directions of the ayllu territory and beyond, as married per-
sons together with their contingent of new babies (human, animal, and vegeta-
ble), and in the modern world with a plethora of material objects.
The “desire” mentioned in commentaries on the unraveling of different path-
ways is perhaps part of the same process of externalizing captured forces, a sign
of the restlessness of hands and sight to expand the personal domain, just as this
expansion is accomplished by imitating the former pathways of others.

T he G reat P ath w ay of O lder P eople

With the accumulated experience and knowledge gained from the pathways
of youth and marriage, older people, in their role as communal teachers, pass on
to the “great pathway” (jach'a thakhi), with its far-reaching and imposing aims.
As the “pathway of the grandfathers and grandmothers” (awila achach thakhi), it
dictates the sociocultural dimensions of ayllu life. As the “pathway of the Gods”
(yusa thakhi) and “pathway of memory” (amtañ thakhi), it dictates spiritual values
and practice. And as “consuetudinary law” (thakhi) and “custom” (kustumri), its
political dictates still function in parallel with the laws of the nation-state.
On initiating this pathway, older people enter formally into the duties (car-
gos) of communal authorities, in which they prove themselves to the community
before passing on to the next duty. Beginning as the traditional school mayors or
postilions, they proceed to become ayllu heads, and culminate as the greater may-
ors of all Qaqachaka.
At a formal level, other selected men pass on to the “other life” by serving the
community as “wise men” (yatiri), while selected women serve the community
as “midwife-healers” (qaquri-yatiris), though the great majority of women help as
midwives after menopause. At a semiformal level, all older people serve their fam-
ilies and community giving advice in the wedding feasts to the new communal
duty holders and feast sponsors, to whom they teach the pathways of libations,
and to the mourners at wakes, to whom they teach acceptance of their new con-
dition. In their capacity as communal teachers, they are called “pathway guides”
(thakhichiri) or “leaders” (irptiri). Finally, at an informal and everyday level, elder
people guide children at home in the daily tasks of farming and herding.
According to Juan Maraza, the thinking of older people had a special rele-
WA R R I O R S A N D W E AV E R S 131

vance in the past: “it was braided well with the pathways of the ancestors and
their luck,” and so they knew when to accomplish rituals at the communal sites
(uywiri and iskina) and their obligations to the school, working “for the common
good, and with a single heart.” With modernization, however, he notes a change;
though they continue to do these things, “now it is less so, and each one does it
on their own account and within the confines of their own home” rather than
the wider ayllu setting.
Given the widespread ability of all older people to interlace thought and
pathway, Rivière (1995) has argued that even the shamanic pathway of the “wise
ones” is not simply the prerogative of specialists. It is rather a “pathway in com-
mon” among all the older people who participate in the religious, political, and
ritual duties of the place together with the ritual specialists, tending natural life to
guarantee production, luck, and communal health (1995, 112).
The comunarios of Livichuco agree that the shamanic and spiritual powers
(above all lightning) “found the legitimacy” of the elders by recording its found-
ing myths, and so establishing a social and symbolic mediation with the gods and
ancestors, guaranteeing the success of the agricultural cycle. The strength of
knowledge (qamasa) centered in the “pathways” comes ultimately from the gods
and ancestors, and so in managing them, older people, like the ritual specialists,
come to manage the luck of the whole social group (Rivière 1995, 112).
In Qaqachaka, the symbolic center of communal strength is attributed to
the Guardian Mountain (uywiri) of the place, the communal “head” where ritual
pathways converge in an incessant to and fro of tracks inscribed into the earth
and into memory. The rituality of offerings made there, permits communication
between the worlds of above and below, present and past, the living and the dead.
At the same time, they articulate the necessary ties between sites of offered heads
(uywiri) and the communal heart (of which Don Juan Maraza spoke) so as to re-
lease the rains (cf. Rivière 1995, 113).
As authorities in the ethical backdrop of communal education, older people
and local specialists value their own textual practices as a necessary part of the
continuity of their obligations to former Andean states. As a corollary, they are
most conscious of the failure of traditional state schooling to consider and re-
spect these practices. Their own educational system functions within a network
of pathways that constitute the culture, law, and custom of the place, and through
communal education with its organization into groups according to gender and
age, this is integrated into the wider state system. This ethical and moral system
draws, in turn, upon its own deontology of family necessities and communal,
state, and spiritual obligations.

These discursive practices emerge from a regional theory of knowledge en-

acted through the marital institution of the ideal couple of warrior and weaver
and their initial possession of a captured being. As in other Amerindian groups,
the development of personal capability depends on capturing a surplus of the
exogenous energy that exists in the world. In the past, Andean cosmology or-
ganized humans and nonhumans according to the degree of subjective activity
made manifest in depredatory capacity, in which the warrior was the prototype
of the powerful being. For these reasons, we find in communal education even
today a dynamics more corporeal than intellectual.
Although the present educational reform is an ideal historical moment to re-
duce the conflicts between overlapping educational systems and textual practices,
so far it has overlooked in its planning and execution the juridical and authorita-
tive foundations of communal practice and has simply reinforced existing textual
struggles in rural areas as well as generating new ones. In comparison, the local
dialogues initiated by comunarios themselves are probing the roots of the prob-
lem and seeking new pathways toward a solution.

Th e C yc l e s of Li bations i n
Scho o l Ri tuals

Soldiers, as well as the whole gamut of military history and activities, are regularly glorified
in school contexts.

Aurolyn Luykx, “Discriminación sexual y estrategias verbales

femeninas en contextos escolares bolivianos”

Could a Nation, in any true sense of the word, really be born without a war?

Michael Howard, The Lessons of History

The school premises, as a primary mediating space between the community and
the state, articulates a historical stratification of multiple intercultural encoun-
ters. What happens, then, in the ceremonial activities of schooling organized by
the state in rural areas to instill in local participants a sense of nationhood? More
specifically, how do community members, pupils, and teachers interpret school
rituals from their differing perspectives and varying degrees of involvement as
citizens in the Bolivian nation-state?
Given the Andean theory of textuality we posited earlier, let us now examine
what role those national rituals with their host of symbols (flags, uniforms, tri-
bunes, parades), centered in the school, play in these overlapping interpretations,
and what parallel social memories they might draw on.


Textual and Corporeal Struggles:

T h e S c h o o l a s a M i l i ta r y C e n t e r

In Imagined Communities, Benedict Anderson concludes that a vital part of the

formation of nation-states in the nineteenth century depended on textual forms
and processes of unification. This enterprise demanded the interchangeability
between men and documents in the state bureaucratic apparatus. Homines novi
(new men) were recruited as state functionaries to implement its intentions to
the letter, according to a shared language at state level (1991, 55). In parallel, the
interchangeability of documents demanded the standardization of documentary
language and forms, above all at a written level.
However, given the long history of massive illiteracy in countries such as Bo-
livia, the printing industry could not control the state bureaucratic apparatus sys-
tematically, nor could the state develop an educational system sufficiently to ac-
complish the task of creating a literate nation. In situations like these, Anderson
points to the historical importance of training bilingual teachers as key interpret-
ers in the genesis of nationalist movements, especially in the ambit of illiterate
peasants (1991, 54). This enterprise had not only secular but, as Anderson argues,
ritual functions: the state selects a regiment of possible functionaries who go an-
nually in pilgrimage-like rites to state centers to meet with other colleagues and
so ascend the rungs of the state system (1991, 54). Once trained in the language
and apparatus of state ritual and ceremony and hardened by the demands of this
rite of transition, the functionaries return to the periphery to enroll another gen-
eration of recruits in the same instructional system. The nation-state is thus re-
produced through this very instructional system.
Even so, Anderson’s model of the ritual and ceremonial apparatus for repro-
ducing the state system at a local level does not give us the theoretical means for
understanding the nature of these rituals from the communities’ point of view.
There is no doubt that school rituals are eminently bellicose, whether these are
rites of the nation or the daily school ceremonies. Each morning, the end of the
daily school rituals in Livichuco is accompanied by the sounds of the nineteenth-
century march of the United States Marine Corps, “From the halls of Monte-
zuma to the shores of Tripoli .....” or another North American march from the
First World War, “The Stars and Stripes Forever.” We asked the teacher where
his inspiration came from, and he replied that he wasn’t sure but he preferred the
Marseillaise. How can we understand these bellicose rituals in the context of the
formation of the nation?
In Blood Rites: Origins and History of the Passions of War, Barbara Ehrenreich
criticizes Anderson (as well as Hobsbawm) for having considered the forging of
T H E C Y C L E S O F L I B AT I O N S I N S C H O O L R I T U A L S 135

the nation as a purely cognitive task of the imagination, without paying atten-
tion to the function of war as a vital factor in the forging of societies. According
to Ehrenreich, this omission can be traced to Marx, who overlooked the critical
nexus between economy and warfare in a historical moment in which the mili-
tary as a class was suffering a crisis of confidence (1997, 196). For Ehrenreich, the
nineteenth century was a time of experiments in the sublimation of self to the
new demands of the greater national community. The men of the new nations
experienced this greater totality through the barracks and the celebration of the
national “imaginary past.” In practice, then, the nation expressed the unity of
the warrior lineage; recall that a vital part of the democratizing process was “the
right to bear arms.”
In many countries, especially Andean ones, state educational rituals empha-
size the child’s incorporation as a new citizen, in this military sense.1 Since inde-
pendence from Spain (1825), the status of citizenship has depended on carrying
out military service.2 The military techniques still used in Bolivian state schools
are part of a policy introduced in 1907 requiring the compulsory military service
of rural youths, aimed at converting them into new citizens.3 An important part
of barracks life is civic education and the incorporation of rural youth into the
nationalist consciousness. As we saw, these policies were reinforced after the War
of the Chaco (1932–1935) and then in the heyday of the MNR party (1952–1964)
with the formation of peasant militias.
Bolivian military history, then, is closely interwoven with the history of rural
education and the formation of the nation, at least for men. In The Citizen Factory,
Luykx shows how teacher training institutes (institutos normales superiores) man-
age this institutional production of citizens by transmitting military educational
history to each new generation of teacher trainees, and hence to rural schools
(1999, 128).
Yet, to understand the community’s point of view concerning the rituals per-
formed on the school premises, we have to go back much further, to the very
roots of national history. Here, we find that rural stereotypes of the nation are
founded in the “monumental time” and indigenous armies of former Andean
states, which serve as prototypes for more recent struggles over land.4 In prac-
tice, there were still indigenous armies when the Bolivian Republic was formed
in 1825. Even in 1829, Mariscal Santa Cruz was compelled to create two paral-
lel military forces, perhaps because of his awareness of this underlying military
tendency.5 During the War of the Pacific (1879–1883), too, high-ranking officers
commanded indigenous armies, in Inka style. Two decades later, the same hap-
pened under the leadership of the indigenous general Zárate Willka (1899) in his

struggle to defend community lands against incorporation into large landhold-

ings (latifundismo).6
This duality in practice concerning the military and schooling history of the
nation underlies the comments of graduates from the teacher training institute
of Caracollo destined to work in rural regions of Oruro. For them, the military
traditions of schooling are essentially foreign, coming from Europe (and the in-
fluence of the Belgian educator Georges Rouma) and with immediate roots in
the First World War and more distant ones in the European military tradition of
the medieval period. While they admit a more regional input in the militariza-
tion of schooling (through the famous ayllu-school of Warisata), they view the
military aspects (such as the school uniform of white smocks, and compelling
pupils to goose-step) as “invented tradition.” It is the same with raising the flag,
although again they are ambiguous about which of the flags should be raised: the
Bolivian national flag or the Andean checkered wiphala.
The militarization of schooling is reinforced by the fact that almost two-
thirds of teacher training institute students have passed a year of military service,
and almost all rural male students have. A certificate of completed military ser-
vice is an admission prerequisite for all young aspirants who are not physically
handicapped. The teacher training institutes of the Bolivian highlands, then, are
an integral part of this same military tendency.7 In spite of the many hardships of
military service, it is still a prerequisite for Aymara men to achieve adult status.
Aymara parents do not easily accept a son-in-law who has not completed his mili-
tary service before marriage, considered the most important step toward achiev-
ing full adult responsibility and identity.8 So, the meaning of barracks in the life
of young men is both symbolic and pragmatic, a kind of men’s hut “which marks
the end of childhood and the initiation into manhood” (Luykx 1999, 114).
Luykx uses a Foucaultian model to compare the life of a teacher training in-
stitute intern with that of the barracks (1999, 118). As “regulated subjects,” teach-
ing trainees are like interns in a total institution, bodies “dispossessed of their
roles” (in Goffman’s terms) and watched over by a Panopticon-like order. Luykx
perceives teacher training institutes as the sites of corporeal socialization; of disci-
plining and punishing to produce docile, controlled bodies; as a “political field” in
which the body is contained and suspended between relations of power.9 Draw-
ing on Bourdieu, the schooling system is a habitus with the same goals, a system
of durable dispositions that integrates the past experiences functioning in each
moment as a matrix of perceptions, appreciations, and actions to achieve the lo-
cal reproduction of state objectives.10
But again, these theoretical interpretations fail to locate this struggle in its
historical context; neither do they explain the motivations of the social actors,
T H E C Y C L E S O F L I B AT I O N S I N S C H O O L R I T U A L S 137

nor the nature of the enemy against which they have to struggle. Closer to the
lived reality of this struggle is Luykx’s perception of the school milieu as a site
of resistance, in which the social memory of the indigenous struggle for educa-
tion is sensed as a historical watershed of communal sacrifice, a popular mem-
ory marked by state violence and the creation of Andean martyrs for indigenous
education.11 This may be so, but the militarization of male education has always
been a fundamental part of Andean educational systems, long before the nine-
teenth century and the founding of the republics, as we saw in the Inka ways of
Guaman Poma, and the present-day pathways of Qaqachaka.12
It seems more plausible to us that the popular memory of struggle and resis-
tance of alternative communal history has infiltrated the state educational sys-
tem. As a “house of struggle” (ch'axwa uta) the school itself seems to reshape
the internal wars of state colonization (as a consequence of conquest), while ex-
pressing the dynamics of the body politic and the fragmented and headless sov-
ereignty of a nation such as Bolivia. It is this set of social memories, founded in
Andean monumental time and an alternative model of the state, that we felt with
such poignancy in Livichuco during the rites of 2 August, Campesino Day, and 6
August, National Day. Let us turn, then, to the comunarios’ own interpretations
of school rituals.

School Rituals

Sources that describe school rituals in the Andes, or evaluate local interpreta-
tions of them, are exceptionally few. A short essay by Ortíz Rescaniere (1973), de-
scribes a contemporary myth of the school in the Cusco region against the histor-
ical background of a textual struggle between the Inkas and the new evangelizers
of the Word of God at the beginning of the colonial period. Llanos (1992, 1998)
describes briefly a school ritual in the region of Chari (Bautista Saavedra Prov-
ince) to initiate the annual activities. Additional insights can be found in historical
studies such as Platt’s (1993), which highlights certain continuities between the
symbols of the Inka state and those of the first decade of the republic, embodied
in the personage of Simón Bolívar. But little more than this has been written.
The oral history of Qaqachaka, with comparative commentaries from north-
ern Potosí and Carangas and from local teachers, is a much richer source of the
local ideas underlying school rituals and the textual theories that generate them.
These ideas find their most concrete expression in the daily and ceremonial prac-
tice of school rituals, founded in the very elements of school construction (flag-
pole and pyramid, rostrum and proscenium).
The school year begins with the opening rituals and the initiation of school

activities in February, when the rains end. These are followed by the rituals of
Pentecost in May, during the harvest, and then the patriotic rituals of August,
when the new period of sowing begins. The rituals at the end of the school year
accompany the opening of a new cycle of rains.

T he “ O n - guard R ituals ” in F ebruary

At the beginning of the school year, the reception of a new teacher in any
school in Qaqachaka demands an “on-guard ritual” (rito de posición), a military
term used when a soldier stands on guard to face an adversary. During this cer-
emony, the local authorities must instruct him (most rural teachers are men) in
his duties and the history of the place (of the school itself, and the origin of all
its elements, included the teacher himself ) through a series of libations. Then the
teacher must confirm the fulfilling of this ritual by going to the district director
and confirming his presence in the community.
The on-guard ritual formalizes the school’s ties to the state military-educa-
tional hierarchy. This is even more pertinent in the lower levels of the ayllu (in the
zone of conflict) where the teacher is presented with a Mauser rifle by the comu-
narios during the same ceremony, to ensure his status as a local military and edu-
cational chief, and “because he must be courageous and not fearful.”

T he R ite of I nitiation of the S chool Y ear

Classes in all the schools of the region begin with iskuyl anxata, “the rite of
initiation,” practiced since long ago by the ancestors “so that the children read
well, and also grasp well what they’ve learnt.” As we saw in chapter 4, the term
anxata signifies “putting animals to graze.” In the context of the rite for initiat-
ing classes, the pupils, as lambs, are being put to graze “on letters.” They must
extricate themselves from their own animals to go to school, where their parents
leave them on the new pasture of paper (the new vegetal covering) and letters
(the seeds that will sprout the following year). The word anxata relates the rituals
of schooling to the principal communal ritual sites in Livichuco: the masculine
guardian mountain (uywiri), also called anxata, and its feminine counterpart (the
corner shrine, called iskina), in a nested hierarchy of offering places, and a cycle
of blood offerings to raise the community’s good health.
The school authorities accompany the rite with a long series of toasts to the
school. To this end, the school mayor (alcalde escolar) is charged with making liba-
tions for the pupils all night long together with the president of the school com-
mittee, which is made up of the school assistant (auxiliar escolar), speaker (vocal),
T H E C Y C L E S O F L I B AT I O N S I N S C H O O L R I T U A L S 139

treasurer (tesorero), corregidor or magistrate, and deputy officer (agente). This ritual
can only be carried out on a Wednesday or Thursday, which are “good days.” Pre-
paratory rituals begin in the afternoon. The next day, at about seven at night, the
sacrifice, butchering, and blood offering (wilancha) of two rams (either a sheep
and lead llama, two sheep, or two llamas) are carried out in the name of the
school. The parents explained how the two animals are shut into the school com-
pound beforehand, as “these are the children,” in a clear reference to the sacrifi-
cial nature of parents giving up their children to the school as a form of tribute.
This part of the ritual is called iskuyl apxata, “gift to the school.”
Don Juan Maraza regards the function of this ritual as “accommodating
the school to the culture of the place.” That is, state interests are incorporated
into the practices of the community, while the community is reproduced in the
school compound. On the contrary, “if they were to forget these communal rites
and customs, then the pupils or school authorities might die, or else the teacher
would not teach well, and the whole school year would fail.” Fulfilling the rite in
Livichuco requires that the school mayor must (ideally) offer a horned white ram
(turu), or a baby llama, to represent the pupils. The corregidor, for his part, must
offer a male leading llama (tilantiru) to represent the community.13
The school authorities lead these animals to the school patio, near to the rit-
ual site of the corner shrine, where they take hold of the llama, lifting it up by
its legs, and circle the shrine three times. (See figure 14 for a map of the school
and its ritual sites.) Then the comunarios must circle the spot three times, passing
beneath the llama’s belly, in the direction of the rising sun, before sacrificing the
llama in the same direction. These actions as a whole are called salur aptaña, “rais-
ing good health.”
In a second stage, the authorities repeat the action, this time with the ovine
ram, lifting it on high and circling the shrine three times, followed by the pu-
pils who pass in single file below the ram’s belly, giving three turns around the
iskina, always directing their actions toward the rising sun (figure 15). They are
directed in this ritual by the teacher, “with all his heart.” Later on, the pupils must
butt heads with the sheep in a competition between them (figure 16). In a meta-
phor of textual struggle, the children compete with the sheep “to measure their
strength and grasp their studies in their heads,” especially those of reading and
Doña Antonia Espejo and her husband have fulfilled the duties of postilion
and school mayor in Livichuco, observing this ritual twice from close quarters.
She holds that the pupils butt heads with the ram for two reasons, “to raise good
health, and to achieve good heads.” In “raising good health,” a wise one (yatiri)
from a rural hamlet must be called to prepare the necessary medicines. Then
Turu Mallku

kurus muqu
(uywiri) Rm. 1 2
1 Iskina

Anxata Tambo
(uywiri) ruins

Rm. Kitchen Class 1 2 Dir. House House

Uma Jalanta Uma Jalsu

(Where the water goes in) (Where the water comes out)


Fig. 14. The school ritual sites in Livichuco.

Fig. 15. The anxata ritual in the hamlet of Livichuco. Photo by Laura Pusateri.
T H E C Y C L E S O F L I B AT I O N S I N S C H O O L R I T U A L S 141

Fig. 16. A child butting his head against a ram. Photo by Laura Pusateri.

the pupils must butt heads not only with the ram but also with these medicines,
as a defense against the bad spirits of the place, and in order to get a good head.
Doña Antonia clarified that as the children butt heads with the ram, they are
given “courage” and “spirit” (jañiyu) against their adversaries. “It’s to defeat the
enemy........‘We’ll give them a head,’ they say, ‘we make the children butt heads
just so, we make them courageous and spirited.’” Still she has not specified who
is the enemy.
Customarily, in the supper celebrated that evening, the teacher eats parts of
the ram’s head, “in order to teach the children well,” while the children eat the
body. At issue is a battle in which the pupils, as soldiers, overcome the enemy, and
then give over his trophy head to the teacher, as their leader, to eat. Doña Antonia
clarified the nature of this enemy by explaining that some children cannot learn,
and must be given strength, as must the teacher in order to teach well. In her
opinion, “It’s as if the children who gather up the letters [in school] were eating
the enemy,” with the help of the teacher. She explained how, in the sacrifices, the
lead llama represents the commune and the sheep, the school. That is why the

school, pupils and teacher, eat the sheep, and the school authorities of the “little
commune” (jisk'a kumuna) eat the lead llama, all cooked by the school mayor.
After the feast, the bones, all of them, are gathered and given to a wise one
(yatiri) who prepares two sets of offerings: one destined for the guardian moun-
tain (uywiri) of the school, and the other to its corner shrine (iskina). The school
guardian mountain, located on the slope behind the rostrum, “replaces” the
guardian mountain of the whole community, called anxata. The school shrine is
located in the patio, in front of the rostrum. The heads and hard parts (bones and
feet) of the sheep and llama are offered to the male uywiri and the softer parts to
the female iskina.14 With this in mind, the wise one takes the bones “all counted
out” to be placed “to the side of the little plants” in a particular spot on the moun-
tainside. “None of the bones should be lost, they must be offered all complete.”
Through different stages in this school ritual, the sacrificial animal is incor-
porated into a wider cycle of blood offerings. In this way, the school ritual sites
(iskina and uywiri) are incorporated into the local ritual hierarchy in which the
great uywiri anxata and the iskina of the place dominate, and also into the wider
ritual hierarchy of the former Inka state, when local sites came under the domin-
ion of Cusco, at least in the memory of the participants.
According to Elvira Espejo, the nexus between the school ritual of butting
heads with the ram and learning to read and write is even more emphatic in the
pueblo of Qaqachaka than in Livichuco. There, you have to “hit the two white
sheep with books, with paper covered in letters or something else related to
the school, so that they grasp reading.” Only “by hitting them,” and their being
“trampled on,” will the pupils learn.
The person going to butcher the animal makes it chew coca first, “in order
to rest,” and wafts the smoke of burning resin over it. According to Elvira, they
make the sheep chew coca “so that the letters (litra) won’t die” and “in order to
rest, and it seems to be so.” Her brief commentary makes a direct connection
between the white paper inscribed with letters used to hit the animal, the coca
chewed before sacrifice, and the “eternity” of the letters, implying a practical link
between the productive school cycles and other local and state cycles, in which
the children act as intermediaries. This is grounded in the analogy between green
coca leaves and the green covering of the earth, or the fleece covering of animals,
and the implicit opposition between these local textual media and the paper used
at school.15 Elvira explained how, on ingesting the coca, the sacrificed rams (as
substitutes for the community children), just “rest” instead of dying, to sprout
anew the following year. She relates the resting of the sheep/children with the
fallow period of the lands in the dry season when the green pasture disappears to
T H E C Y C L E S O F L I B AT I O N S I N S C H O O L R I T U A L S 143

sprout again with pasture/fleece/letters in the next period of cultivation. “The

writings are not lost........They continue. So they say.”
Community ritual, then, relates the school cycle, taking place in the dry
months of the year, with “eating letters” as the enemies of war, in order to di-
gest them with greater ease in an intermediate “resting period,” and then “sprout
them anew” in the coming rainy season, or else “regurgitate them” in the exams
at the end of the school year. This is another facet of a regional textual theory in
the context of ontological depredation, in which the children appropriate the en-
ergies of the Other (alphabetic writing) to convert them, during the course of the
school year, into something of their own. The letters are first internalized, and
then externalized again with the coming rains, in the production of a new har-
vest of babies (wawa).
Here, the comunario idea that the children are “eating letters,” consuming
them through the mouth that usually produces speech, implies that this Andean
textual theory actually distinguishes between “text” (written and read, woven and
spoken) and “text artifact” (the perduring sense of text, as written script on paper,
woven cloth, or spoken words). In this regional theory of making artifacts of
texts, the performed text of oral speech is sensed as perduring through the letters
that stand in for speech, letters that can “rest,” “sleep,” and be “awakened anew,”
that is to say texts that can be reanimated by the letters that constitute them.16
According to Elvira, the remains of the sacrificed animals are taken farther
away in the pueblo of Qaqachaka than in Livichuco. The burying of the bones
on the guardian mountain (uywiri) converts this place into a fortress (pukara) that
protects the community. The sacrificial offerings are an “alert” that reading (as
a kind of artifact) is “interred there” and that “the children now know how to
read.” The letters, as enemies, are digested not only by the schoolchildren but
also by the principal community ritual sites. The school milieu is incorporated
in this way into the wider ritual ambit of the whole ayllu, within whose limits
the desired transformation occurs. This constellation of rituals, especially those
directed to the guardian mountain, also helps raise the children’s health: “It’s so
that they learn........Their health is there, so it’s for their health, for the health of
the children.”
In both Livichuco and Qaqachaka, the school authorities (men and women)
take their respective places to make the necessary toasts, once the offerings are
made. Both the men (seated in a raised place above the corner shrine, with their
ritual altar and cloth containing coca) and the women (sitting on the ground near
the shrine, with their own coca cloth) face the rising sun. They begin with toasts
to the corner shrine and guardian mountain, and then to the pupils under the

charge of these ritual sites. As always “they must toast well” for the community,
so that the teacher teaches well and does not renounce the job, and so forth.
The libations in the pueblo are made from what Elvira calls the “reading hol-
low” (liyiñ phuju) near the school patio that serves as the corner shrine, and from
there “they serve a round of corn beer.” She learned the whole schooling history
of Qaqachaka in the toasts made there, beginning from the city of Sucre, where
they named the first ancestors who brought the first and successive teachers, and
all the elements of the constructions “from their very foundations.”
For the comunarios, it is important to carry out a ritual of this kind before ini-
tiating any school activity or constructing any element of the school compound,
such as new classrooms according to the educational reform design. Although
they continue to use the guardian mountain of the old postal way station, a wise
one must nevertheless select and open a new ritual site for the corner shrine, clos-
ing the former one with a “ritual key.”17 According to Doña Antonia, the danger
is that “the children are like little animals” and anything could befall them—the
same for the school authorities and the classes. The comunarios enact this rite of
initiation with the intention of articulating their local cycles of production with
those of the state, and they fear that if they were to ignore it, then both cycles
would fail. This vital process of mediation between both cycles occurs through
the comunario manner of converting into artifacts the key texts in play, as the let-
ters/seeds common to both sets of textual practices that underlie these cycles of
production and reproduction: the community’s and the state’s.

O ther R ituals at the B eginning of the S chool Y ear

Two other rituals take place toward the beginning of each school year. The
first, so that the children “don’t get sad” (jan llakisiñataki) and “don’t hurt them-
selves” (jan usuchjasiñapataki), assures they have sufficient physical strength to
struggle against the enemy letters. Attention is paid to “protecting them against
any misfortune or accident” and “giving them pluck” through a mixture of me-
dicinal herbs that includes a spirit strengthener (sebario de ánimo) diluted in wa-
ter. This medicine is given to the children so that they “grasp the letters” (litr
katuni) and, strengthened by the courage running in their blood, they learn more
The second, celebrated at Pentecost, the Day of Ispiritu, at the time of the
harvest, is accompanied by libations for the pupils. This ritual, at the end of
the farming year, provides evidence of the success of scholarly efforts during the
year up till then, of having digested the letters well enough to make them sprout
with the new harvest of letters/babies. On this occasion a toast is made “for Fa-
T H E C Y C L E S O F L I B AT I O N S I N S C H O O L R I T U A L S 145

ther Spirit” (Tata Ispiritutaki) who gives them the spiritual strength necessary “to
learn the letters.”

T he L ibations for 6 A ugust , N ational D ay

The most important libations of the whole school year take place during the
rituals of 6 August, National Day, “when the earth is open” at the beginning of
sowing. In the play of discourses during these libations, the comunarios interpret
the school precinct as a part of “their own” by incorporating there “the foreign
element” of the enemy, incarnated in letters. This is done through the poetics of
libation making, which integrates the school into the pathway of “one’s own”
and into the local textual traditions. As Don Feliciano affirmed, these customs are
part of the “pathway of the ancestors,” maintained as “things that have emerged
from the place itself.” For him, the series of libations “are like school subjects,
such as the natural sciences.” Moreover, the libations “are not unitary, but fol-
low one after another,” so you must “learn them mathematically.” It is not as in
the city, where you simply repeat “good health” (salud) when toasting with beer,
and then you get drunk. Rather, you must ask the person who is “remembering”
(usually a wise one), “Now what should I toast for?”
The libations are made with wheat beer (k'usa or chicha) and cane alcohol (al-
kula). It is customary to “give beer” to the teacher, “although he shouldn’t drink
hard like the people of the place.” They take advantage of this occasion to toast
him in reform terms: “For the person who teaches” (yatichir jaqitaki).
The libations, or ch'allas, begin on the previous night (5 August), after the
social gathering and emotive parade and the various theatrical skits described in
chapter 3. Doña Antonia explained how those designated by the school must take
these toasts very seriously indeed, without drinking too much, to record all the
different components of the school precincts. The toasts are carried out within
a performative context that deictically grounds the text event within the histori-
cal and social memory of the place (in parallel with the official history of the
nation presented by the teacher). Again, their ritual nature implies a struggle be-
tween different forms of recording texts, where comunarios seek to appropriate
the power of new modes of textualization, by giving priority to letters, locality,
and cloth.
Thus, the opening libations name the outstanding mountains of the whole
ayllu (the grandfather Lord Mount Turu and the grandmother Lady Mount
Jujchu) to define the ritual territory included in the rounds of toasting. Naming
these ritual sites “aids the memory” at local and state levels, and by alluding to
the past, the libations reinforce comunarios’ strength to go forward to the greater

reproduction of the place with all its different elements. In this order of things,
the toasts to the outstanding mountains reiterate others made when young peo-
ple form couples and marry. A key idea is that the patriotic toasts directed at the
activities of the schoolchildren should help the general reproduction of the place,
its people (through the new married contributors and the new harvest of babies),
and lands.
Another hierarchical order followed in the patriotic toasts commemorates
the pyramid of power in the state educational system, beginning with the na-
tional directorship in La Paz, another for the regional directorship in Challapata,
and finally one for the local directorship in the main school of Qaqachaka.
Parents participate, especially in the toasts for their children: for the babies
and “their benches.” They request that they learn better in the coming year and
don’t forget their teachers’ explanations. As those responsible for school costs,
parents also make toasts for the annual budget (with its ritual name phaxsima,
“moon money”) and for the school materials, pencils and notebooks, given that
they are not financed easily and money has to be obtained by working in the cit-
ies or selling animals. Typically libations are made in the same sequence for the
origins of money, with toasts for Simón Bolívar and Antonio José de Sucre (given
that they “founded the Republic of Bolivia with the money obtained in the mines
of Potosí”).
Another series of libations begins for the flag and the coat of arms, although
these toasts do not distinguish between the emblems of the state and those of
the community, as other studies suggest.18 Instead a communal interpretation of
Bolivian nationalism draws on its own stereotypes and the authority of former
Andean states. As Don Juan Maraza explained, “to record the national flag is an
honor,” but more importantly when the flag waves “toward the rising sun.......it’s
good, as it’s something that’s ours and not the teacher’s.” At this moment, the
wise ones of the community comment that the flag “has ispiritu.”
The actual siting of the flag evokes older historical memories than the offi-
cial history of the nation expressed in the teachers’ annual discourse in the first
week of August. For Don Domingo Jiménez, the combined flag and pedestal are
a “stand-in for the Inka” (Inka lanti), perhaps even “the same as the Inka.” This
is why the pyramid of steps “is the same in all rural schools.” Even so, the true
Inka flag for him is the wiphala that accompanies the Bolivian flag daily in many
branch schools, and annually during the communal tax-gathering ceremony (to
such a point that the leftist MIR party has recently suggested that it should re-
place the national flag).
He revealed other reasons why the school flag is a stand-in for the Inka. The
place where the flagpole stands “would be the Inka grandfather and grand-
T H E C Y C L E S O F L I B AT I O N S I N S C H O O L R I T U A L S 147

mother,” and “so you must toast them before toasting for good luck” in certain
months of the year. In the actual time of the Inka, the place where the flag stands
was ornamented with “gold and silver” (uru qullqi) and inside there was a “good
pathway” (suma kamiri) and the Inka’s “life of riches” (qhapaq wira). His term ka-
miri denotes a “pathway,” perhaps because of its association with a birthplace
that has Inka ties.19
For Don Domingo, good luck is assured by toasting the flag in June and again
at the beginning of August. On each occasion, the rite consists in a long series of
libations that name the flag and then all its elements from above to below (like a
kipu), from the tip of the flagpole to the steps at its base. The verticality of the
sequence has to do with the earth being open in this month of the year, ready for
the next sowing.
At the beginning of August, the dual nature of communal historical memory
is particularly explicit in some more extended libations for luck to the two flags.
First a toast is made “for the wiphala” and “all its seven colors,” each one of which
has its meaning, for having emerged from the watery places where the spirits of
dead warriors reside (those called jira mayku). Yellow is “the warmth of the sun,”
dark red is “evening,” white is “to lighten the heart,” green is “the new pasture,”
and so forth. Only after recording the wiphala are toasts made to the Bolivian flag
with its three colors, toasted in detail before passing on to the flag seam, called
“golden thread” in memory of the gold and silver threads in its background.
Then they toast for the pole (lawataki) that supports the flag, and “conducts
luck internally” to the earth. For Don Domingo, the specific power of the flag-
pole is that it “helps the sowing begin each year, by introducing energy (animu)
and soul (ajayu) into the food seeds.” It also sends the solar warmth (lupi) that
emanates from Pachamama herself. He affirms: “We shall subsist with this, both
animals and people. And when we no longer exist, the children and grandchil-
dren will go on living with this.”
Finally toasts are made for the pyramid’s steps (grara grarataki), “which are
three at present, but will be some six to twelve in the future,” with their differ-
ent names (“upper step, middle step, and the six of our own”) and functions (“for
the steps of silver and gold, what a good pathway that is!”), since “one step is of
solar warmth, another is of silver, and the last one is the step of life.” The series
of toasts ends recording the way the flag flies as “it makes the spirit play” (ispirit
anatt'ayi), and “the wind plays” (wayr anatt'i) since the flag is “calling everything
The world inside the pyramid is not named as such, but some comments im-
plied its fearsome nature. As the domain of the flag concerns the health of the
children, it was the custom in Livichuco to bury something under the pyramid,

where the flagpole is, “so that the children don’t get ill.” According to Doña Anto-
nia, the yatiri in charge knows about this, putting ground corn and other ingredi-
ents there, and then you dance at the place on 6 August. Elvira Espejo mentioned
other elements put in the equivalent place in the central school in Qaqachaka,
among them a “dog’s fang” (anu kiwu) buried at the base of the pyramid to give
strength to the pupils in their struggle with the letters.

A bout the I nkaic U s n u

These unexpected commentaries about the communal meaning of the Bo-

livian flag, flagpole, and the pyramid base in the school compound, transmitted
through oral tradition, derive from specific symbols in Andean history, above all
the archetypes of the Inka state. In Nacimiento de una utopia, Manuel Burga men-
tions a similar play between national and subaltern symbols in the Peruvian An-
des, in which the Peruvian flag has this double sense, its subaltern equivalent be-
ing the lance of the Inka warrior, Rumiñawi (1988, 54–55).
A key symbol they draw on is the Inka usnu (or ushnu), a ceremonial plat-
form with steps, whose form was similar to the flagpole pyramid where the Inka
was seated when he met his military chiefs, or supposed allies, such as Francisco
Pizarro. This use is illustrated in Guaman Poma’s drawings ([ca. 1613] 1989, f.
369, 374, 398), and his comment that the usnu was a place of encounter (between
state and locality), and in Cieza’s commentary (1553) that the usnu of Cusco’s
central plaza was “a war stone.”20 The long history of these sites has an even
older Andean precedence in the pre-Inka Moche pyramids of what is now north-
ern Peru. (See figures 17 and 18.)
Zuidema’s essay “El ushnu” suggests other meanings behind the present-day
commentaries on the flag-flagpole-pyramid complex. He writes of how, in Inka
times, these constructions served as observation sites of the setting sun on two
precise dates in the year, August and April, when agricultural activities began and
ended. On these dates, the sun passed the nadir, while the moon was at the op-
posite pole, in the zenith. To make the astronomical observations, a gnomon
was placed on the usnu. This gnomon (like the modern flagpole) also facilitated
communication during the rituals with the subterranean world, as a kind of axis
mundi (1989a, 407–8) that controlled the entire textual production of the empire.
When the sun passed over these constructions, the same site, in the form of a pil-
lar, served as the “seat of the Sun.” According to Zuidema, these astronomical
observations had important state functions: “it was the most important instru-
ment that the Inkas used to achieve the integration of local calendars, tied to re-
gional farming and herding activities, within the more general and abstract sys-
T H E C Y C L E S O F L I B AT I O N S I N S C H O O L R I T U A L S 149

Fig. 17. Inka usnus.

Fig. 18. A Moche usnu.


tem demanded by an imperial organization” (ibid., 402). Perhaps the way that the
Livichuco comunarios locate the school compound and its ritual sites as local cer-
emonial outliers dedicated to the lands of the sun god, that are integrated in turn
into the more extensive state hierarchy of ritual sites centered in Cusco, harks
back to this kind of function. Within the Inka state context, the usnu also had mil-
itary functions necessary for the successful management of the empire. Zuidema
cites the colonial chronicler Molina, on military ceremonies practiced there:
And in the plaza, in its center, in the manner of a fountain, where they cast the sacrifice of
corn beer (chicha) when they came at the point of warfare, were found four hundred In-
dians around the said fountain, a hundred with their faces turned to Collasuyo which was
the birthplace of the Sun, and another hundred with their faces turned to the West, which
is the road of Chinchasuyo, and another hundred to the South, which is the road to Anti-
suyo, and a hundred with their faces toward the Midday zenith, and they had all manner
of weapons that they used. (1989a, 414)

Another anonymous chronicler confirms how offerings were made to the usnu,
and the nexus between the beginning of sowing and the tipping of chicha into its
fountain, which he describes as a place “coated with gold” (like Don Domingo
Jiménez’s memory of the present-day school flagpole base).
Other data collected by Zuidema indicate the more feared aspects of the usnu,
that out of its interior came “dangerous emanations” associated with the burial
of the ancestors and gold, which had the power to “suck” the souls of people,
making them ill (1989a, 421). Even so, Zuidema perceives in the usnu’s telluric
powers the necessary ritual criteria for opening the agricultural cycle, by “suck-
ing” great quantities of water (represented by the chicha) from the underworld in
anticipation of the coming rains in the next period of farming. These practices
also acknowledged the power of certain underworld deities, such as the serpent,
to release waters in the near future.
The way Zuidema interrelates the opening of sowing, offerings of food and
drink, and the wind and solar heat with the Inka usnu’s power to procure the
“bringing of the rains” (ibid., 452) could equally well be describing the contempo-
rary school usnu in Livichuco. Hocquenghem (1989), writing on Moche culture,
also sheds light on the possible meanings of the offerings buried at the school
usnu base, such as the dog’s fang mentioned by Elvira Espejo. In Moche culture,
canines signified “the power of the wak'a” while, together with the serpents, they
signified “the forces that animate,” being the “immortal power of the ancestors”
(204, 208). Hocquenghem also narrates how a large chain of gold (related by Mo-
lina with the serpent) surrounded the plaza in Cusco during the feast of Qapaq-
jucha and the initiation rituals of the young warriors as they danced wearing fe-
T H E C Y C L E S O F L I B AT I O N S I N S C H O O L R I T U A L S 151

line skins—how the Inkas danced and sang in the plaza at the end of these rituals
(according to Cieza), “taking a thick and colored rope, called muruurcu, that was
then deposited into the ground, rolled up like a serpent.......and it was offered a
sacrifice when.......the rains fell” (1989, 204–5).
Both the modern school flagpole-pyramid complex and the social body of
the nation that stirred on 5 August in the tricolor serpentine parade find homolo-
gies in these former Andean symbols of power. Even though these historical as-
sociations continue as lived memory only for the older people of rural areas, nev-
ertheless they are not entirely forgotten by the present-day teachers and school
directors of the region. For example, Don Amado Cahuana, director of the Co-
legio Litoral in San Miguel (Carangas Province), confirmed that his college called
the flagpole base “usnu obelisque” in acknowledgment of the same historical ar-

L ibations to the C oat of A rms and the P upils

Other school libations to the Bolivian coat of arms name all the elements
of its design, although these do not have the same resonances as those to the
flag. As Don Juan Maraza explained, “The coat of arms is just for remember-
ing.” Even so, toasts are made “for the Tata Mallku” (directed to the condor in
its design), “for the nine stars, for the llama, for the tree, for wheat bread,” and so
on. Parents record in this manner the first elements their children learn to read
at school, as “the teacher always teaches them that the nine stars are the nine
departments and that with the whole symbolism of the coat of arms (its llama,
tree, and wheat), you make the daily bread.” The Red Mountain of Potosí that ac-
companies these elements in the central part of the coat of arms is remembered,
and there are other libations for the mint (casa de moneda) and the Condor Mallku
as the national emblem. Finally they toast the outstanding personages (jilir awkis)
of Bolivia, above all Simón Bolívar and his followers. Nowadays, the comunarios
add to the list the eighteenth-century indigenous leaders Tupaq Katari and Barto-
lina Sisa, “who were not toasted before,” as part of the new tendency to remem-
ber their own.
As a whole, the subsequent toasts for the pupils follow communal education
practices, emphasizing certain parts of the body over others according to corpo-
real theories of learning and the regional norms of memorization and recitation.
So on 6 August, toasts are made “for the hands of the children, so that they write
well, for their heads, so that they receive well, for the eyes, for the mouth, so that
they pronounce well, and for the ears, so that they listen well.”

L ibations to the S chool R ostrum and

I ts A ndean M eanings

The rites directed at the school rostrum (tribuna) form another ceremonial
point of reference in the parades of August, and a pertinent nexus between social
memory, textual struggle, and learning to read. They draw on other archetypes
related to the ancestors, as the rostrum is a crucial place of transformation in the
bellicose cycle of ontological depredation, in which ancestral remains are con-
verted into a new harvest of babies.
In the community of Don Domingo Jiménez, the school rostrum is called
litrira, “noticeboard,” and the acts performed there on 6 August transform it into
a meeting place for all comunarios: “We dance there, marriages are performed
there, with the godparents and all those present.......We all used to go there to
watch, women and men, it’s usually full.” These begin with a sample of “read-
ing” (again artifactualized) in which the schoolchildren are covered all over with
paper and writing, with wings of paper (like white nymphs) as appendages. They
dance like this, with writing instruments (pencil, pen, and notebooks) hung by
threads to each side of the body, saying “It’s reading matter.”
Drawn on the paper-covered bodies of the children who “know best” (of the
fourth and fifth grades) are the different elements in the rostrum construction,
above all its adobe bricks and the mud, “that sucks and makes the rostrum stand;
if not it could just fall over.” The president of the auxiliary school committee,
who directs the teacher and pupils, also has to dance on the rostrum dressed in
a blanket (a covering of fleece), accompanied by panpipe music (although other
groups dance with band music, considered part of the world of the urban whites,
or q'aras).
Communal meanings of the school rostrum give attention to “Father Spirit”
(Ispirit Tatala) and the “mountain peaks” (kumrira) where offerings are made,
both addressees reiterating the historical memory that binds the ritual action of
the place to a wider cycle of blood offerings. In this respect, the school rostrum
has strong ties to the ritual site called taqawa, important in the interethnic wars in
Qaqachaka and in their more Christian echoes in the “church mounds of stones,”
which young girls go around on their knees as they pray aloud in the rituals lead-
ing up to Easter.22 It was on the rostrum of stones called taqawa where they dis-
played ancestral heads and the trophy heads of enemies conquered in battle. Of
all the ritual sites replaced long ago by the school, the principal one is said to be
Don Domingo confirmed that the rostrum called litrira “has the power of all
the mallkus, the great personages of the place,” because their bones and heads
T H E C Y C L E S O F L I B AT I O N S I N S C H O O L R I T U A L S 153

are “inside there.” When the comunarios make the adobes for the base of the ros-
trum, they grind the ancestral bones with the mixture of mud, and then the rest
of the platform is constructed on top. The bones placed there “give spirits to the
children.......in order to bestow on them reading power” for their school tasks, so
the complete name of the rostrum in his village is liyiñ taqawpa: “their platform
for reading.” “The great ones are there, grandfather Inka, grandmother Inka. We
put their bones inside the litrira.......the bones of the chullpas and Inkas. We con-
struct the litrira over this.......they say. ‘It will give spirits to the children.’”
In Don Domingo’s village, Carnival time around February is the moment for
celebrating the transformation of the remains of enemies and ancestors into a
new harvest of baby-foodstuffs, when the pinnacle of the rostrum is made to
bloom, and a cross is placed on the taqawa “grandfather.” They climb up there
to cast down sweetmeats, complete the toasts for luck, and direct others for the
different food crops in each household. At the end of the toasts, the taqawa is
named, as are the protectors of the “knots” of the libations, the “Lord and Lady
of the Knots” (taqaw chinu mayku, chinu t'alla), a reference to the nexus between
the place and the Andean practice of reading by kipus. Don Domingo’s commen-
taries about the school rostrum not only confirm its ancestral ties with the place
called taqawa. They also remind us of the Inka platforms sited strategically where
military exercises were carried out in the great plazas constructed not so very far
from the ceremonial platform (usnu), whether in Cusco, Saqsawaman, or Huá-
nuco Pampa.

T he S ite of t a q a w a , and the D efense of

L ands and F oodstuff

Don Domingo clarified the nature of the present-day enemy that the school-
children must face. For him, there will always be fights (ch'axwa) in the world,
and the children’s duty is to defend the comunarios: “They’ll go in front and we’ll
go behind, we say. When they are bigger, they’ll practice like the adults do.......
They study making libations: ‘Chullpa grandfather, World War,’ they say. ‘He
will fire well. He will fire first,’ they say.”
The schoolchildren themselves provide a vital link between the struggle for
education and that for lands by acting as the vanguard of defense against any in-
tent to expropriate their lands. Still, the constant “March, march” of the inscrip-
tional practices of the school compound are not only to defend themselves, but
also to defend the national territory and regain the sea beyond, which successive
governments ignore. “They’re not their lands, are they?” In this sense, the comu-
narios are more preoccupied over territorial boundaries than the state itself. And

while the schoolchildren are preparing to defend national territory (and the dis-
tant coast as well), the school taqawa rostrum has the additional function of “de-
fending the foodstuffs.”
From the community’s point of view, the function of teachers in this pend-
ing warfare is the defense of national territory and the defense of communities
against the dispossession of their food. The best teachers are those that teach sub-
jects related to the defense of ayllu lands, just as in the regional barracks, rather
than those that are weak and unstable, like the alphabetical writing of state tex-
tual forms on paper that they teach. As Don Domingo says:
He teaches well! He teaches them like soldiers. We called him profesor coronel! “Let’s even
beat Christ,” we’d say. He’d exercise and manage the students like a colonel, just the same,
Phax phax phax phax, and make them do a half turn, to the right and left. The children
wanted him to return again........Others are weak, and without energy, and they don’t
have any spirit in their hearts or their heads. Just the wind could sweep them away, like
bits of paper.

It is common to hear that the children made to march like soldiers “grasp the
letters quickly afterward.” But there is also a general consensus that teachers who
do not prepare the children sufficiently for the everyday struggles to survive out-
side the ayllu make them quarrel with their parents instead. The teachers’ insis-
tence on making boys into “men” also induces them to be macho. “Then they’ll
be lazy, they won’t obey their parents, and they won’t want to go with the ani-
mals either.” The children will have forgotten who is the real enemy: the present
state, which ignores Andean values.

Modes of Inscription in the

M i l i ta r y H i e r a r c h y o f t h e N at i o n

While communal interpretations of the bellicose activities centered in the

school precinct (in civic performances, marching in file, and singing patriotic an-
thems) focus on learning the letters of the nation, the interpretations of children
tend to focus more on their inscription into the nation at a bodily level. This is the
result of the ways in which children have been socialized into the nation-state at
Elvira Espejo told us how, at the outset, “all is just play.” As a girl she had
performed these things at school, and for her, the marching about at school was
“more like dancing,” as were the hours of civic education they underwent, rais-
ing the flag and singing the national anthem:
T H E C Y C L E S O F L I B AT I O N S I N S C H O O L R I T U A L S 155

The flag was the same for us, we didn’t know what a flag was. And when we sang, it hadn’t
any sense, it was just for the sake of singing. We’d just sing by repeating, looking about
us........The national anthem was the same. For us, it was just another song to sing........
The bigger children were singing: “Bolivianos helados [frozen].......bolivianos, bolivianos.......”
That was all we knew. Others sang: “Bolivianos, qaqachakas.......” Some sang in this way,
twisting the verses (laughing)........They still do it. It’s just like singing a traditional song,
where they invent it all, the children. It’s all invented to fit the tune: “Bolivianos, alimentos
[foodstuffs], qaqachakas” (laughing).
And the teachers listened and said, “Not like that. Louder.......it’s as if you were over-
coming [the enemy].”

But gradually the children begin to sense their transformation at a corporeal

level from community babies into “new citizens” of the nation, especially when
the teachers use brutal techniques to instill in them a memory of schooling. El-
vira Espejo recounts how in writing lessons the teachers used to hit them with
stinging nettles:
Sometimes they’d beat us. It depended on the teacher. They’d hit the very bad ones with
stinging nettles and sticks, or they’d hit us just on the hand. They’d hit some of the big-
ger ones to the point of breaking the bone weaving-pick they used. But I didn’t experience
being hit like that. They’d do it to me with stinging nettles though, when I was at inter-
mediate level. I hadn’t done the exercises as we hadn’t understood them well. The teacher
went through it very quickly and we hadn’t understood........The following day they asked
us: “Have you done it?” “We haven’t,” we said. So they told one of those who hadn’t done
it, “Go and get some stinging nettles.” And they hit us all on the hands with it. Some of
us scratched the place because it hurt a lot.......and it got infected with mange. We almost
lost our hands.

From the teacher’s perspective, this meting out of punishment served to teach
the children that they were still not hardened, giving them the pretext for fright-
ening them with other threats, such as the dead souls of the place that came out
at night.
School techniques of corporal punishment have much in common with
those of former Andean rites of passage, for example when Inka boys were
whipped with slings to make them remember.23 They also echo those of present-
day boundary struggles, when “through a ‘grinding with blows,’ Macha warriors
sought to ‘grind the spirit’ of the enemy, making him ‘smooth,’ like flour on the
grinding stone” (Platt 1987a, 90). Equally pertinent is the custom whereby teach-
ers (as combined military chiefs and Andean priests) carried a makana or “Raspu-
tin,” a ruler often decorated with special designs, to punish children who had not
fulfilled their tasks.24

On formal occasions, the children begin to feel the teacher’s rejection of lo-
cal dress in favor of city clothes, linked to their imminent textual transition from
local weavings toward the reading and writing of the state. Elvira narrates how:
“At Civic Hour, some teachers don’t want us wearing the clothes of the place......
they don’t want to see us with our country clothes. ‘You must get used to dress-
ing in another way, because now you’re literate.’”
The children begin to internalize other foreign values under the teachers’
instructions, for example the habit of eating white bread for breakfast, both to
“nourish” and “refine” themselves (in the same way as they must cover them-
selves with white smocks, the result of having “eaten” the letters).
But here, too, state attempts to regiment schooling though the changing of
clothes and food in a military context is reappropriated according to the commu-
nity’s own values. The school uniform becomes one of war. Just as the pupils’
white smocks play their part in this war of identities, so other elements of the
school uniform constitute fields of the same struggle. For example, the bags that
children take to school, and the food contained in them, have local meanings.
Elvira remembers how she took to school a small bag woven by her mother, to-
gether with the folder containing her notebook and pencil. In the woven bag, she
carried her midday snack of toasted peas and potatoes cooked in an earth oven.
In the region, peas are associated with the dead, and pea soup (alwirij lluxru) is a
traditional dish served at All Saints’ Day. This is because dried peas are hard, so
they will last out the whole year ahead as food for the dead. It is also common to
call the snacks of toasted grains “war food” (ch'axwa manq'a).
In Qaqachaka, where boundary wars reached a crisis point in 2000, women
spoke about the food they had to prepare when their men went to war. In this
case, they say you shouldn’t give your warrior husband grains toasted too hard,
as they could bring on a “rain of bullets” from the enemy. Alconcé and López
(1998) cite a similar case in Qaqachaka where women “selected any peas that had
not cooked and gave them to the men folk to wrap in their hands or tie to their
rifle.” The women’s hope was that “just as these peas that had passed through fire
and come out uncooked, so they too would resist combat and death” (139).
Don Domingo confirmed how in his community, too, the toasted peas chil-
dren carried in their bags served to defend them as ammunition. In the bellicose
context of confronting reading and writing, the children had to be strong, acquir-
ing the “strength of men.” For Don Domingo, all these customs are an inheri-
tance from the ancestors and should not be forgotten. As a parent, he is aware
that school teaching, above all the attempts to read and write, “diminishes the
strength” of the children. “It’s like nourishing them with other kinds of food”:
T H E C Y C L E S O F L I B AT I O N S I N S C H O O L R I T U A L S 157

Fig. 19. Schoolboys with their knitted caps hanging down behind, like trophy heads.

you shouldn’t give them too much to eat during the school day, just a measured
amount; “if they eat a lot, they become stupid” and can’t “grasp the letters.”
Even more pertinently, the woven bags containing peas (as “bullets”) replace
the trophy heads of the past. Formerly women wove war textiles to help their
menfolk in their bellicose excursions, and when the men returned from battle
they carried trophy heads. In the way of trophy heads, women now weave the
bags (wallqipu) children take to school. Even so, the old battle customs persist
in the way the bags are carried. For Don Domingo, a man’s woven bag with its
hanging decorations (saxsani) substituted for the trophy heads of the past, and so
they were worn hanging down behind, “to defeat the enemy.” However, the chil-
dren must not literally wear their bags hanging behind them like the men folk,
“until they have finished growing.” It is more common, then, to see them with
their knitted caps (ch'ulu) hanging behind them “instead of trophy heads,” to give
them courage (see figure 19). The woman’s equivalent coca cloth (wistalla) does
not have hanging decorations and represents the Pachamama and production
from the lands.

In this context of imminent warfare, the teacher in the community plays a

dual role, looking after the interests of the state as one more functionary and
those of the community as a “brother” in the common struggle to read and write.
As part of the state military apparatus, teachers receive their military formation
by sleeping in barracks-style dormitories, experiencing a disciplinary regime, and
learning foreign values as they become transformed into the new citizens of the
nation.25 There, they must give up their communal appearance (braids, tradi-
tional clothing) to “wrap themselves in” the uniform of the nation. This is most
obvious on formal occasions, when the men dress in suit and tie with white shirts
and the women in Western dress with a white smock on top, as if wrapped in
paper. On graduating from these civilizing centers and returning to their rural
communities, they assume the function of “motors” of progress, and military in-
structors of sorts.
Nevertheless, as intermediaries between the community and the state, teach-
ers are in the unenviable position of managing an army of children participating
in two struggles simultaneously. On the one hand, they direct the children’s ef-
forts to become new recruits of the nation, wrapped in their “paper uniforms”
as sacrificial contributions to the state bureaucracy. On the other, they direct an
indigenous army wrapped in the color of innocence, intent on defending their
lands, the frontiers of the nation, and recovering Bolivia’s access to the sea. In this
sense, the teacher divides his (or her) loyalty between this Andean army and the
present-day army of the Bolivian state.

A n d e a n T e x ts a n d T h e i r
I n t e r p r e tat i o n

C yc l e s o f M emory
The Inka’s Voice

These barbarians.......not having letters.......could not hold in the memory so many

details........To this they reply that in order to make up for the lack of letters, these
barbarians had a very good and sure curiosity, and so it was that one to another, fathers
to sons, went on referring to past and ancient things.......repeating them many times—
like he who reads the lesson from a professorial chair—making the listeners repeat them,
until they were held fixed in their memories.
Pedro Sarmiento de Gamboa, Historia de los incas

Regional ways of reading and writing alphabetic script have been a fundamental
part of Andean textual practices for centuries. Emergent in the historical strug-
gle for lands through the competent handling of written documents, these tech-
niques were later transferred to the classroom and the traditional didactic meth-
ods practiced there.
Now we trace the historical roots of these ongoing regional reading and writ-
ing practices, their status “in accordance with the law,” and their ties to notions
of corporeality. We show how fundamental ideas about memory, vocal power,
and the genealogy and continuity of life, based originally in weaving, were trans-
ferred to written texts. As a result, regional interpretations of writing are just as
rooted in the theory of textuality we propose, whereby Andean societies repro-
duce themselves (their people, land, and other resources, even communal texts

162 A N D E A N T E X T S A N D T H E I R I N T E R P R E TAT I O N

and textual practices) by appropriating the strength of the Other, and then in re-
vivifying the Other, but now as a part of the Self.

T h e E d u c at i o n a l R e f o r m a n d A n d e a n D i s c u r s i v e P r a c t i c e s

Seen from this wider historical context, the new teaching and learning meth-
ods in vogue with the educational reform are often at odds with centuries of re-
gional practices. The same happens in other daily conventions related to reading
and writing. One of the most surprising facets of the educational reform is that
with the introduction of reading corners in the new classrooms (see figure 20),
the new books have tended to be put away under lock and key in glass cases, or
else deposited in cardboard boxes in some corner of the director’s office, and only
brought out when an inspector should happen to visit the school. This is the case
in the libraries of traditional schools and the new reform schools,1 and in both ru-
ral and peri-urban schools.2 It is as if the new books embodied some kind of spirit
that has still not been brought under the control of these new institutions.
At the core of the reform proposals about reading and writing, then, the pre-

Fig. 20. A typical reading corner. Note how the books are under lock and key.

vious situation continues. Neither teacher nor pupil reads alphabetic script in a
Western sense; rather, they read and interpret texts according to other norms,
and other criteria of textual authority. These criteria concern the historical con-
frontation between the canonical texts that figured as points of reference, and
their modes of transmission.
In the colony, some of the new didactic practices of memorization and reci-
tation had their basis in the Christian pedagogy promulgated by the Church, as
happened in other parts of Latin America, but others derived from the regional
practices of local specialists and textual communities. Similarly, while the eccle-
siastical models of knowledge were based in reproducing canonical texts (e.g.,
Holy Scripture) through oral expression, these ecclesiastical practices of memo-
rization and recitation were also adapted by Andean communities according to
their own norms of discourse and textual authority, founded on other canonical
texts. Regional mnemonic practices equally demanded the incorporation of the
text, in a religious sense, and recognized that the text, once incorporated, had
to be memorized completely, and all of this was in the hands of men. Here, the
bodily (and not just mental) concept of memory as the formative basis of An-
dean teaching determines the effective oral and verbal transmission of knowl-
edge between speakers.
A fuller examination of these concepts of memory, in the context of orality,
the corporeal seats of remembering, and regional texts themselves, must first ap-
prehend the status of speech in Andean cultures as against written practices. For
example, older people often contend that verbal communication has more status
and truth than other modes of communication, such as writing. Don Domingo Ji-
ménez clarified how interpersonal verbal communication (aruskipasiña) achieves
an “eternal” status by being repeated verbally (and so “textually”) throughout the
generations. Added to which, the genealogy of this interpersonal verbal commu-
nication derives its authority “from the time of the Inkas, when the moon was in
the place of the sun.” Don Domingo’s faith in the eternity of speech and its sta-
tus in the world, mediated through the Inka’s authority, reveals the textual norms
upon which respect for verbal communication over writing was predicated, when
speaking was the basis of interpersonal communication. “The letters were com-
municated and the transmission was forever.” Another clue to the nature of the
power attributed to oral transmission was his assertion that the power of words
is comparable to “irrigating the fields.” Just as water irrigates sown seeds to make
them sprout, so verbal communication has the power to make ideas sprout
through the generations.
We saw in chapter 5 how alphabetic reading and writing is interpreted by the
164 A N D E A N T E X T S A N D T H E I R I N T E R P R E TAT I O N

comunarios of Livichuco in terms of the letters/seeds/pastures that the pupils are

charged with “making sprout anew” during the school year, as part of the lo-
cal farming and herding cycles of production and reproduction. The ideas about
the power and status of orality that underlie local interpretations of reading and
writing share an etiology in the production of the lands and the perpetual reno-
vation of its resources.
In practice, communal attitudes toward written texts, as the legal safeguard
of rights in perpetuity to local lands and their resources, avail themselves of the
regional norms of orality in such a way that the intellectual genealogies of tex-
tual transmission also derive from regional norms of recitation and memoriza-
tion. So a great deal of importance is given to personal ties of instruction, the
authenticity of recited transmission, and local values in general. As a corollary,
comunarios are often opposed to the idea that local knowledge “should go far,” in-
terpreting this as the “robbing of culture.”
Since many aspects of local knowledge are organized along gender lines, gen-
der relations also come to bear upon local practices of textual transmission. Don
Donato Inka of Livichuco commented once that meetings about boundary dis-
putes (like assemblies in general) are “a matter for men,” who tend to “guard
information” within the confines of the ayllu. If women were to attend such
meetings there would be no security, as they “let words loose into the world”—
another indication that ayllu knowledge must be “well guarded” locally and not
“let loose” outside the ayllu bounds.

L ocal and R egional A ndean T e x tual H ierarchies

The transmission of knowledge, both verbal and written, is controlled within

a social world based on interpretative relations, founded in turn on a hierarchy of
power and authority with communal and state dimensions. According to the lo-
cal criteria of historical precedence, molded by social memory, the original text
constituting the point of reference for predicating the truth of an original autho-
rized presence dates back to monumental time and the archetypal figure most
pertinent to questions of land rights: the Inka himself. The different levels of this
textual hierarchy are articulated through the metalanguage of the thakhi, in the
multiple sense of “pathway,” “custom,” and “law.”
In this context, instead of appealing to a body of written law, the authori-
ties of each minor ayllu in Qaqachaka interpret their consuetudinary laws orally,
founding their interpretations in turn on other hierarchies. A great part of in-
terpretative practice depends on the stewards (mayordomos) charged with church
activities, above all the care of the place saints. In this instance, a key genealogi-

cal tie between past and present relates the adoration of the local saints to the
prior adoration of the ancestral mummy bundles.3 From the community’s point
of view, the present system of ecclesiastical authority has continuities with Inka
dominion in the region and with the figure of the Inka as the supreme mediator:
the Inka king (Inka-Riy) or Inka-Christ.4 Hence, in their interpretation and execu-
tion of ayllu law, the mayordomos resort to the authority of the ancestors of the
place, and ultimately, to the religious power of the Inka.
In the case of a major offense in the ayllu, from adultery to murder, it was
the mayordomos (sometimes under the direction of the corregidor) who examined
under oath the testimony of the accused and then made a judgment. In the local
court held at the foot of one of the church towers, or in the church itself, they
arrived at their sentence and its execution. The accused was stripped completely
and made to pass over a cross, laid over the ara or altar stone under their jurisdic-
tion. (In ecclesiastical history, ara was the name for a consecrated stone on which
the priest laid out the Eucharist for the mass). After declaring verbally his or her
testimony, the mayordomos arrived at their sentence, appealing at times for the
help and supervision of the “Justo Juez Cristo el Señor.” According to local his-
tory, if the accused was guilty of a very serious offense, then the mayordomos car-
ried out the sentence with an axe from the time of the chullpas, which they used
to guard in the church itself.
Such cases inspire Platt (1992) to postulate ties between many textual aspects
of contemporary sacred language and the juridical discourse of past centuries.5
But it would be a serious mistake to perceive in these parallelisms more than an-
other play of textual contact and transferences. As Platt affirms, the principal me-
diators in this juridical language are the wise ones (yatiri), who simply borrow
certain elements from the colonial language of power according to their own
interpretation of local and regional hierarchies. In spite of superficial parallel-
isms with the colonial language, the language of wise ones, with its constant re-
quests for license and permission from superior powers (especially in their own
ceremony of investiture), has other foundations. And whereas the pinnacle of
the colonial juridical hierarchy was personified in the King of Spain (or his re-
gional homology, Inkarriy), that of the Andean hierarchy came under the power
of lightning, the sun and other stars, the wak'a and the Inka, and was mediated
by vocal power.
This is the case with oath making (juramento), which not only acknowledges
the sacred in the name of the Christian God and his written word in the Bible,
but also the power of lightning. The authority of lightning, its power to go to the
truth directly, and its inspirational force in the male domain of warfare, all im-
pinge on the gendered aspects of textual authority and its genealogy. As we have
166 A N D E A N T E X T S A N D T H E I R I N T E R P R E TAT I O N

seen, the interpretation of historical documents pertaining to land is generally

in the hands of the title bearers or apoderados, the majority of whom are men.
In this masculine domain, access to land titles has to do with the transmission of
masculine ancestral substance (muju, semen) from one generation to another. As
a consequence, it is common for the lineage of title bearers to pass in the male
line from father to son. It is also common for the title bearers to be yatiris, men
who have been “struck by lightning,” whose genealogy also tends to pass in the
male line, although this is not always the case. At the very least, the title bearer
must have a “good head.”
Only in exceptional cases, for example when there is no son in the family, does
the knowledge and handling of documents pass from the title bearer to his son-
in-law. In the cases of women who “know” the contents of ancient documents,
these are usually family relations of the great title bearers, their wives, daughters,
or granddaughters. This is because the genealogical and textual transmission of
title bearers are traced from the pinnacle of Andean authority, incarnate in the
powers of lightning that reach directly to the truth, and in the religious power of
the Inka to impose his rule vocally.
This is evident in the characteristic recitative style of oral transmission of the
families of title bearers, for example the Inka Maraza of Livichuco, when these
families dwell on historical or genealogical precedence. Their dissemination of
knowledge also has certain similarities with the hermeneutics of scholarly writ-
ing, in which the oracular quality of the voice lends itself to the Inka’s authority.
In our interviews with Don Donato Inka Maraza and his son Santiago (from
Livichuco), Don Donato always opened his discourse about the title bearers he
knew personally by naming all the links in a masculine chain of transmission
from the well-known title bearer, Feliciano Inka Maraza, in the early twentieth
century, up to himself. Although he did not remember his grandfather well (“as
he was very old”), Don Donato recorded that he was a “person of a high stature”
and that “he spoke very well.”
Qaqachaka oral history tells how Don Feliciano was a descendant of the
grandfather Inka, Mateo Inka Maraza, who hid himself from the Spaniards, liv-
ing from the time of the Inkas on top of the ayllu grandfather mountain, Lord
Mount Turu. Don Donato’s narrative of this event dwelt on the vital nexus be-
tween the powers of speech of this grandfather Inka and the flow of water. The
grandfather Inka made the water arrive there; it came out of the top of the moun-
tain “as if from a tap,” and then provided for the wet pastures at the foot of the
mountain. Finally, the people of the place felt for his solitude, made him come

down mounted on a mule, and sent him to be baptized (in the place called Misk'i,
in the Cochabamba valley).
As a descendant of Mateo Inka, Don Feliciano was considered the “last Inka”
of the place, the only ancient person descended from those people of long ago
(“the most respected”) who had come down from Lord Mount Turu after the
Flood, to found the settlement of Livichuco. This privilege compelled the then
ayllu lord mayor (alcalde mayor) to elect Don Feliciano as great mayor (jach'a al-
calde) of the place, charged with the defense of its interests in the world outside
its limits.
With his good voice and gift of persuasion, Don Feliciano struggled against
the big landowners “in the time of colonization” to defend ayllu lands against the
imposition of the haciendas. This entailed his traveling throughout Bolivia, carry-
ing his land titles and “walking with the laws,” where his struggle “against Spanish
law and in favor of Andean law” brought on the hatred of the Spaniards, leading
to his imprisonment for a year in Copacabana, where “he suffered a great deal.”
In his heated defense of the comunarios of the place from serving as “can-
non fodder” in the Chaco War, Don Feliciano presented his case first before the
lesser justice (justicia menor) and then the greater justice (justicia mayor). As a re-
sult, he achieved the freedom not only of the ayllu members from conscription
for that war, but also of the ayllu as a whole against the landowners. Qaqachaka
as a free community has never had landlords (hacendados), and always had pro-
ductive lands. He died in Sucre, in the midst of his struggles, with all his papers,
but when the other ayllu members found his body, there was not a single paper
on him, a fact that has had severe repercussions throughout Qaqachaka’s long
struggles for lands until now.
In their accounts of this former struggle, both Don Donato and Don San-
tiago were careful to distinguish the oral transmission of the history of the place
from that contained in the ancient documents accessible only by reading. They
also pointed out how the special nature of reading the ancient writing “written
with hens’ quills” meant that it had to be learned outside schooling “because it
was prohibited.”
The same masculine ties of transmission are found in the traditional tech-
niques of guarding the ayllu’s written documents. Communal commentaries
about these techniques reveal other corporeal aspects of masculine genealogy
that figure in regional notions of memory, the power of the ancestors, and their
revivification through reading, and how these ideas still influence school reading
168 A N D E A N T E X T S A N D T H E I R I N T E R P R E TAT I O N

For example, in spite of centuries of colonial and republican laws, it is still

considered important to guard the venerable ayllu documents (the “grandfather
and grandmother papers,” awil achach papila) in the family house of a line of ti-
tle bearers, although in theory the papers are the possession of the whole ayllu.
Traditionally such papers were guarded in a leather envelope, or a chest of hide
or wood (the so-called petaca), and special rites had to be performed before they
were consulted.
If these ritual obligations were not fulfilled beforehand there was the danger
of “losing your memory” (chuym chhaqtayasiña) and being left totally “peeled of
ideas” (q'ararpacha). The same could happen if you saw a bare corpse. Don Juan
Maraza warns: “You shouldn’t see a naked corpse; if you do, it will leave you
without anything.”
These communal fears about the danger of losing one’s memory on seeing a
naked corpse, or approaching ancient parchment documents improperly, imply
two constellations of ideas. In one, memory and clothing are somehow related,
as if the ideas held in the mind (or heart) of the living provide a temporary woven
covering. In the other, ancient documents are imbued with some ancestral and
fearsome force, at once dead and skinned, like the hides from which they have
been made. Even so, faced with these dangers, the comunarios still feel obliged
to take the inanimate parchments out from the chests in order to “read them”
(orally). It is as if the reader’s voice made them relive by “breathing life anew”
into the empty skins. If this is so, then a regional theory of textuality should ac-
knowledge the potential energy of regeneration in this act of revivifying some
dead element through vocal power.
Another central idea is that the efficacy of transmitting orally the contents of
ancient documents depends on the fertilizing ties of consanguine kinship along
a chain of human transmitters, who enact this transmission in an intermediary
vocal domain between the living and the dead. At the same time, the documents
(like captured beings) have to be well guarded “at home” and not “released into
the world.” Regional textual practices, then, attempt to overcome the alienation
of writing and the rupture it brings on in the oral transmission of truth, by incor-
porating written documents into preestablished ayllu reading practices. In addi-
tion, the original writer of the document appears to be revived by the oral activ-
ity of dictating (“reading”) the document’s content in verbal testimony.
The contrary case of the loss of vital ayllu documents (and the subsequent
rupture of oral-written transmission) always occurred when the title bearers
wandered about, carrying their charge to faraway cities. Similarly, when the cul-
ture of the place is spoken about and “loosed into the world,” it too is severely
endangered. In this sense, the textual attitudes of the title bearers and comuna-

rios of Qaqachaka, faced with Hispanic writing practices, are again like those of
Socrates faced by Plato in the Phaedrus, when Socrates too defended oral trans-
mission faced with writing “loose in the world.”

D o n F e l i c i a n o I n k a M a r a z a ’ s C h e s t , t h e C o mm u n a l C h e s t s ,
a n d t h e Fa m i l i e s o f D o c u m e n t s K e p t T h e r e

The ways of guarding documents practiced by the last generations of title

bearers, according to past precedent, were structured in the main by the ongoing
agrarian and herding contexts of the productive system the documents sought to
defend. However, the ramifications of the documents’ contents were perceived
differently by the state and the community.
Communal descriptions of title-holding practice compare the piles of docu-
ments with food produce from the ayllu lands. For example, Don Donato told
how his grandfather, the title bearer and great mayor Feliciano Inka Maraza,
guarded the communal documents at home in a kind of “chest,” where they
were “well ordered,” one behind the other, and “piled up” (suk'antata) in heaps
of paper. The verb he uses here is equally applied to food sacks full of seeds.
However, Don Feliciano’s methods for guarding the documents at home also had
to fulfill the conditions stipulated much further afield than the immediate ayllu
boundaries, in state policies concerning the methods of keeping documents that
interested them, centered on the so-called “communal chests” of the colonial ad-
In Las comunidades indígenas y la economía colonial peruana (1997), Escobedo
Mansilla traces the colonial history of these communal chests. In 1556, under
the viceroy marqués de Cañete, a new institution of the Hispanic state was born
when the viceroy decreed “that they were ordered to collect together in a house
all they had to give in tribute, and whatever was left over should be held in a chest
to be called community property” (1997, 111).6 From this date onward, the chests
“were already designed theoretically in their general outline,” although there was
little advance in their implementation in indigenous communities because of the
many difficulties of applying the new decree. In 1565, under the monarch’s rec-
ommendations, the demand was repeated that tribute should be “put in a chest
that the community should have, of which there were to be three keys, one held
by he who teaches doctrine, another by the cacique and the other by one of the
Indian mayors (alcaldes).”7
It was still difficult to fulfill this demand “while they did not first make con-
centrations of the indigenous population.” But within a decade, with the Toledan
reforms, the viceroy himself attempted to specify the details of the communal
170 A N D E A N T E X T S A N D T H E I R I N T E R P R E TAT I O N

chests that now received the name of “deposit box” (caja de depósitos). With To-
ledo, the system of community chests were disseminated throughout the vice-
royalty as a “neurological center of tributary administration” for taxation, and
local extensions of the royal coffer (ibid., 113, 136).
To this end, Viceroy Toledo ordered that the head of each moiety was to
have, at the side of the cacique’s house, a hut (bohío) with its respective chest,
where the tributary revenue would be entered, and that the said chest “must have
three keys” (ibid., 112). The principal state interest in the chests was as a means
of receiving tributary revenue, to compel the Indians to pay their tribute and de-
faulters to leave securities that would be kept in the chest while they completed
the month’s time limit (ibid., 113). Another stipulation required that the commu-
nal chest “would only be opened to make the entries and pay the respective credi-
tors.” Once the messenger (encomendero) and priest were paid, the rest remained
“in deposit” (whence it got its initial name from the viceroy); it couldn’t be spent
without “leave of the justice” (licencia de la justicia) (ibid., 114). Besides, “it was
ordered that a strict account be made of absences in the chest, in a book” that
“must be guarded in the chest of the town council (cabildo)” (ibid., 127).
Even so, there were the different ways of interpreting the communal chests.
For the state, the function of the chest was for tributary revenue, but for the
community, it proceeded from communal property (livestock and lands, tribu-
tary land, the possession of other collective landed property, set of working tools,
and so on). As Escobedo Mansilla points out, the institution of the communal
chest “has its bases in the indigenous community organization itself ” in a “pre-
existing communal reality” (1997, 122). Another purpose of the communal chest,
apart from being a means for tributary administration, was that of keeping cus-
tody of communal capital “to assist the collective needs of the village or assist the
underprivileged.” Its funds could meet the costs of public works in the commu-
nity, for example, for irrigation, repairing the church fabric, or more importantly
“the upkeep of schools and the sustenance of the teacher” (ibid., 128–33).
Regarding this last point, a key part of the reformulation centered in the com-
munal chest was Toledo’s stipulation that in each Indian village there must be a
so that the boys, especially the sons of the caciques, principals and other rich Indians, are
taught to read and write and speak the Spanish language as His Majesty commands......
and this [teacher] the priest will name, as he who seems to him to be the most able and
sufficient, to which he will give a salary in each one year of two outfits of rough home-
spun and six fanegas [approximately 24 bushels or 192 gallons] of corn or chuño, that which
most easily can be given to him, twelve sheep of Castile in each one year, that will be
bought for him from the community’s property.8

These stipulations are fulfilled by the comunarios to the letter, even today, espe-
cially the communal maintenance of the teacher in food and sheep.
Escobedo Mansilla does not tell us much about the practical management of
the chests, apart from mentioning among its minor costs the purchase of three
keys or desk materials “for the scribe” (1997, 134). In fact, those who were to act
as the chest’s turnkeys were a constant problem for the colonial state administra-
tion, who feared the “embezzlement of funds” on the part of the corregidores or
town mayors (alcaldes). But present-day commentaries in Livichuco concerning
the chest of Don Feliciano Inka and its role in Andean textual practices seem to
relate back to the fact that in the 1575 administration of tax tribute in Aullagas
and Uroquillas (not far from Livichuco), instead of the alcalde the “turnkey” was
the kipukamayuq, secretary or notary of the town council. Concerning this point,
Escobedo Mansilla emphasizes “in the first place, that the quipucamayo whether
he had under his charge one of the keys or not, really fulfilled an important func-
tion in the administration of the chest, since he was in charge (as we have already
seen in the chests or stores of communal property) of registering in his quipus, or
better still in writing, the control of operations in the chest” (1997, 141). Hence
in 1575, well into the colonial period, the principal person charged with the chest
was still the kipu reader, suggesting that the immediate historic precursors of the
communal chest, at least from the community’s point of view, was the Andean
tradition of registering accounts on a kipu.
Citing Guaman Poma, Escobedo Mansilla also shows that the charge of the
chest, at an even higher level, remained in the hands of the historical figure of the
alcalde mayor (formerly the Inka Totrico, which was the communal title granted
to Don Feliciano Inka) (1997, 142–43). If this is so, then the local management of
the chest of piled documents in the house of Don Feliciano Inka and his son can
be situated in an Andean textual institution of much greater authority. Title bear-
ers like Don Feliciano, with their other titles of mallku and alcalde mayor, would
come to occupy their hereditary role in Andean legitimate textual transmission,
and likewise, the contents of his chest of papers would be based on the content
formerly stored on kipus.
This historical perspective on the ways that Andean and Hispanic systems of
state administration were articulated, even in their hybrid forms, sheds light on
present-day comunario attitudes toward the new school libraries, especially the
“book corners” of the educational reform. Former state demands on the local
administration of communal production (through entries in kipus or communal
chests) even now structure how the new textbooks (and other materials) of the
educational reform are deposited in similar chests, under lock and key, to which
only the school director has the privilege of access. In extreme cases, even permis-
172 A N D E A N T E X T S A N D T H E I R I N T E R P R E TAT I O N

sion from the director’s office was not sufficient authority to open the chest, be-
cause the director himself feared that any loss of books would affect the “whole
system of inventories” outside his immediate charge.
Similarly, the idea (in Livichuco) that any malignant place in the school com-
pound must be “closed” by a wise one with three keys (this time in the form of
medicinal plants) must also draw on this historical precedence.9 It is as if written
documents still embody some alien danger that cannot be allowed to be “loose
in the world.”

Andean Oral and Writing Practices

In practice, the predominantly masculine practices of controlling the circula-

tion of communal documents often appeal to the genealogy of the document
and its reproductive character. The roots of many comunario fears centered in
matters of life, death, and alien presence are given expression here in a regional
theory of artifacts concerned with the reproducibility of originals.
The very manner of classifying documents sometimes refers to the genera-
tion of the documents themselves, in the sense of their revivification and prolifer-
ation as offspring of a more ancestral document. These cases refer to a feminine
corporeality, expressed in a specialized Aymara vocabulary for this textual trans-
mission from one written document to another. This act of “copying” (waraqaña)
an “original document” (tayka papila, “mother paper”) to a “transferred copy”
(waraqa), takes as its point of reference a reproductive model in which a “mother”
document (tayka) gives birth to a “baby” (wawa). Don Juan Maraza explained
how: “That mother paper must not be lost, no one must handle it. People are
made to understand from this document [but] they speak apart from the docu-
ment itself........The papers of before, the mother papers, were made with hens’
quills, even the letters were made in another way; there is no way of reading
them easily.”
These same terms replicate the terminology of Andean textiles in which the
“mother” of the pampa background gives birth to her “offspring/designs.”10 But
apart from this additional nexus between regional textile and writing practices,
the commentaries on these documents situate their venerable authority in the
antiquity (even eternity) of the “mother documents”; hence the prohibition on
handling them and the difficulties of understanding them. Ancient documents,
like weavings, are considered to be living beings with their own life trajectories.
Like an older person who enters the system of duties called cargos, a new doc-
ument is considered at the “beginning of its pathway” (qalltañ thakhi), whereas

one that is “complete” (and has reproduced itself ) has its “feet” and “head” (kayu-
chaña, p'iqiñchaña).
In other cases, the genealogical aspect refers to the creation of new charac-
ters through the discursive practices that accompanied the written preparation of
colonial documents, especially the periodic stocktaking of land called composición
de tierras. In practice, many people participated in the preparation of such docu-
ments in the dialogues and other interventions in their execution, for example be-
tween the local authorities (caciques, the boundary markers called lindero, or the
mallku of an ayllu or federation), and the state functionaries: scribes (escribano),
reporters (relatador), or magistrate (corregidor).11 The relative authority of such in-
terventions was expressed in the language of discursive authority, whereby some
interventions (most probably on the part of the local boundary markers) were in-
corporated as more mythical insertions, intercalated with the state formulae con-
cerning the King of Spain and the exercise of his dominions.
In the case of Qaqachaka, there emerged in this manner the hybrid personage
of Taqi Mallk(u) Astete, really a combination of the local cacique, Mallku Fer-
nando Taqimallku, and the commission judge (juez factor) of Potosí, Bartolomé
Astete. This local culture hero figures throughout the pages of colonial docu-
ments and in ongoing references to them in the recitational readings in matters
concerning ayllu lands by the title bearers of the place.12 His hybrid birth comes
from welding the two factions (cacique-mallku and state corregidor) present in the
original elaboration of the 1646 regional stocktaking of lands (composición de tie-
rras) by José de la Vega Alvarado, in a context where the state power at play was
reappropriated according to communal values and their own interpretations of
writing. In textual terms, the discursive practices in a document’s elaboration,
with its numerous points of reference (perhaps from other previous documents),
gives birth to new beings, characters that are revived on the dead paper, as were
the personages of novels in the following centuries.
Other daily practices based in these Andean theories of orality and writing
derive from the same faith in genealogical ties grounded in the corporeality of
the social body. This is the case with relation to libation making and the dialogues
that accompany sacrifice.
Older community members in Livichuco often make comparisons between
the practices of writing on paper and those of making libations in the mind, as
if they were common elements of a regional art of memory (rather similar to
the way that Derrida regards such genealogical exercises in oral cultures as re-
sulting in a writing-like mental inscription). But without exception they agree in
174 A N D E A N T E X T S A N D T H E I R I N T E R P R E TAT I O N

the many advantages over writing of making libations in the mind. Don Santi-
ago’s view is typical: “The libations (ch'alla) are unfolded more easily, while the
very quantity of letters produces lots of confusion.” Libations, then, emerge as a
whole discourse from a condensed code, while written words seem busy with the
multitude of their constituent letter elements. The implications of this assertion
can be heard in the communal memories of libation making. Don Santiago tells
how he recorded through libations what is now recorded on paper: “In ancient
times, when there was not this thing about reading.......making libations was like
writing on paper. So, to record something, they say, they had to record it with al-
cohol and in libations, whilst in present times it is with paper.”
This act of recording through libations is glossed in Aymara by the verb
amuyusiña, “to realize, be aware of, not forget.” His father, Don Donato Inka,
used this same verb to refer to the practice of recording libations to the way sta-
tion (tambo) of Livichuco, where the classrooms were later built.13 When he held
the office of postilion, Don Donato named in toasts the sequence of tambos on
the route to Sucre (Livichuco, Janq'u Qhawa, Pukuwata, Macha, Qulqa Phujyu,
Rutiyu, and so on). Other toasts, to the “mit'a of Potosí” followed the same route,
with its different “resting places.” The postilions also toasted the different ritual
sites of the place, especially the stones where they make offerings, with their
drinking names: Wooden Stone, Little Saint Monica, Knotted Cross, and so on.
These were followed by toasts to the different foodstuffs that arrived at the tambo,
which recorded their place of origin and their route into the ayllu.
In each case, the series of libations unfolded from the simple and universal-
izing notion—tambos—and all the rest followed in a predetermined order; hence
the community proposition that it was much easier to remember with libations
than with writing on paper.
These two textual practices, libation making and writing, shared other fea-
tures. Just as the recitations of toasts demands a preliminary sacrifice (in the case
of libations to the animals, or to the school), so a prerequisite sacrifice is neces-
sary before releasing the words of written documents. Given the immense power
of ancient documents, one precaution before consulting the grandfather and
grandmother papers entails sacrificing a sheep (or llama) and sprinkling its blood
in the place of consultation, resulting in the blood stains that spatter the cover
and initial pages of such documents. If no animal was sacrificed and none of
the appropriate libations made, the consequences could be serious. Don Donato
had seen title bearers from many parts practice this custom “with much respect”
when they visited the house of his grandfather. Don Feliciano even made sacri-
fices when he went off to consult documents elsewhere:

He knew Santos Marka T'ula, “We used to walk with Feliciano Inka Maraza,” he said. He
had many friendships, I’m aware of that. They arrived at my house in Jalaqiri, the people
of Yamparaes, wearing their hide helmets. They’d make libations to the documents there.
I don’t know where they came from; I think they were elected from all over. They came
saying “Alcalde Mayor,” and then they’d make libations and a blood offering to those docu-
ments with a llama.

In the opinion of Don Donato, his uncle, Robustiano Inka Maraza (the son of
Don Feliciano), a modern son who had learned to read outside the ayllu, suffered
the punishment of the laws (castigo de leyes) precisely for having ignored these lo-
cal customs of reading:
They say it was “punishment by the laws.” They say that my grandfather Don Feliciano
used to make a blood offering before going anywhere, but he [Robustiano] must have for-
gotten to do it. Even my grandfather became poor. He no longer had his animals; they
were all finished. This is because he no longer remembered the mountains. He used to
record all of these, and Tiwanaku too........This punishment was directed at his son, that
Robustiano; he was like an idiot. We buried him when he died. He was like an idiot; he’d
even sold the papers.

An essential aspect of ancient law demanded a general respect toward the docu-
ments, and more specifically the carrying out of a blood offering before vocalizing
them, as if the very act of reciting had to do with releasing the flow of blood.
This is why comunarios comment among themselves about the lack of respect
of researchers who visit them to consult ancient documents without taking into
account these precautions. It also accounts for their reluctance to accomplish the
educational reform’s request to have a new “open” reading corner in the school.
The fact that the traditional didactic practices of schooling derive from these ties
between reading in a local way (releasing a flow of blood), other regional modes
of recitation and memorization, and the recognition of authority in specifically
regional powers, makes them particularly difficult to reform.
We saw how the sacrifices at the initiation of the school year form part of the
counterhegemonic communal practices of reading and writing. In everyday class-
room practice too, teachers depend on communal (and state) permission to dic-
tate a text to the pupils, then they must repeat the same text orally, from memory,
resorting to writing only secondarily, at an intermediary level. This technique of
“legitimate textual transmission” (to use a term from Messick 1993) is known to
the majority of rural teachers, familiar from childhood with its dialogic use among
the local textual community of title bearers and their secretaries (secretarios).
In the local history of these practices, the descendants of title bearers (such as
Don Donato) perceive in the act of dictating out loud a form of teaching rather
176 A N D E A N T E X T S A N D T H E I R I N T E R P R E TAT I O N

like literary composition, in which what counts more than the written word is
the genealogy of the text, whether this was the “word of someone” (khitis satäwi)
or “the hand of someone” (khitin amparapasa). Once again, the corporeal aspect
of knowledge is emphasized.
Even in the classroom, legitimate textual transmission depends on ties be-
tween the “original” text (to be memorized), the elucidating commentary (non-
memorized understanding) on the part of the teachers, and its memorization on
the part of the pupils. Less attention is given to the “dead” writing on paper and
more to its vocalization, memorization, and transmission. In essence, this process
of transmission implies a sacred revelation, what Derrida and Foucault intuited in
the written practices of the Old World.14
This takes us back to a key question: To what point were Andean textual prac-
tices founded on the voice, as Derrida proposes for Old World textual practices in
his theory about the conceptual priority of the “word” over “writing” in the Eu-
ropean written tradition?
As López G. (1998) has argued, while Andean legitimate textual transmission
is founded on the voice, in practice the participation of at least two voices is re-
quired for its execution. This form of transmission has practical reasons, that dia-
logue as a more social practice has advantages over monologue as a mnemonic
device, and conceptual ones, that knowledge forms part of a woven field, in
which a “plied thread” formed by braiding two voices is considered stronger than
a “simple thread,” a widespread idea in the marriage advice giving that we exam-
ined in chapter 5.
In this case, the dialogical form of traditional classroom practice is simply a
variant on traditional techniques of teaching in which one person acts as “guide”
(ira) and the other as “follower” (arka), as in weaving practice, music making,
and so on. In the didactic process of transmitting the original written text via
the voice, the work of two men (title bearer and secretary) is interrelated: “They
make the paper between the two of them.” “He who recites out loud” (irpiri) be-
gins the transmission and “he who listens” (or “imitates”) receives it (arkiri). Don
Donato used this terminology when he spoke of the practices of his grandfather,
Don Feliciano Inka: “He dictated and his secretary received the transmission.”
These examples reveal the strong sense of corporeality in dialogue; it is not
simply a verbal communication of ideas in the air. In the intergenerational trans-
mission between himself and his grandfather, Don Donato Inka remembers viv-
idly the words Don Feliciano transmitted at a verbal level, although he hardly
recalled his grandfather’s face, “only as if it were in a dream.” He also records
with utter clarity how his grandfather used to speak “just by thinking,” and then

Don Donato received these ideas in his own memory. Given these characteristics
in common, Don Donato agreed that the techniques of teaching and learning at
school followed the same communal pathway of teaching and learning and writ-
ing norms.
In the context of schooling, this didactic technique was reinforced by the so-
cial conditions of the pupils, who had no textbooks, either at home or at school,
so they had to borrow them to copy the appropriate passages and then recite
them from memory. With this pretext, more attention is given to the opening and
closing moments of the didactic cycle: the recitation of the teacher and the aural
reception of the pupils.15 Following the same logic, less attention is given to the
teaching of writing techniques (how to form letters, connect them into words,
and connect words into sentences and paragraphs). As is well known, these skills
are hardly developed in the Bolivian educational system.
Parents judge their children’s progress by the speed with which they finish
each notebook, and not by the quality of its contents. And they assess their chil-
dren’s progress not by their ability in reading and writing, but by their form of
speaking. As Don Domingo observed, “First, you don’t notice any change, they
are the same. But when they’ve been for a few weeks, then you note a change
in their way of speaking.” More informal instances of the same traditional sys-
tem of teaching and learning take place between schoolmates in the breaks be-
tween classes, when pupils who have already learned some subject by memory
teach the smaller ones. In all these cases, the dominant theory of transmission on
which the legitimacy of knowledge is based depends on an intellectual tradition
in which writing is conceptually subordinated.
These dialogues in the classroom derive their power, in turn, from other more
powerful principles of legitimate textual transmission. Certain aspects of this po-
tency emerge in ritual contexts, when the voice is interwoven in dialogue with
the sacred powers of a place. It is possible to identify in these instances a meta-
language concerning vocal power and its expression in a sacred context. At times,
this metalanguage alludes to an “original dialogue,” probably that between local
ritual specialists and the gods of the place (the wak'a or sayxata as they are called
in Livichuco), upon which other textual practices have been elaborated, whether
in ritual or the classroom.
Certain principles of this metalanguage underlie the whole system of mascu-
line authority in the Andes in which the power of the voice plays such an impor-
tant function.16 The fact that this dual division of ritual labor in dialogue with the
sacred powers of a place emerges in the context of sacrifice, suggests that this,
with the attendant spilling of blood, is what moves participants to adopt certain
stylistic forms of speech (and song) with their marked parallelism and dualisms.
178 A N D E A N T E X T S A N D T H E I R I N T E R P R E TAT I O N

In this sense, the dialogues enacted during sacrifice provide an authoritative chain
of intertextual links between the community members with their gods and an-
cestors.17 At the same time, such dialogues serve as a communal and intercom-
munal (intercultural) source of power for expressing the authority of the com-
munity’s voice in its educational messages.18
Could it be that such dialogues (between ritual specialists and sacred powers)
serve as the model for all the rest?

T e x t u a l C o n t i n u i t y : C o r p o r e a l i t y, V o i c e , a n d M e m o r y

That Andean textual practices are founded on the voice challenges Mignolo’s
recent critique of this proposition in favor of Derrida’s original emphasis in Of
Grammatology on the voice as the basis of European writing practices.19 In the An-
des, however, sacrifice seems to be an even more fundamental element than the
voice as the basis of textual practices. This fundamental importance of sacrifice is
what both Derrida and Mignolo overlook.
Regional ideas concerning the physiology of the voice within the body con-
firm this critical nexus between the ritual context of sacrifice and its revivification
through the voice, by alluding to the historical fact of the Inka’s death and his re-
vivification as the authority of precedence.
From the commentaries we heard about the title bearers and their practices,
the importance of vocal power resides both in the genealogy of oral transmis-
sion, and also in a metaphysics in which birth through the voice brings to light
new beings (babies and the hybrid personages born through processes of docu-
mentation). As a descendant of his grandfather’s voice, Don Donato affirms: “As
his grandson, I don’t forget. I tell my sons, ‘This was so, and that was so,’ in those
times, they say. I heard it and I too am speaking the voice of my grandfather, up
till now.” Don Donato was equally specific about the physiology of this process
of transmission: the knowledge received in the memory of his grandfather “went
up to the head and then began to go out by the mouth, through the voice.”
Descriptions of the physiology of the voice and its source of power delineate
a sequence in which the voice originates in the nourishing of the body through
the blood system. They say that well-nourished blood goes to the heart, where
“it goes on boiling” (t'imphtasisan). From there, it rises via the “large vein” (jach'a
wina, meaning the windpipe/aorta complex) to the head, and there the strength
in the blood passes to the mouth and hence to speech through the voice. Elvira
Espejo of Qaqachaka explained the process: “Once the blood is boiling (wilpa
t'impki), it comes out of the throat and you talk well, they say that you talk well.
If their blood is not boiling well, it doesn’t come out, the voice can’t externalize

itself, they don’t speak well. They say those are the ones who stammer.” Elvira
suggests that speaking well depends on the temperature of the blood and the
force of its circulation. If the blood is boiling well, then “the words circulate,”
but if the blood is sluggish in the heart and lungs, then words do not circulate
well. This happens sometimes in the case of babies at the moment of birth, and
as a consequence “the baby cries and cries and never learns how to speak well,” it
only “begins to stutter.”
The heart, where the blood is boiling inside, dispatches air (samsuña) and this
in turn externalizes the words of “ideas” (arsuña). As Don Domingo says, “That’s
how the words (or ideas) are dispatched from the heart.” We saw how homolo-
gous physiological ideas operate in the body of woven cloth, when outside forces
enter like food into the textile “mouth” (laka) and so into its “body” (kurpu). In
this case, the food enters the intestines through the to-and-fro movement of the
weft, compared to digestion in the belly. In this way, outside energy enters into
the textile equivalent of the blood (the threads), and hence to the textile “heart”
(chuyma), its repository of memory and consciousness.
The eternal voice of the Inka, as the principium of speaking, functions ac-
cording to similar physiological notions. A conversation some five years ago con-
cerning an Inka burial near the main school of Qaqachaka gave us a pretext for
conversing with Elvira Espejo about the Inka’s voice. The burial had been dis-
covered when her father, with two other men, was excavating the site of a new
school building. A large rectangular and pitted stone, “just like a coffin” had been
laid over a corpse:
With its arms crossed, the neck cut, and between the teeth of its open mouth, they say
there was something like a metal wire, as if it were saying “Jaaa!” On its crossed knees was
another flat stone and on top of this, held in its hands, was an Inka drinking cup in the one
and a plate in the other. The whole configuration frightened people so much that they
only took out the drinking cup and the plate, and then they buried the rest as it was, with
the huge stone put on top again. “You shouldn’t make fun of it,” they said.

Elvira compared the burial, surrounded by worked standing stones, with co-
lonial drawings of the beheaded Inka. For her, the two drinking vessels were bur-
ied with the Inka “in order to forget.” From his head, cut off and placed to one
side, there seemed to be blood sprinkling into empty space, “as if he were speak-
ing in this way.”
Her explanation of this extraordinary burial was that “the Inkas were always
buried with a knife that pierced the hole in the crown of the head” (phuju, the
fontanel). “The blood used to come out in this way, and that blood irrigated [the
180 A N D E A N T E X T S A N D T H E I R I N T E R P R E TAT I O N

land].” Then they covered them over with a flat stone on top. (In this case, the
Inka’s blood is irrigating the earth). She related this custom of burying the Inkas
with the present-day practice of tying the neck of a corpse with a white thread.
“When we tie it so, this is instead of the Inka........It’s as if we were tying it with
his hand, and as if we were seeing his body, so they say.” For her, tying it in this
way gives courage, and “nothing will frighten us easily.”
For Elvira, the voice of the Inka is eternal and gives courage to the members
of his line. She says that his voice can still be heard at the dark of the moon and
the full moon, when the Inka speaks like the wind in those places where one of
his lineage is buried. “He is saying ‘Waaaa!’ or else we hear it together with the
wind, saying ‘Qhaaax’ and then it passes.” The same sound can be heard in the
month of August, when the voice comes out of the earth with the fire that is
burning there within. The Inka says Waaaax! as if he were shouting, and usually
frightens people. He speaks as he would have shouted out when he died, with
a great deal of energy, courage, and without fear. Besides, the Inka would have
spoken in Quechua and not in Aymara.
These ideas about the Inka’s voice form an essential part of a regional textual
theory that focuses on textual creation through the processes of revivification of
a captured or dead being. The very continuity of the Inka’s voice forms one of
the primordial texts of origin, a moment pregnant with possibilities on which
other practices of corporeal and spiritual continuity are founded. This is why, in
an Inka burial, the crown of the head is always pierced with a knife, and a stone
is always placed on top, “so that he goes on speaking.” As Elvira affirms, “My
grandmother said, if he died and we buried him just like that, he wouldn’t speak,
they say. We couldn’t hear anything.”
The relationship between piercing the crown with a knife, the blood spurting
out, and the Inka’s voice is not clear. But it warrants reflecting on how Andean
people would have interpreted the death of the Inka during the colonial period,
in the light of comparisons with the Christ of the new religion, by incorporating
elements from their own cycle of death and resurrection into the new Christian
These own elements are sometimes introduced through wordplay. Elvira and
her grandparents, when they describe how the blood spurting from the crown
transformed the Inka into a cockerel with a splendid crest (kiristu or k'ari k'ari),
deliberately make the wordplay kirista-kiristu (crest-Christ) to imply a likeness be-
tween the Inka and Christ:
Because, at the end, they say that the Inka used to sing just like a rooster: “Ququruqu!,” he
said. That’s why they punctured a hole there, they say, when he died. On dying, they say
that he sang like a cockerel for three nights. And then the Inka died, they say. That’s why

they say he’d died. We thought that perhaps they’d made a hole there so that he had his
crest. He died and then he was converted into a cockerel, we say.

When he was pierced in that way, they say, “it’s so that he sang, so that he’d say
‘Cock-a-doodle-doo!’ (ququruqu in Quechua). If there was no hole, there’d be si-
Her narrative linking the blood that spurted from the Inka’s crown with the
“crest” of a rooster that has power to sing “Cock-a-doodle-doo!” each dawn, not
only plays on the name of the last ruling Inka, Atawallpa (wallpa is generic for
rooster or hen in modern Aymara and Quechua), but on echoes between this
name and the powerful Inka origin myth founded on the birth of the Sun.20 As in
the West (according to Foucault’s arguments), Andean foundation texts concern
the origins of the world and powers that hark back to this liminal and revivifying
Communal memories of Inka writing form another key textual support
with which other regional matters of reading and writing engage. Conquest is
bypassed here, and colonization is presented in terms of regional intercultural
struggles. For example, innumerable narratives in Qaqachaka tell the textual his-
tory of the region, beginning with the first textual struggle between the chullpas
of the place (the ancient occupants of the burial ruins) and the Inkas from Cusco,
who came with their own writing and writing practices.
Evidence for this struggle is sometimes found in the buried remains of knot-
ted threads, “similar to an Inka kipu,” in the lower parts of the ayllu. Pieces of an-
cient clothing are also found in the soil or washed up, usually nothing more than
other threads sprinkled with nitrate powder. Other “Inka” textual elements found
in the remains of the place are deerskins (taruk lip'ichi), which the Inkas suppos-
edly used “to send messages to their family members.” According to Elvira, “they
used to pit the hide with sharp stones to incise it and give it form.”
Her grandparents told her how they used to send messages in plant roots
“tied together hard, however you could.” The Inka knew how to write messages
on these roots, and also how to read them, so as to “speak the same content later.”
These facilitated communication between the invading Inka and the local people
of the place. According to Elvira there were also special messengers before (like
chaski), who entered for terms of a year, taking messages.
She interprets as letters the Inka writing in roots, on stones, hides, and threads:
“The letters appeared in this way, they say.” Interestingly she applies to the notion
of Inka “letters” the Aymara verb ch'imthaptayaña, which indicates a surface pit-
ted with different colors, something germinal and fertilizing. The writing with
threads was in different colors, some with pieces of hide or hair added to clarify
the message, rather like the threads that the older people of the place still use.
182 A N D E A N T E X T S A N D T H E I R I N T E R P R E TAT I O N

Oral history also tells how these knotted threads were the primary means of
communication by which the Inkas, with their developed reading abilities, de-
feated the rest, including the local chullpa people. The Inkas “read everything,
they say” on these knotted threads, but the chullpas “didn’t know how to read
them, and so the Inkas had to teach them.” In these tales, the Inkas thought of
the chullpas as idiots who had to learn everything: how to weave, cultivate, and
construct their houses. They were particularly stupid because their heads were
shaped differently; they did not have Inka-style heads. To remedy this situation,
they had to learn from the Inkas “how to wrap the head to make it grow,” and so
achieve the necessary intelligence to read the kipus. Elvira relates this Inka cus-
tom with modern-day ayllu head attire. Girls still wear a wide white-colored band
on their knitted caps, and its tip hangs down “to the backside,” to stretch the
form of the head. Similarly, boys’ caps are knitted with a pointed tip:
The Inkas taught us so, they say, to make the head grow, tying it with cord (like a damned
one). They say that the chullpas made their heads grow only as an experiment........The
chullpas have their heads oval here, don’t they?
As the chullpas didn’t know how to read the kipus, they did this, they say, they did it
like this. “It’s to have a good head,” they said, “this person knows how to make it grow
here,” they said. That’s why they have a large band in their caps. It’s embroidered and usu-
ally hangs down to the backside. It’s to make the head grow.

The longue durée of this regional textual history implies that European writ-
ing was received in the Andes as an intermediary form to facilitate oral repeti-
tion, and not as an end in itself. We might face the difference that Derrida draws
between concrete writing (more mundane and human) and metaphoric writing
(more tied to the spirit and the voice). In the Andes, the recited word was privi-
leged over the written word for being closer to the sources of power and author-
ity of legitimate textual transmission, in the social memory of the figure of the
Inka. But even when the written word was at issue, then this too drew on the
same source of power and authority, this time in Inka writing.
It is improbable that there would have been Andean practices of “reading” like
European silent reading for comprehension. If such practices existed in the Andes,
they happened outside the instructive milieu in subjects, literary genres, and activi-
ties considered more mundane in comparison. At an everyday level, vocalization,
recitation, and memorization were privileged as modes of teaching, face-to-face
encounters, and oral and aural techniques, based on other original texts.

C yc l e s o f S ound
Prayers and the “Rain of Letters ”

They say there’s a skull in the school, but I don’t know much about it.

Elvira Espejo Ayka

Now everything has been changed, now there’s another way to teach reading and writing.

Domingo Jiménez

While many traditional oral and written schoolroom practices in the Andes derive
from a regional interpretation of European reading and writing, they derive most
specifically from the reception by Andean populations of colonial ways of teach-
ing (and learning) Christian doctrine. At stake was a struggle over doctrinal cor-
pus. Each party in the struggle (Andean populations and Christian colonizers) had
distinct sources, rationales, and methods of reasoning that they drew on.
To demonstrate that this is so, we first locate the evangelization program of
the first decades of the colony, especially Christian instruction carried out in na-
tive languages, in the atmosphere that rural children would have experienced in
that period. Then we consider how the daily activities of recitation and memo-
rization in the traditional classroom replicate these colonial forms of indoctrina-
tion, above all instruction in prayer. Finally we suggest how Andean populations
might have assimilated these enforced practices by appropriating them into their
own religious practices.

184 A N D E A N T E X T S A N D T H E I R I N T E R P R E TAT I O N

A n I n t e r c u lt u r a l E n c o u n t e r o f
R e l i g i o u s I n t e r p r e tat i o n s

We describe elsewhere how, in the struggle over doctrinal corpus and the
constitution of doctrinal authority, Andean populations reformulated Christian
doctrine according to their own textual theories.1 In this reformulation, Christian
ideas about the death and resurrection of Jesus Christ were incorporated into
Andean ideas about reviving the body of the Other. In this instance, the transub-
stantiation of the body of God was understood as the ingesting of the “letter” in-
carnate of his Word, according to an Andean interpretation of the Eucharist and
salvation. In the working out of this process, especially in the rites around Easter,
the presence of the ayllu children was vital for their capacity to pray for rains and
in their function as the little lambs of God.
In their encounter with ecclesiastical teaching, Andean populations seemed
to opt for internalizing certain aspects of Christian doctrine taught to them in the
colonial period, above all that concerning Logos (Verbum in Latin): the idea that
the second person of the Holy Trinity was made incarnate in Jesus Christ. Histor-
ical echoes of this contact between religious practices have been influential up to
the present, so that regional reformulations of Logos still influence community
ways of thinking about formal state education and its instructional practices.
Although the Church promulgated this idea initially, an early secularization
of these ideas, taken up by the state, can be encountered in the linguistic interpre-
tation of “verb” (verbo) in the first Spanish grammar, Antonio Nebrija’s Gramática
de la lengua castellana, published in 1492, the same year as the Conquest of the
Americas. Both ecclesiastical and secular currents coincide in present-day debates
in the educational reform.
The ecclesiastical history of the region around Qaqachaka (the Cordillera de
los Frailes) shows the immediate influence of Augustinian evangelization during
the colonial period, although neighboring Macha, with which Qaqachaka had ec-
clesiastical ties, came under Franciscan influence. Both schools of religious ideas
were to influence many aspects of the regional teaching practices of Christian
doctrine, whether in ecclesiastical or more secular contexts.2
Particularly striking are the similarities between Augustinian religious ideas
and regional ideas concerning “seminal thought.” As we saw in chapter 4, semi-
nal thought in the Andes concerns growth generated from a germinal base, ex-
pressed through the language of sprouting and generating offspring (wawa) in
different life forms (animal, human, vegetable) from an original seed. In its refor-
mulation, the attraction of seminal thought for Andean populations was proba-

bly its association with the impending processes of production and reproduction,
in particular the capacity of the land to produce in and of itself. Kusch develops
this idea in El pensamiento indígena and many scholars of the Andes have identified
this same constellation of ideas (although they do not name it as such, nor have
they examined its origins).3
Beyond the simply productive, the Augustinian interpretation of Logos trans-
forms the seminal thinking of those times to a metaphysical level. Focusing on
the second person of the Trinity and the idea that God is truth, Augustinian semi-
nal thought sustains that the nature of God, in his function as Logos, is incarnate
in Christ the Son, while God the Father is being, and the Holy Spirit is love. For
St. Augustine, God, as the foundation of all there is, created the world through
his Word (Logos), incarnate in his Son. This meant that the Son, as the Word,
contained in himself the ideas and immutable reasons of things. St. Augustine
identified these ideas or immutable reasons with “seminal reasons” (of which the
Stoics spoke), a way of reasoning that explained why the creation, even though it
was a unique and indivisible act, produced its effects successively over time.4
Ecclesiastical influences in colonial pedagogy were to incorporate certain as-
pects of Augustinian seminal thought, for example the art of germinating the
potential “seeds” inside you, when they are stimulated with opportune experi-
ences.5 These Augustinian philosophical ideas would have passed into the class-
room through the filter of Andean seminal thought.
The regional reformulation of these ideas, absorbed through the complex
of ontological depredation, has even more striking parallels with Augustinian
thought. We find in regional thought the religious and metaphysical basis of
the cycle of metamorphosis of a baby/wawa/Son, first into a captured and mal-
treated being, then into a bloodstained head, and ultimately its resurrection as a
new harvest of offspring (wawa). In the final transmutation of the trophy head
into a new generation of babies, it is the energy of the Holy Spirit (ispiritu) that
generates new life through the impetus of breath. Hence, this new life is gener-
ated through the power of the voice and the Word.
Basing their reasoning, historically, on this constellation of ideas, the resi-
dents of Qaqachaka seem to have internalized the Augustinian notion of Logos
in the sense of the Word of God as a creative force and reinterpreted it accord-
ing to their own criteria. Evidence also suggests that Andean populations (such as
that of Qaqachaka) came to reinterpret these Christian ideas at an everyday level
through religious instruction in prayer and song, in the very practices used by the
Church for evangelization.
In 1615, referring to New Spain (Mexico), Fray Juan de Torquemada (a Fran-
ciscan) in Monarquia indiana (chapter 18, vol. 3), considered the particular im-
186 A N D E A N T E X T S A N D T H E I R I N T E R P R E TAT I O N

portance of evangelizing in native languages. It is worth scrutinizing his com-

mentaries, as the same ecclesiastical teaching techniques used in Mexico for
indoctrination purposes were later applied in the Andes. His advice “so that a
heart may be illuminated and put on the pathway of perfect clarity” was that
populations “were granted the language” for doing so, by learning to “bleat” like
the sheep of the shepherd Jesus Christ (1615, 43).
This Biblical language about the bleating of Christ’s flock would have ap-
pealed to Andean populations interested in the successful management of herd-
ing, above all of the flocks of the Inka. The Franciscan method of teaching the
faith was apparently successful, given that “a half year after their arrival, they
could not only understand those who talked to them, but they could answer
them back sufficiently in the Language, and in this way they communicated one
with the other” (ibid.).
In translating “The Principal Aspects of Christian Doctrine” into indigenous
languages “they put it into a very gracious plainsong, so that those who heard it
would take it better to memory........Thus was this mode of Doctrine, to stamp it
into the hearts of these Indians: because with the suavity of the song, they took
more pleasure, and with the sweetness of the words, they grew fond of it” (ibid.).
In practice, this technique was applied first with the schoolchildren of the period,
and only then in the church: “This medium that these apostolic men laid out to
smoothen this Doctrine, they first exercised in the schoolchildren for some days, up
till the point that they knew it by memory, and then they taught it to the others,
to which so many people attended that it was good to see them, as that which
moved them was God himself, who desired them and who wanted to count them
already as his own, like sheep, unloved until those times” (ibid., our emphasis).
We should not ignore the crucial point that it was the school and not the church
that served as the primary institution for indoctrination.
Torquemada gives a vivid impression of the quantity of voices praying and
raising “the voice, and sound of the Word of God” everywhere (ibid.). Once the
prayers had been learned, the “common people” from the villages used to teach
the other children: “even in the villages where a Minister of Doctrine does not
usually attend, every morning they gathered in the churchyards, the common
people of the village, who sang the Doctrine, teaching it to the children” (ibid.).
Such was the fervor of the period that boys and young men, without being
compelled to do so, gathered in the churchyard “at the agreed hour” to pray from
evening until almost midnight (ibid.). The prayers of the children were particu-
larly valued, for their innocence and their power of uplifting the praise of God
against the “Infidels or enemies of God” (the same idea we encounter today in

their ritual ability to pray for rains), especially the children of the local interpret-
ers of the faith, the doctrineros (1615, 46).
The friars were equally impressed by the power of memory in the Indian vil-
lages. Village residents heard a sermon or a history of a saint once or twice and
remembered it (as in the present-day schooling practices in Andean villages), and
“afterwards they said it with much grace, daring and efficacy” (1615, 44). Torque-
mada observes how, in the beginning, the “common people” of the Indian villages
did acts of devotion that not even the Spaniards did, for example going down on
their knees to recite the Lord’s Prayer or Ave María (ibid., 4–44). The paradox is
that the friars interpreted all these activities as evidence that “barbarous peoples”
were learning the Faith. For them, this change in sound and voice provided evi-
dence of the letting go of their previous “great errors and blindnesses” of having
“Idolatrous Rites” and of having “instead of the true Gods, false and lying De-
mons, Enemies of the Human Lineage” (ibid., 44).
The new Christian practices gradually replaced former religious practices
(just as Biblical practices had their roots in a much more pagan world). Torque-
mada refers to these previous indigenous practices (“the accustomed sacrifices
in which they used to kill men, in secret, on the mountains and in out of the
way and desolate places”) and how “they did not give up making sacrifices”; in-
stead “the Diabolic Temples were served and guarded with their ancient ceremo-
nies” (1615, 46). He also worries that “the priests themselves, sometimes heard at
night the uproar of the dances, songs and drinking bouts in which they engaged”
(ibid.), and which the words of God should have been replacing.
So in spite of the confidence of ecclesiastics such as Torquemada in the pro-
cesses of evangelization, their interpretation of what was happening in the pueb-
los de indios and what was really happening were worlds apart (just as at present
in the schools of the Bolivian highlands).

A More Secular Encounter bet ween Andean

and European Theories of Language

A more secular attempt to dominate the voice of other peoples, in order to

gain control over their territories, was conceived by Antonio de Nebrija on the
basis of a European theory of reading and writing.6 His Gramática de la lengua cas-
tellana (1492) served as the basis of linguistic policies implemented by the Spanish
Crown in its expansion in the New World, where it was the reference text for the
application of language standardization programs in Castilian Spanish. His Latin
grammar served as a model for the written grammars of Amerindian languages,
188 A N D E A N T E X T S A N D T H E I R I N T E R P R E TAT I O N

which gradually replaced Amerindian texts with European alphabetic writing.7

Nebrija presented his philosophy of writing for the first time in the prologue
to his Gramática, dedicated to Princess Isabela of Castile. There, linguistic ideas
of the times with respect to theories of the development of languages, phonet-
ics, and phonology are mingled with contemporary European ideas about gene-
alogy and corporeality. In the way of similitudes, Nebrija appreciated in each lan-
guage its “infancy, plenitude or flowering” (in a vegetative sense), when it showed
its “strength,” and then their decline or lapse (1492, 5–6). In a more openly colo-
nizing manner, he applies the organic metaphor of uniting the different anatomi-
cal parts spread all over Spain “into one single body,” to the unification of the lan-
guage. Uniting the two ideas, he saw in the same process, taken further afield than
the frontiers of Spain, a flowering “of the arts of peace” and through the union of
language with writing, the immortality of Spain’s “achievements” (ibid., 6–7).
His scheme initiated a new kind of colonization in which the body of Christ
is stretched over the body of empire. To Nebrija’s way of thinking, the immer-
sion of barbarous peoples into learning the colonial political and religious lan-
guage would incorporate them into the new divine body. Thus, Nebrija saw in
the standardizing of a common language (Spanish) a vital colonizing practice (a
point underlined by Illich in his essay of 1981) since “many barbarous peoples and
nations of alien languages, with defeat, had the need of receiving the laws that
the victor imposes on the vanquished and together with them our language......
not only the enemies of our faith, who have the need of knowing the Castilian
language, but.......all the others who have some dealings and conversation with
Spain and need our language” (ibid., 8). To this end, he recommended teaching a
standardized language to the vanquished starting from childhood: “if it does not
come in childhood, that of learning through use, the sooner they may be able to
understand through this work of mine” (ibid., 8).
From this linguistic foundation, Nebrija proceeded in the second chapter
of Grámatica to announce the more religious goals of his work, declaring the
power of letters to “reveal the divine” (ibid., 14). He sought to understand the
origin of this power through the linguistic and ecclesiastical history of the pe-
riod (asking from whom and whence did they bring letters originally to Spain).
In the same quest, he appealed to a theory of the period concerning the physi-
ology of the voice, illustrated in his description of the phonological apparatus.
His ideas there about the impulse for the voice from the lungs, via the windpipe
(gargavero, or the windpipe/aorta complex) and toward the mouth, were trans-
ferred to Andean thought concerning the physiology of the voice and its power
of putting forth words, or ideas, by externalizing the knowledge incarnate in the
carnal and spiritual wrappings of the body: “So it is none other than the figure

for which the voice is represented; nor is the voice aught else than the air that we
breathe, thickened in the lungs, and then moved in the rough artery that we call
“windpipe” (gargavero), and thence onward successively through what is called
the uvula, tongue, palate, teeth, and kissers or lips. Therefore the letters repre-
sent voices and the voices signify, as Aristotle says, the thoughts we have in the
soul” (ibid., 19).
Aristotle’s idea (borrowed by Nebrija) that “letters” represented the voices
and that the “voices” signified the thought we have in the soul, was repeated by
Saussure many centuries later. This same idea is at the basis of the “phonetic
writing” (where writing is presumed to represent the voice) that has become the
dominant school of thought in applied linguistics, implemented in educational
reforms throughout Latin America.8

The Struggle bet ween Different Textual

F o r m at i o n s : T h e E a s t e r P r ay e r s

In the colonial period, the ways of teaching children Christian doctrine con-
stituted the primary way of teaching literacy. Colonial books of catechization of-
ten incorporate a syllabary for teaching reading and writing, literacy being the
essential condition out for the transmission of foreign ideas.
The ensuing struggle between different textual practices and the relative
weight of the doctrinal authority behind them centered on the articulation of
distinct textual theories about “being” and “knowing”: those from Christian Eu-
rope (in their Augustinian interpretation) and those from the Andes. In the search
by Andean populations to reformulate their own ideas and practices in accor-
dance with the new norms, the Christian idea closest to the cycle of creation of
their own textual theory (of the transformation of the dead into a new harvest of
babies, or wawa) was nothing less than the death and resurrection of Christ.
Paradoxically, the most persuasive evidence about an Andean interpretation
of the death and resurrection of Christ is found at the heart of the ecclesiastical
rites of Easter, which take place in the main church of Qaqachaka. In the essay
“La trama revitalizante” (Arnold and Yapita 1999a), we describe how the girls and
young women of the ayllu had to pray in the church with the objective of making
the bloodstained head of Christ “relive with their voices,” and how they did this
with the “letters” learned from Christian doctrine. By performing these rituals,
each young woman hopes to “achieve salvation,” although this signified prepar-
ing herself for war, in which the braiding of protective prayers acted as a means
of defense for her and her future husband against the “enemies of the world.”
The performance of the Easter prayers took place first within the church and
190 A N D E A N T E X T S A N D T H E I R I N T E R P R E TAT I O N

then in the procession of the stations of the cross (istasyuna), the local name for
piles of stones in the church precinct. Older people call these “stations” taqawa, the
name of the rostrum where ancestral remains, and the trophy heads of enemies
captured in interethnic wars, were formerly displayed. So the Easter prayers derive
from a common semantic domain as those of the school precincts, whose rites
also focus on a place called taqawa (the school rostrum or tribune built with an-
cestral bones) where schoolchildren pray each year to make the ayllu lands green
These common meanings revolve around the instructional practices of prayer,
in particular the vocal skill exercised by children in prayer that enables them to re-
quest the rains of immortality (the ritual ability in their charge). We examine first
the ecclesiastical rituals, then the communal prayers and finally the school prayers,
focusing in each case on their ties with regional textual theories and practices.

T he E cclesiastical P rayers

The primary instructional methods of prayer in Qaqachaka derive from the

ecclesiastic practice of “novena” (borrowed into Aymara as nuwina): prayers and
acts of devotion that last nine days in their execution by way of suffrage and of-
ferings for the deceased. The community identifies different classes of novena,
each of which forms part of a wider ritual cycle with its respective territorial ju-
The ecclesiastical novenas practiced in the church and its precincts during
Easter are under the charge of the young people of the ayllu, ideally those who
are still innocent, in the sense that they have not yet had sex (the same reason we
encountered in chapter 4 for young children’s ability to pray for rains). These go
accompanied by another group of smaller children who “must cry” as they don’t
yet know how to pray.
At a communal level, some of the “small novenas” (jisk'a nuwina) are per-
formed in the immediate family compound with other community members
(men, women, and children), while others are enacted further away, when some
children accompanied by older people go to the nearest hills, and the women and
other children cry close to the fields. The reason given for the participation of the
whole community was that “we usually eat the same and we usually drink the
same,” which clearly relates the prayers to production from the land. The com-
munal “great novena” (jach'a nuwina), on the other hand, are performed on the
high mountains far away from the community by the older men, ritual specialists
and selected ayllu authorities (the ayllu leaders or jilanqus, the mayordomos, and
feast sponsors), if it does not rain in December and January.

Finally, in the novena practiced in the school, the practitioners are the school-
children under the instruction of the teachers and some older comunarios. These
take the form of the interminable reciting and memorizing of the contents of
daily classes, as if they were reciting for rain.
These different kinds of novena (family, ecclesiastical, communal, and school)
form another hierarchical ritual system that binds the immediate locality into the
higher levels of a remembered Andean state. We shall describe each one in turn.

L earning the P rayers

A decade ago, the Easter prayers were learned through didactic techniques like
those of the colonial period (or those still practiced in the classroom). These tech-
niques were instrumental in the articulation of two distinctly textualized worlds,
generating transferences of interpretation and hybrid modes of significance.
The Easter novena are recited during the Sundays of Lent and Good Friday.
Church performances during this period mark the culmination for preadolescent
young people of a period of oral learning of Christian doctrine (lutrina) in their
respective hamlets, under the tutorship of the local doctrinero (lutriniru), mainly
older women.
The doctrinal prayers are learned by memory through a didactic technique in
which the doctrineros formed mnemonic aids out of clay and placed these figures
on dishes. Ideally, the twelve figures on each of twelve dishes represented a key
idea of the prayer “like the beads of a rosary.” In the case of the Lord’s Prayer,
the Aymara word kamisa (“whatever” in the sense of “whatever is thy will”) in
the translation of the verse “Thy will be done on Earth as it is in Heaven,” is illus-
trated with a hand lifted toward God, and awkisaraki (“Our Father”) in the same
prayer, with the figure of an old man.
Resorting to this didactic technique often results in the double meanings that
characterize so-called Andean syncretism. For example, the Ave María (Jamp'atqha
Mama) was taught with two clay figures: a toad (jamp'atu) and a woman that repre-
sented Mary. In this case, the technique appeals to the aural play between the Ay-
mara word for toad, jamp'atu, and that for “saluting” the Virgin Mary, jamp'atqha.
As a result, in Andean Christianity, the Virgin Mary, made manifest as the Virgin
Earth, comes to be associated with the toad, also called “Virgin” (wirjina).
In Qaqachaka, the prayers learned and then sung in Aymara through these di-
dactic techniques are taken directly from examples in the Doctrina cristiana of the
Third Lima Council of 1583, or from a contemporary Manual de la doctrina cris-
tiana (like that of 1577 prepared by priests from the Company of Jesus), an indica-
tion of the power and longevity of the instructional influences of that period.9
192 A N D E A N T E X T S A N D T H E I R I N T E R P R E TAT I O N

Elvira Espejo remembered how she learned prayers from her paternal grand-
mother, Doña Domitila Quispe, who taught these methods as a doctrinera when
she was younger. In the center of each dish was a cross of corn that announced
“the completion of one cycle of prayers with an Amin Jisus.” In each dish, a com-
plete round of the prayers was done, and then the last prayer in this dish ceded to
its “offspring” prayer, for which you passed on to another dish to begin another
round, and so on, all in the form of a spiral (llawuntaña). So the conceptual orga-
nization of oral apprenticeship in prayer followed the same logic of progenitor
and offspring (wawa) found in the textual genealogy of both written documents
and textiles. In addition, the tracing out of the spiral shape began in the center
and then expanded outwards according to the round of prayers, first defining its
spatial limits in one dish before passing on to the next, in another example of
the ontology of regional patterns of constructing knowledge, this time through
sound. Reciting and memorizing the prayers was made easier by the exaggerated
rhythm of versification (shown with diacritics in the colonial texts) that made of
the prayers a murmuring almost without sense (rather like that heard in Bolivian
classrooms today).
The girls learned their repertory of prayers under the instruction of the doc-
trineros in the small chapels of each hamlet. Then, at the indicated time, the older
ones went down to the main church in the pueblo to fulfill their religious duties.
If they did not learn the prayers beforehand, they would arrive “mute” at the
church, and be reprimanded. The youngest ones attended classes but did not go
on the main church; they just prayed in the temple on the appropriate days. All
of them were taught in their hamlets in the later afternoons, from the moment it
got dark until bedtime.
In the past, there were more stringent demands on this form of praying. The
learning period was more extended: girls had to pray each Sunday after Carni-
val and then each day of Holy Week. And the gendered division of ritual labor
was more dialogical. According to Elvira, although praying is now a matter for
women, “long, long ago, both men and women had to know the prayers, and
there was competition between them”; “if a woman didn’t know, then the man
had to. But now the men don’t pray at all!”
As another requirement, both groups had to sense the inscription of Christian
doctrine and Christ’s suffering on their bodies. First, you were made to recite com-
pletely all the prayers for the end of Holy Week and if you didn’t complete all the
prayers in order, a menacing corporal punishment was that “mange would appear
on your face.” Then, for the final performance of the prayers, you had to make
rounds of the church kneeling on the hard ground, and the young men would of-
ten put there on purpose some sharp stones, until they made the girls bleed.

P rayer P erformance

Nowadays Qaqachaka girls only recite “pure prayers” each Sunday of Lent
and all night long on Good Friday inside the church, reaching an ideal of twelve
prayers. Elvira used to sing the Credo (Iyaw satpi), the Lord’s Prayer (Nanakan
Awkiña), the Ten Commandments (Tunkapän Palawra or Tunkapän Ixwa, which
seem closer to the replies of the catechism), St. Peter and St. Paul (San Pilaru, San
Palaru), the Act of Contrition (Qullana Jisus Kiristu Tatay), and “another that is
pure counting” (wakhunti), evidently the Seven Sacraments.
The timing of these prayers had astronomical indicators, another sign that
they alluded to other symbols of authority. The authorities in charge (jilanqu and
mayordomo) “observed the position of the Goat kids (Pleiades),” which informed
them how much time was left to end the prayers, and if necessary they would
instruct the girls to hurry. Ideally “on ending the twelve prayers, the Goat kids
would disappear on the horizon.” The imminent rise of Father Sun was another
incentive to hurry, and the girls “used to pray faster and faster” accentuating the
mixed rhythm of the versification (0 10 , 0 10 , 10 0 , 10 0 ), as “whoever
prayed out loud was really praised and desired as a wife.”
The novena of the stations of the cross were equally concerned with salva-
tion. But while the prayers sung inside the church gave utmost priority to “re-
vivifying the bloodstained head of Christ” through the braiding of voices and the
letters of Christian doctrine, those sung at the stations of the cross were addition-
ally concerned with “generating” the seeds/letters at the different ecological lev-
els of ayllu production.
Those called “songs of the stations” (istasyun kirki) are sung after the young
women leave the church to do the round of the stations of the cross (istasyuna).
The principal axis of the nave served as the point of reference (taypi) for the pro-
cession, and each time you passed this point, the tone of voice had to change to
another novena. All participants went praying on their knees “to cleanse them-
selves of sin” (juch pampachañapataki).
In the past, according to Elvira Espejo, they went out of the church in small
groups belonging to the different hamlets, each of which made up a larger group
from the same ecological level of the ayllu, to direct their prayers at production
from the land. In the linguistic criteria expressing group identity, those from the
“upper levels” (patxarana) sang in Aymara, those “from the middle” (taypirana) in
a mixture of Aymara and Quechua, and those “from the lower levels” (manqha-
rana) in Quechua. Each of the three groups had to know two songs of the sta-
tions, reaching a total of six variants in all. The groups of young men went round
the church on their knees first, going to the right.
194 A N D E A N T E X T S A N D T H E I R I N T E R P R E TAT I O N

“The group that began was always from above” and they sang a “novena from
above,” followed by those from the middle who sang another novena, and finally
those from below with their own novena. When the first group arrived at the
axis of the nave, another group set off. Then, when the last of the three groups
of young men reached the axis, the first group of young women (from “above”)
set off, and so on. Each group, a little separated from the others, had to do three
rounds of the church, so the three groups reached a total of nine rounds of the
church, that is to say a novena. The nine novenas were sung in an elaborate braid-
ing of voices, when differences of gender, language, and the order of rounds fa-
cilitated the future production from the levels of the mountainous terrain.
The procession around the stations was called istasyun sarata. As to the mean-
ing of the processions, Elvira and the other young women gave more importance
to recognizing the U shape of the three “walls” of the ecclesiastical enclosure
as they did the rounds. Older people, though, related them with the ritual site
called taqawa, and in some communities they actually call the stations taqawa.10
Elvira from Qaqachaka pueblo calls the mounds of stones ataqawa (a play be-
tween taqawa in Aymara and atacar, “attack,” in Spanish), and she was aware that
“bones used to appear there” thinking of it as a former pantheon where the an-
cestral bones were buried. The combination of the nave’s straight axis and the
piles of stones related to the dead, was analogous to the thread of a rosary (or cir-
cular kipu) with its knots marking the resting places of contemplation.
Elvira also remembered that, while the girls prayed in church, one of the au-
thorities walked around holding a skull which “he used to butt against anyone
who was sleeping.” The similarity between this church butting with skulls, and
the schoolchildren butting with sheep during the initiation of classes to awaken
their strength against the enemy letters, has to do with overcoming the same fear
of death in order to reach salvation. Because of the dangers from the dead of the
place, the smallest ones were not sent to the church, “only when they knew the
prayers sufficiently” to defend themselves.
The verses sung in procession along the stations of the cross reiterate a pri-
mary concern with the overcoming of death through salvation, reinterpreted
from a regional perspective. For example, Elvira sang the songs of the stations
specifically “for Mama Tulurisa” (Our Lady of the Sorrows), “because Dolores
Mamala decided this about making novena” and “she orders us.” Elvira considers
her to be the wife of Animas Tatala or Tata Risuriksyuna (Lord of the Resurrec-
tion), “put with his bloodstained head onto a wooden cross.” In a gendered duty
based on this precedent, Elvira felt obliged to sing the Easter prayers in order to
“revive” the Lord of the Resurrection, “as the prayers function as a wife’s novena

for her dead husband.” Identifying with the Lady of the Sorrows, Elvira sings as
she ordained to revive her dead husband, although her immediate concern is to
protect her future husband from such dangers.
Elvira still remembers six of the novenas she sang as a girl in the pueblo of
Qaqachaka (at the middle level of the ayllu), although school and college tasks
over the last years have “dulled her memory.” She also complains that “now they
don’t pray well, mixing them with Quechua” and how the young men are badly
brought up and pester the girls, shining torches in their eyes when they come out
of the church. “They even take their mantles and from then on they are consid-
ered married,” so now, “there really is reason to be afraid!”
The first novena, “Aruray,” “Let us salute God,” lays out the instructions from
Our Lady of the Sorrows to perform the novena, both in the church and the sta-
tions of the cross. The refrain names her husband, the Lord of the Resurrection
(Tatitu), as a beautiful little flower (suma panqaray), in a reference to the babies
(wawa) of the coming harvest:
Aruray aruray, siluy aruray I salute you, I salute you, heaven I salute you
Tatitu sumay panqaray Little Father, pretty flower
Khistiraki kamachistuy, Tatituy And who directs us? Little Father
Kamsasaraki kamachistuy, Tatituy And what does she say as she directs us,
Little Father
Nuwinasim sasan kamachistuy, Tatituy Do the novena, she says as she directs us,
Little Father
Kamsaraki kamachistuy, Tatituy And what does she say as she directs us?
Little Father
Istasyuna saram sasan kamachistuy, Do the station, she says as she directs us,
Tatituy...... Little Father......

The second, “Yus ti salwi,” “May God save you,” associates salvation by God
with doing the round of the stations of the cross. The third, “Risyurist tala,” con-
firms that the order to do the rounds of the stations is also directed by the Lord
of the Resurrection, husband of Tuluris Mamala (Our Lady of the Sorrows). The
fourth novena, “Sinsalsismay,” is a wordplay on the Spanish “Sea santísimo,” “May
he be the holiest,” whereby the Aymara version transfers the original meaning to
the domain of mountain offerings without salt (sin sal), and blood sacrifice:
Wila purismay sinsalmay sinsalsismu Blood, you must come without salt, may He
be the Holiest
Sima purismay sinsalmay sinsalsismu You must reach the summit, may He be the
Ay Tatituy...... Oh, Little Father......
196 A N D E A N T E X T S A N D T H E I R I N T E R P R E TAT I O N

Wila purismay sinsalmay sinsalsismu Blood, you must come without salt, may He
be the Holiest
Sima purismay sinsalmay sinsalsismu You must reach the summit without salt, may
He be the Holiest
Ay Tatituy...... Oh Little Father......

The fifth, “Patxarana,” “Above,” names the flowers growing at Easter at the dif-
ferent altitudes of the ayllu that must be woven into the verses of the different
groups of singers, organized on ecological lines, as they do the round of the sta-
tions of the cross.
Patxarana tanitaniy, Ay Tatituy Above is yellow gentian, Oh Little Father
Taypirana chanachanay, Ay Tatituy In the middle is chanachana, Oh Little Father
Manqharana putuputu, Ay Tatituy...... Below are little blue bells, Oh Little Father......
Kunasaraki kamachistuy, Ay Tatituy And what does she say as she ordains us?
Oh Little Father
Istasyunasim sasan kamachistuy, Do the station she says as she ordains us,
Ay Tatituy...... Oh Little Father......

The sixth and final prayer, “Siryu siryu,” plays on the Spanish name (cirio) for the
large wax candles in a church, sometimes incised with pieces of incense in the
form of a cross, which are blessed on Easter Saturday. It also alludes to Our Lady
of the Sorrows as the arbiter of compassion (misericordia) and healing (médico), al-
though a more submerged meaning of misericordia might refer to a dagger used
for killing foes (as it does in a Salve María collected in the 1940s).11
Siryu siryu misirkurya Large candle, wax candle, misericordia
Miryu miryuy misirkurya...... Médico, médico, misericordia......

In her commentary on these prayers, Elvira emphasized how the quality of

the songs had to do with their orality. In her opinion, the girls’ prayers are much
better than those of the church sacristan, “as he doesn’t really pray” but only
“reads what is written down on paper.” Her claim for the superiority of Andean
orality as opposed to European alphabetic writing again drew on the precedent
of Inka practices of prayer. For Elvira, in Inka times they used to have “buildings
much bigger than churches” called “The tambo palace of the pampa” (tamp palas
pampa, perhaps a memory of the Inka palaces that measured the movement of
the sun). “It was huge, you even went down steps to go in.”
In her narrative, Inka prayers had to do with warfare, according to the period-
icity of the moon (that some chroniclers mention),12 as “they prayed and made
offerings at the dark of the moon, and fought at the full moon.” In Elvira’s view,

there inside their palace, the Inkas “prayed terribly when there were wars,” to ac-
company the movement of the sun and moon: “Inside there, the Inkas prayed as
soon as the moon rose, they say. They used to pray at night, they say. When there
were wars, they used to gather there inside, they say. They used to gather at the
full moon, they say........There inside, they used to play the cow horn, they say.
Even Viracocha and Mama Ocllo were there, they say.” “Lots of people used to
gather, they say; and they prayed hard, they say.”
Of course, in those times, instead of the Our Lady of Sorrows and Resurrec-
tion Father, there were other Inka gods, the principal gods of prayer being Mama
Ocllo and Viracocha. “Mama Ocllo ordained them, they say.” In the wars, “the
Inkas used to confront [the enemy] first with decorated slings, then with bows
and arrows and lances, and then with swords or knives. And they had shields to
defend themselves. They used to pray at night as soon as the moon rose.”
As she says, “it was the same as now”; the Inkas organized the prayers ac-
cording to the ecological levels of the ayllu. “Those who came from below en-
tered from below, and those from the upper levels entered by the upper door.
And those who came from the middle levels entered by the door in the middle.
They say the site was on top of a hill.”

T he P rayers as a R evivifying Weave

In this warring context, local doctrine has its own explanation of how Resur-
rection Father (the former Father Sun) was brought back to life through prayer,
under the orders of Our Lady of the Sorrows.
Elvira explained how, as they prayed in the past, the girls had to “braid” their
voices together (k'anaña), and do it “really tight” (suma q'ara, literally “really
bald”), and “fine or sheer” (llusk'a). Making it “loose” (llanka) like a sieve was no
good. Besides, they had to “pray really clearly,” like the sound of running water
and not “all mixed up.” She related this nexus between braiding the prayers really
tightly, and weaving in the same way, with warfare. For her, the weavings women
make for warfare should be really tightly and densely woven, so that they can lift
stones and help the men in arming their slings. As a measure of this quality, the
cloth, or any knots involved, “shouldn’t let water through.” Another character-
istic of war textiles is that they were “well beaten and tight,” like a dead enemy
who had been “well pulped and beaten black and blue, like papaliza” (a color-
ful Andean tuber). These definitions refer to the same semantic domain as Platt
mentions in his study “Entre ch'axwa y muxsa” (1987a), in which the brutal beat-
ings of the enemy in boundary disputes (ch'axwa) provided the necessary condi-
198 A N D E A N T E X T S A N D T H E I R I N T E R P R E TAT I O N

tions for an imminent stage of fertility (muxsa), in other words an anticipation of

the revivification after death.
In the Easter context, centuries of indoctrination have inculcated the idea
that the image of the dead enemy is centered in Christ, with his injured body and
head sore wounded. But regional reformulations of this same idea insist that the
cycle of re-creation that must have followed his death could only begin by releas-
ing the renovating rains. This is why the main goal of the Easter prayers, with
their weave tight and beaten (like the dead enemy), is to “release the rains” and so
renovate any remains from warfare. Specifically, the reciting of the Lord’s Prayer
(Nanakan Awki), the Credo (Iyaw satpï) and Ave María (Jamp'atqaman) are said to
have “helped the rain.”
Another function of the prayers in Qaqachaka relates the defeat of the en-
emy through the struggle of braided voices with “defending yourself ” against
counterattack, and at the same time “saving yourself.” From this perspective, the
soteriology of Andean populations is centered in “saving yourself ” from the en-
emies of the world (as they were taught in Christian doctrine), but transferred to
a context of warfare. So saving yourself from the enemy, in its Qaqachaka guise,
demands the recording of dead souls, whether of the ancestors or the defeated
enemies. As Elvira says “They used to pray for the dead souls.” You saved your-
self in this way, with the following words:
Yus Tata almanaka yanapt'ita...... God the Father, dead souls help me......
salwasikiristxa. How would you save me?

The regional notion of “defending yourself ” (tiphintasiñataki), as reinter-

preted Christian doctrine, is extended to suffering from physical toil. So the help
of the dead souls “is to save yourself from any kind of labor,” in the sense of
accomplishing it quickly and without suffering. “You ask God, and whatever is
asked for comes to mind rapidly.” Dead souls are thought to have more power
than God, so more help is requested of them.
The reason given for the greater power of the dead souls in your own defense
and salvation is that “they are dead, whereas God is still living.” Elvira clarified
this point of view: “My mother says that prayer is to ask of the dead souls. Even
my grandmother used to say the same. It shouldn’t be mixed up with God. Be-
cause God seems as if he’s dead, but actually he’s alive”: “‘It’s as if we were kill-
ing God, but how could God be dead?’ they say. ‘They’d say that God is living, but
actually he has to be resuscitated in this prayer. You shouldn’t forget that,’ they
said.” In this context, for Elvira, reciting the Ave María “is for the dead women,”
while that of the Credo “is for the dead men.” You must always ask of the God-
souls praying out loud, as “it’s something else to ask of God directly.” Most im-

portant of all is asking of the souls “remembered and not-remembered” (amtata

jan amtata), as this includes all the dead souls in the world.
In these reinterpretations concerning death and resurrection, another funda-
mental function of prayer for young people is that of “getting a spouse.” You can
even direct prayers to acquire the desired husband, and protect him at the same
time. In this search “you should pray well and weave well,” as then you are act-
ing as the warrior-women of Christ, and defending the desired husband from the
mortal dangers of war.13 The essential part for young women in making of this
constellation of ideas a reality is that, in the regeneration that follows warfare
and the defeat of the enemies of the world, they will be in charge of the harvest
of new babies (animal, human, and vegetable). With this end in mind, they must
learn to beseech rains, in order to make the babies develop and grow.
Elvira emphasized that “in the past, in order to marry, young men had to be
strong warriors” while “the girls had to be good weavers.” Unfortunately “now
the school has made us forget all this. Now there is no time to weave, as every-
thing is written down.”

T he G reat and S mall C ommunal N ovena to R e q uest R ains

Praying for rains is concentrated in the great and small novena in the com-
munity that take place in December and January, when the sown land is already
sprouting and there is the maximum danger of the plants drying out if the rains
do not come. Likewise the animals will suffer if there are no pastures. The very
uncertainty of the highland environment and its productive capabilities is prob-
ably why, in the past, Inka Cusco would have demanded state control over local
rainmaking rituals.14
In each hamlet, persons specialized in “making the land strong” at this criti-
cal time of year (the paqu yatiri) do so aided by the efforts of the children. In the
great novena, two kinds of rituals concern the cosmological circulation of wa-
ters, echoing the same semantic domain as the school rituals at the beginning
of August. In the pueblo of Qaqachaka, they practice the ritual called “changing
water” (um turkaña), in which portions of water collected from different sources
(springs, ponds, and so on) are gathered in a large pot with local medicinal herbs,
to “make them mix together.”15 In surrounding hamlets, such as Livichuco, they
specialize in another ritual, “raising the health of the water” (um salur aptañani),
but collecting in a round pot only the water nearest the place. The big pot (wirkhi)
used in the pueblo is considered masculine, while the round pot (p'uytulu) used in
the hamlets is feminine, an expression of the political organization of a ritual cen-
ter and periphery divided on gender lines.
200 A N D E A N T E X T S A N D T H E I R I N T E R P R E TAT I O N

Two aspects of this ritual interest us here. The first concerns the importance
of children as “flowers”; as the comunarios comment, “the babies are flowers” and
“always have to help.”16 The second concerns the relationship between the rains
and sound. When the comunarios go to the mountains to collect the different por-
tions of water in the “water changing” ceremony, they must pray and sing, and
on returning, dance and sing wayñu rainy season songs “with a good siren,” ac-
companied by pinkillu duct flutes and large guitars. This is because, as Don Juan
Maraza affirms, “the rousing up of song and dance also rouses the rain.”
In this respect, the long history of ties between local rainmaking rituals and
religious prayer, the powers of the voice, and the ritual participation of boys and
girls, all under state control, have simply been readapted to the new ecclesiastical
setting. A key aspect of the reformulation of these rites in Qaqachaka is “the call-
ing upon the Lord of the Waters” mentioned in chapter 4, a modern version of
the Inka God of the Waters, the great Viracocha, that Elvira remembered when
she spoke about the Inka prayers.
In this wider historical context, the relationship between modern rainmaking
rituals (and their counterparts in the school precincts), and the communal mem-
ory of past Inka rituals, hinges on the links between the highest mountains of
the locality, and the power of the Inka hierarchy in Cusco. In the region of Don
Domingo Jiménez, oral history narrates how, in the time of the Inkas, you had to
climb high mountains to request the rains, and so “you do the same today.” As a
result, the matter of praying is even entering the new school curriculum.
It is the same in Livichuco. You must go to the high peaks to request the
rains, as “you have to bring the rains with lots of energy.” It is no good for the
community to “bring a heavy rainfall or else hailstorm” which could happen oth-
erwise. Usually the indicated older people from some two families, accompanied
by various children, “go to the novena” (nuwin saraña) for rains in the hills where
their fields are, at the ritual site called jat'i, which means literally “to scratch with
the nails,” as “the earth there is scratched to put the offering inside.” The aim is to
make the smallest ones (of four or five years) cry, as “they don’t know how to talk
much, and only cry,” and their “crying” brings the rains. To this end, the adults
are insistent that they cry as they walk to the novena. According to Elvira, they
even strike them with their mantles, ponchos, belts, whatever there is, and make
them go down on their knees:
Karaju, q'asam, jall puriyañani, Dammit, cry, let’s make the rains come,
Apur pur, qunqur qunquri, Hurry, hurry, on your knees, on your knees,
qunqurt'am.......! go down on your knees.......!

As a result, the highest mountains (called achachila), the guardian mountains, and
the corner shrines “on seeing them crying pitifully” have compassion on them
and call the rain.
The most knowledgeable, ancient, and respected person among the group,
on reaching the mountaintop, begins to pray, going on his (or her) knees and ad-
dressing the high peaks. Don Juan Maraza recorded what they say:
Jilarat mallkunaka, Mallkus that stand out,
may aka awiyatururu, ask the Provider Mountain,
May aka Awar T'alla, Awar Mallku, Ask the Lord and Lady of the Waters
Ukax jawst'anikimay Please call them,
jay ukham, sasa. so saying.
Jan jawsanjanta ukatsti For if you don’t call them, then
kunantirak aka wawanaka mantinxasti what will maintain these children?
kuns maq'apxä sas...... what shall we eat, saying......

These oral verses remind us of the so-called Inka prayers collected from the
Cusco region in the early colonial period and transcribed onto paper by various
chroniclers, among them Guaman Poma and Molina.17 Although almost five
hundred years separate these recitations, their formulae are almost identical, in-
dicating that they drew on similar sources, whether these were Inkaic, ecclesiasti-
cal, or a reinterpretation of the latter in a more regional mold.
The lamentations that follow echo these historical resonances; older people
always begin and the children follow them, in the roles of “guide” (ira) and “fol-
lower” (arka). Don Juan Maraza narrates how they are charged to recite:
Kuns aka wawanakar liwäxa maq'a What food shall I serve to these children?
jan Awar Mayku, Lord of the Waters,
jani jallu milagrosinjanta ukatsti If you do no miracle for the rains, then
kuns aka wawanakñarux maq'a liwäxa what food shall I serve to these my children?
Kuns maq'apxä, sas jay ukhama...... What shall we eat? so saying......

Then the children answer:

Kunakraki nanakasti maq'apxasti, What things are we ourselves going to eat?
yanapt'anipxakitay Please help us,
lastapxay if not we shall suffer.
Akan anchhichhan Here, at this moment,
kunas janis puqhxiti sas...... nothing is being produced, saying......

If the rains still do not arrive, even the specialized older people cry (the yatiri
and amuyt'iri), “those who know how to manage these things, all of them used to
cry at this moment.” As a result, on many occasions, it starts to rain at that very
202 A N D E A N T E X T S A N D T H E I R I N T E R P R E TAT I O N

moment. Some people add to these cries the Lord’s Prayer, and libations to the
sacred sites, the Pachamama, the corner shrine (iskina) and guardian mountain
(uywiri), named as the local provider (awiyatura) of the great commune, beyond
the limits of the immediate locality.
Afterward, they sing twelve songs of different types, to salute (arunta), re-
quest (mayiña), and take their leave (tipirinataki). Don Juan Maraza narrates:
Yus Tatitu, sas Little Father God, saying,
Kunäms yanapxt'anipxakita, Please help us as you can,
Aka jallu akaru tispacht'anipxita Send us the rain here,
Kuna urkamas lastapxästi, jay ukham, Until when will we suffer, so then
Kunrak maq'apxästi what shall we eat?
uywanakñas jiwarji, Even our animals are dying,
Yapunakñas wañarji Even our fields are drying,
Yus Tatitu, sas ukhama Little Father God, saying so
Aka wawanakamar mä jalluk tispachanipxita Send us the rain to these your children
sas jay ukham so saying,
taqimanaw jaqi utjtana We have people with all kinds of ideas
janiw mä amtaniktanti...... We don’t have one single way of

The Livichuco schoolchildren, eleven-year-old Elías Chukicha Wallpa, ten-year-

old Eloy Wallpa Maraza, and eight-year-old Roberto Choque Wallpa, record with
pride their wailing efforts to beseech rains on 25 January when the older people
asked them: “We used to go off to request it, we’d pray wailing; we used to go
up the hill for this, to make it rain when the rains weren’t coming. Then we came
down wailing, and the following early morning the rain arrived really heavily.”
The following day a heavy rain fell “that almost made the houses collapse.” Their
prayer was a curious skewing of the Lord’s Prayer toward warfare and vengeance
on the enemy:
En el nombre del Padre, del Hijo, el In the name of the Father, the Son and
Espírit Santo, the Holy Spirit,
amén: amen:
Padre Nuestro que estás en cielo, Our Father which art in Heaven,
Santo ahogado, sea tu nombre, “Drowned saint,” be thy name,
Vengador en...... Avenger in......

Then they cried: “Boohoo!! Please let the rains arrive. Boohoo!!” In a division of
ritual labor by age, the older children prayed and the young ones just wailed. As
the smallest one said, “I don’t know how to pray. They pray........I know how to
wail for the rains.” And according to a ritual division by gender, “The girls would
pray and we’d wail.”

The little novenas (jisk'a nuwina) in the community are related to the great
novena in the way of fractals (and the interrelation between parts), with the same
object of “raising the health of the water” (uma salur aptaña). Carried out in the
patio of a house in the presence of the hamlet authorities, the participants must
go twelve times around the house corner shrine, on their knees, praying that it
will rain. Then a dirty-colored ewe is sacrificed, the color of rain clouds.

C hanging Waters

However, the most striking ecological basis to the school rites to the nation
that we witnessed in August 1998 can be found in the water-changing rituals
performed some years ago when the rains did not arrive in time in Qaqachaka
pueblo and Livichuco. Both rituals link locality to state and beyond, in a cosmo-
logical circulation of waters impelled by the vocal powers of children.
In the pueblo, first of all the ayllu leaders (jilanq tala, jilanq mala) gather the
people under their charge to pray; above all “they herd together” the children
with whips to take them to the church, saying “let’s change the water” (um
turkañani). Inside the church, the older people lowered from their niches the vir-
gin-mummy bundles (Mama Kantilayra and her two daughters, Mama Kapitana
and Mama Ch'uri) most associated with the rains, to beseech them on their knees.
While they prayed, the authorities whipped the children “so that they wailed and
prayed before the goddesses.”
These female gods had rainmaking attributes even in their dress. Mama
Kantilayra had her little “pitcher” (p'uñu, now rather cracked) that the comunarios
related to the rains dripping from heaven to earth. In their prayers, they beseeched
her to release more rain from it, addressing her as an Inka princess, “ñusta.” These
local instances of rites perhaps once carried out in Cusco have certain similarities
with the Inka prayer for rains collected by Father Blas Valera and translated by him
and Garcilaso el Inka, whom we mentioned in chapter 1. It is also possible that the
present-day prayers make the same allusions to sacrifice and revivification.
In a later part of the same rite, portions of water from different sources in the
ayllu were gathered into large pitchers, with the help of a wise one, in order to
regenerate the ample circulation of all the ayllu waters in a cosmological sense,
the waters of above and those of below, in a ritual preoccupation that has been
described by various authors.18
In a local variant of this ritual in Livichuco some years ago, Doña Rosa Maraza
told us how each hamlet had to participate, under the direction of the authori-
ties, in the collecting of water from local sources and “only then did the clouds
come.” Another important feature of the ritual was the sacrifice of a white ram
204 A N D E A N T E X T S A N D T H E I R I N T E R P R E TAT I O N

and the eating of its roast meat, accompanied by the thick and spicy broth called
At different moments in her account, Doña Rosa Maraza emphasized the im-
portance of the women’s singing “really hard” rainy season songs (wayñu) and
other Carnival songs, while the men accompanied them on pinkillu duct flutes.
This combination of song and libations “made the clouds appear.” At the same
time, all of them had to do rounds on their knees, murmuring twelve supplica-
tions to God in the places called heaven (silu).
Two women went to collect water from the pool at the foot of Lady Mount
Jujchu (jujchu t'alla, the ayllu grandmother), at some distance, while the men
went to a place called Shining Water (Qhusmi Uma) at the foot of Mount Phiri
Phiri. In the same locality, other older women went to bring water from Uma
Jalsu (“where the water comes out”) in Qallampata, the place where the school
parade began on the night of 5 August. Water from all these different sources was
collected in three big pots, bound around the neck with little belts in which white
flags are inserted to represent the clouds. Seeing this mixture of different waters
in the large pots compels the women to murmur among themselves “the twelve
seas” (tunkapän lamara), a reference not to the sea as such but to the Milky Way,
which they consider to be the source of all the waters in the world, and so of a
truly cosmological circulation of waters.

School Rites to Request Rains

In this light, the prayers in the school precincts are a more secular echo of
what happens in the church and the community as a whole. As Doña Antonia
commented, “they pray just the same in the school” (iskuylanx risapxarakiwa),
although she referred in this case to the pupils’ interminable daily recitation of
letters under the teacher’s direction. If in their daily and ceremonial activities of
memorization and recitation, the pupils are metonymically “praying for rains,”
then what more efficacious manner than releasing a whole “rain of letters”
(founded originally in the Word of God) with the object of making the blood-
stained seed/head of Jesus Christ sprout again, and so generate the crops of the
coming year.
Like the sheep sacrificed at the opening of the school year, the children as
little lambs of God are figuratively “eating letters like pasture.” These enter the
stomach during the rains, in order to sprout again the following year. As Elvira
clarified, “it is as if it were raining meanwhile.” And as Don Domingo states in
relation to this same complex of ideas, “it’s like requesting the rains for the year.”
In this sense, the children are chewing over the letters in their mouths to call the

rains, in time for the next harvest. “‘Let us chew them for the rains,’ they say.”19
Don Domingo went on to relate the teacher and his teaching with the mature
pasture that the children as little lambs must eat in the form of letters. Therefore
a vital part of children’s development, in their role as the little lambs of God, oc-
curs through their participation in this cycle of growth of the pasture/letters:
“the pastures grow out of the earth and are still soft, but gradually they harden
and mature, and in the same way the children keep on growing.”
So in the wider cycle of war and counterattack, the children are figuratively
eating the pasture/letters of the teacher’s cerebral knowledge (rather like a didac-
tic Eucharist—from death to resurrection), to green the ayllu lands.
Taken as a whole, the communal organization of these rainmaking rituals
casts doubt on the success of colonial attempts at indoctrination, while shedding
light on how the performance of present-day rituals by Andean populations and
their interpretations of them draws significance from social memory and the en-
during ties it evokes between the locality and the highest levels of the former
Inka state.
Whether at the level of family, pueblo, or ayllu, the participants (men,
women, and children) under the direction of the authorities and wise ones,
braided together a hierarchical series of ritual sites (mountain peaks, ecclesias-
tical and school sites) through the sounds of their prayers and wailings, songs
and libations, and rogations to God. All of this (including the obligation to make
sacrifices at these sites, nowadays of animals but formerly of children) consti-
tuted a state matter concerning Inka religious jurisdiction, in which the regional
mountains and plains (mallkus and t'allas) serve as the Inka’s ongoing stand-ins to
assure a maximum production from the land through adequate rainfall. In this
sense, the whole expanse of Inka territory was (and continues to be) articulated
through the vocalization of water flow.
Through the vocal medium of prayer, unleashed with a preliminary sacrifice,
a pan-Andean objective within the terms of a regional textual theory is that of re-
constituting the ancestral remains scattered in the different ritual sites (bones and
hard parts here, soft parts over there) and so refertilize the flocks and replenish
ayllu lands. In a moment of peace, cosmological flow concerns the circulation of
waters; in a more bellicose one, the circulation of blood. In both cases, the ends
are the same, that of achieving the revivification of the dead.
With respect to the sacrificial offerings to the mountains, there is a clear
nexus between the innocent pupils wrapped in white smocks and the little lambs
of God. In order to be good Christians, comunarios were instructed from the first
decades of the colonial period in their ritual obligations demanded by the organs
206 A N D E A N T E X T S A N D T H E I R I N T E R P R E TAT I O N

of the state (church and school), and with this objective, the children had to go
“as lambs” to the school or church “corral,” to pronounce in prayer the Word of
God, bleating under the care of their shepherd Jesus Christ. In this context, we
can appreciate the use of derivations of the Aymara verb anaña, “to herd,” both
for the communal sacrifice (in tribute) of sending their children to school, and
also for those other more demanding sacrifices of the past concerning children,
carried out to satiate the hunger of the hills.
In their efforts to articulate outside demands with their own communal ob-
jectives, Andean populations must have assimilated the death and resurrection
of Christ, focused on the rites of Easter (at Calvary, the “place of the skulls”), to
their own textual theory, where the ritual cycle of life and death was centered in
the place of ancestral bones called taqawa. In this knot of ideas, Christian notions
of death and resurrection were transferred equally to the sacrificial lamb, Christ,
the enemy, or the children as little lambs.
In the Inka past, there were evidently real sacrifices of children, but today,
children are “sacrificed” to the school, under the scrutiny of the officiating priest
in the person of the rural teacher, who supervises the accomplishment of com-
munal obligations to the state in order to perpetuate state rights to income from
rural lands. In fulfilling these state interests, the teachers of the military and edu-
cative hierarchy accomplish the center’s dominion over the periphery by convert-
ing the communal “babies” or schoolchildren (wawa) into “conquered heads.”
But according to the community’s own interests, the children, through their con-
stant prayer, retransform the ancestral remains buried in the school (and other
ritual sites) into the communal body and its own property in perpetuity, manifest
in the new harvest of babies.
Inasmuch as the Hispanic Crown and Church extended the body of Christ
and the Spanish language in common toward the limits of empire, the pueblos
on the periphery reacted against this imperial intent. Perceiving it in terms of a
struggle “to eat or be eaten,” they introduced in the school cycle of learning their
own form of counterattack by “butting the letters,” incorporating them as their
enemies, to digest during the school year and finally regurgitate in the end of
year exams. Paradoxically, all this occurred as a consequence of ecclesiastical in-
doctrination in native languages about “the letters made flesh,” according to the
teaching of seminal thought at the heart of Andean Christianity, directed at the
mass of the body of Christ and the celebration of his Resurrection.

Th e Co rp oreal i ty of Kip us
Toward a M athematics Incarnate

The images on the interlaced headbands are a template for the repeating patterns on the em-
broidered fabrics from Paracas Necrópolis and.......both encode similar classification systems.

Mary Frame, “Structure, Image, and Abstraction”

We examine now regional textual practices in their own right, including the nu-
merical operations they facilitate. First we describe the social and corporeal logic
of kipus, both in historical examples and in contemporary ethnographies of the
region, where there do seem to be certain analogies in key concepts between his-
torical and contemporary practices. Then we describe the numerical and mne-
monic systems that might underlie the cultural and textual logic of the knotted
kipu, and finally the uses of kipus and textiles in the cognitive and didactic sys-
tems in the region of Qaqachaka today.
In each case, we show how regional numerical practices, like textual prac-
tices, are related directly to productive and reproductive activities, and hence with
the same farming, herding, and ritual cycles concerning death and revivification
that we examined so far. As a “pastoral discourse” (in Brotherston’s terms), kipu
language is concerned directly with the productive base of pastures and waters
and the fleece quality of the flocks, while other underlying aspects of their use
concern agriculture and warfare, and associated ideas about the potential fertility
of the trophy head as a seed that sprouts.
In addition, the ways that Andean populations use kipus in the generation,

208 A N D E A N T E X T S A N D T H E I R I N T E R P R E TAT I O N

measurement, and scrutiny of their productive cycles indicate that the economy
with which these numerical practices intersect, relates many domains simultane-
ously, such that different notions (of value, circulation, wealth, sacrifice, war, pro-
duction and reproduction, the social and the corporeal) are all based in the same
symbols and the same organic language. This permits us to understand a regional
interpretation of “capital” and its proliferation into a multitude of babies.
Understanding the dynamics between the numerical and writing practices of
Andean populations must also take into account that, in the Inka context, both in-
formation systems were based on the same medium of communication, namely
the kipu. Research up till now has delved more into the kipu’s function in nu-
merical practices; here we consider how kipus might form a conceptual bridge
between numbers and writing.
This entails considering in turn how, in the absence of Western writing and
arithmetic, Andean textual practices functioned on the basis of lived orality and
quite different kinds of texts and writing (kipus, weavings, braiding, libations,
songs). We maintain that textual practices in the Andes do not distance the voice
from the object of research, as do European writing practices; rather they sustain
the processual flow between word, sound, body, and mimetic image that cannot
be abstracted easily into objective signs and fixed meanings.1

The Braided Corporeality of the Inka Political

a n d M i l i ta r y H i e r a r c h y

We first sketch the possible organization of some historical kipus at their po-
litical and symbolic levels, using evidence from different sources. Our contention
is that the social and corporeal codes of the kipu indicate key ways (little exam-
ined until now) in which numerical practices articulated with other levels of or-
ganization of the Inka state.
In the Inka state, the pinnacle of political organization was the Inka himself.
This hierarchy was expressed at a corporeal and symbolical level in the Inka’s
head and hands, the woven expression of which is found in the organization of
Inka kipus, where the principal cord acts as the “head.” However, this idea dates
back to pre-Inka braided kipus and headdresses. Mary Frame’s research (1991) on
headdresses from the Paracas Necropolis is pertinent here as it demonstrates that
the headbands called wincha, whose weaving structure is highly complex, consti-
tuted a template for the patterning of other garments that draw on similar clas-
sification systems. In our case, if the principal cord of an Inka kipu were to repre-
sent the Inka’s head, then we can begin to understand the organizational logic of
other levels of the same classificatory system.

In his Book of the Fourth World, Brotherston cites some evidence for this so-
matic interpretation of kipu structure, centered in the Inka’s head and hands,
from literary fragments of the colonial drama of Ollantay.2 In one example from
scene 5, the kipu’s main cord is compared to the crown of the empire (llawt´u/
llauta) and the pendant threads with its subjects, “all united with him” suspended
as “kernel knots” (ruru), like the pips of some fruits left after the fleshy part has
been eaten.
caycallauta As to the main cord,
nam kahuahua cay umanpi huatascana so the skeins that are bound to his head;
cay rurucanari as kernel-knots
runam tucuy payman tinkiscana the men all united with him.

In the second example (scene 14), the Inka’s hand grasping Antisuyu (having been
conquered by him) is compared with three fivefold knots of a black kipu:
cay kipupim can killimsa As the kipu is carbon-black,
nam Ollantay rupascana so Ollantay is burnt;
cay kipupakmi kimsa piscucuna huatacana as the kipu is bound with three quintuple
nam Antisuyu hapiscu so the Antisuyu is taken
nam Inca makeykipina: so it is in the Inka’s hand:
chaymi huatacuncaypisu here the quintuple knots,
kinsa piscu tucuypinas. three fives altogether.

Here, the main kipu cord expresses the power of the Inka’s head and hands, while
the knots of the pendant cords seem to represent some aspect of the various sub-
jects, in a top-down hierarchy.
These brief fragments possibly indicate other aspects of Inka hierarchy. In
terms of the trophy head complex and ontological depredation, the main cord
might represent the Inka’s head, from which the subjects as kernel knots hang
like booty. Such a use is suggested in the following reflection by Platt (1987a) con-
cerning the early seventeenth-century terminology of an Aymara chinu (equiva-
lent to an Inka kipu), in which the Indians of the lower segments of a moiety
“hang” as subsidiary cords from a main cord, related this time to the highest rung
of Aymara authority, namely the mallku-condor: “The captain with his multiple
dependants seemed to be a great mallku: his followers would be the campesinos
themselves of the lower segment of his moiety, that ‘hung’ from him as the sub-
sidiary cords of the administrative chinu” (1987a, 91).
The comparison between “quintuple knots” and the Inka’s hand echoes the
custom of Inka warfare, of procuring the skull of the enemy and using it as a
drinking cup (qero), which the victor (in this case the Inka) grasped in his hands to
210 A N D E A N T E X T S A N D T H E I R I N T E R P R E TAT I O N

drink festive corn beer. (See the ceremonial song voiced in this situation in chap-
ter 1.) In his Relación del descubrimiento y conquista de los reinos del Perú (1571), Pe-
dro Pizarro’s account of the battle in Ambato between the forces of Atahualpa
and Huáscar implies the same use, when the generals from Quito killed someone
called Atoco, and “a vessel was made from the crown of his head.......for drink-
ing, mounted in gold.”3
The Inka royal insignia, as described in the chronicles of the period and stud-
ies in the history of art and weaving, tell us more about the significance of the
Inka’s head and hands in the political hierarchy of this state. Guaman Poma’s de-
scriptions in his Nueva corónica of the dress of different Inkas, and the royal attri-
butes related to their headdress and the arms carried in their hands, based on the
reading of a kipu, consist of three parts, each one with its proper order ([1613]
1989, f. 84, 85, 87). More pertinent for our purposes is the first part concerning
the royal attributes, arms, and dress of the Inka related to his majesty in war. The
order of the costume description began “from the bottom up,” and then the de-
scription of what the personage carried in his hands, in the order “from right to
Guaman Poma informs us that the Inka’s headdress had various components,
including his headband (llawt'u), royal tassel (maska paycha), the feathers (of his
parasol) and woolen decoration, together with his ornamental flower (wayuq
tika) and earplugs of the finest gold. In the right hand, he used to carry his axe
(kunka khuchuna) and in the left his shield and ceremonial halberd (champi).5
Other sources hold that the elements of his headdress expressed the insignia
of state power, mediated by warfare. Thus Pedro Pizarro’s descriptions in 1571
of his encounters with the Inka emperor Atawallpa mention how the braided
headband or llawt'u (similar in appearance to the main cord of a kipu) served as
a “crown”: “This Indian (Atahualpa) put on his head some llautos, that are braids
of colored wool, the thickness of half a finger and the width of one, as are some
braids as I say, and this made in the manner of a crown, and not with points but
rounded, the width of a hand, that covered the head; and on the forehead a tassel
sewn to this llauto” (ibid., 66).6
As a primary symbol of the Inka’s birth to power, the llawt'u tassel’s crim-
son color was an expression of dynastic hierarchy, at least of the mythical dynas-
ties of the founders of Cusco, where Inka culture first developed its control over
herding and weaving production.7 It also expressed the Inka’s dominion over re-
gional flocks, when local sacrifices served as an opportune moment for revealing
his authority. Cieza’s description (1550) of a communion of sacrifice describes the
moment when the Inka took the tassel for the first time, and notes the correspon-
dence between the meat and blood of the sacrificed animal in the feast, and what

would happen to any enemies of war.8 Cieza narrates how the chiefs (los princi-
pales), on taking the Inka’s tassel, ate the raw meat and blood of sacrifice; in the
same way, their enemies would eat the flesh of the chiefs themselves, in the eter-
nal cycle of eating or being eaten.
Other chroniclers took pains to explain how the tassel was made with the
finest wool, dyed with threads twisted perfectly and all cut the same, and mixed
with golden beads, serving in its totality or in its parts, as the medium of com-
munication of the Inka’s “highness.”9 Yet others related the headdress’s “tuft of
wool” directly with enemy hair and heads. In a rite that Cieza narrates ([1553]
1984, vol. 2, pp. 18–19), the same tuft of wool was related to a rough bunchgrass
that grew on the hilltops, and with a rite of passing from hill to hill, points still
emphasized in present-day kipu practices.
In his Historia del descubrimiento y conquista del Perú (155510), Agustín de Zárate
tells us that the crimson threads of the tassel served, in the hands of the Inka long-
eared elite (the orejones), as a proof of their closeness to the Inka, and evidence
that the orders that these might give would be accepted and fulfilled as if the Inka
in person had dictated them. This implies that the tassel’s crimson threads had a
very direct relationship with the Inka’s voice (rather like Atawallpa’s “crest,” in
Elvira Espejo’s description). For Cieza, the very act of taking the tassel signified
taking possession of a territory, and hence the reaffirmation of the Inka’s power
([1553] 1984, vol. 2, pp. 211–12). So the royal tassel, with its direct allusions to the
camelid flocks and more indirect relations to the wild bunchgrasses that grew in
the lands under his dominion, combined the distinct elements of farming and
herding production in a single semantic domain. In that same domain are found
the customs of war, both of territorial expansion and in its defense (by the ore-
jones), mediated by the power of the Inka’s voice.
In brief, the crimson royal tassel, together with the braided crown (llawt'u),
functioned as a kipu template in miniature. The llawt'u served as its main cord,
while the pendant tassel (like pendant threads) worn on the most powerful fore-
head of the empire, expressed the Inka’s power to take trophy heads, in a visual
symbol of what would happen to you as subject, if you were to contemplate en-
tering into conflict with its user.

The Corporeality of Modern Kipus

Using these historical contexts as our point of departure, we can proceed to

analyze modern kipus and their Aymara equivalents, called chinu. Our sources
in this case are limited and there are many voids in the information, but the frag-
mented memories of handling historical kipus, considered against a backdrop
212 A N D E A N T E X T S A N D T H E I R I N T E R P R E TAT I O N

of changes in kipu practices over the centuries, still suggest a common logic
whereby some kipu structures follow a corporeal model of Andean state organi-
zation, with its chiefdoms and federations headed by Cusco, and handled by the
Inka, through his local representatives.
For example, in the community of Don Domingo Jiménez, until recent de-
cades the kipus used in the annual census of the ayllu population had to pass to
the second authority in charge (curaca, considered to be “their father”) and then
to the regional mallku. At the present time, as Don Domingo says, in the absence
of the upper levels of the political system “they only record the mountains” as
“stand-ins for the Inka.” In a variant on this practice carried out in Qaqachaka a
generation ago, the reckoning of the flocks in the annual census that took place
during the animal-marking ceremony (called k'illpha) was passed on to the church
caretakers called mayordomos as the political and religious authority immediately
superior to that of the household.
In the same region, homologies between the territorial jurisdiction of some
kipus and the political and ritual domains represented by the highest mountains
of the zone reiterate the part kipus played in territorial organization within a
wider political and ritual hierarchy. For example, in different communities of
Oruro department, “in the absence of the Inka,” the territorial domain served by
a kipu is considered to be under the dominion of the highest mountains, called
mallku. Don Mario, a modern kipuqamayuq of the region, narrates his experience:
“This is from Challapata.......yes, from the mountain of.......Azanaque, we handle
it just from Azanaque to Negro Pabellón, all around there, they tell me. That’s
what they’ve explained to me. Now other communities.......they have another
class of kipu. So they’ve told me.”11
Comparatively, Salomon’s study of present-day kipus in Huarochirí (Peru)
shows that when they are exhibited in public, they are placed coiled up in the
pyramidal form called “mountain” or “pedestal” (peaña) (1997, 6). This is similar
in shape to the pyramidal base of the school flagpole in Livichuco. All these ex-
amples imply that the hierarchical organization of the Aymara chinu or the Inka
kipu had certain homologies with Andean political organization, and that the
mountain served as a principal symbol of hierarchy.12
As for the more localized interrelations among ethnic groups or ayllus, Don
Mario from Oruro pointed out another connection between the kipus and war,
in the manner of forming a fighting knot, by tying together tightly two pendant
threads. He implies that the pendant threads may represent other subordinate
territories in a state of conflict: “it’s different, look at the knot, it’s different, this
should be loose, it must be loose........Only the white connects to this color; when

it’s loose, no.......there’s no fight, because the fights are these [where the two
threads are knotted together].”
The modern color codes of a kipu also allude to political hierarchy and war.
For example, just like the Inka’s tassel, the color blood red (or purple) of the mod-
ern kipu that Don Mario handles signifies war: “This blood red is not tied much,
because war between them [would occur]........They don’t use red, they use a
violet, this one, a purple, don’t they? Violet, it’s called violet, they use this pur-
ple for fighting.” The same commentary can be heard in Qaqachaka. For Elvira
Espejo, “red is always war because it’s blood, isn’t it?” The symbolism is the same
in weaving, given that certain uses of the color red unfold the “ribbon of war” or
“warpath” (wila sinta).
Even in the modern handling of kipus (or chinus), comunarios record how the
Inka handled the knotting of threads in his function as maximum authority. Don
Domingo mentioned once, on ending a series of libations, that the Inka used to
“hold” the knots of the threads as if they were the heads under his hands (and he
emphasized the function of the little finger in forming the headlike knot). His
commentary implied that one of the goals of elaborating libations in the pres-
ent day is that of “putting a head” to each food crop so that it produces well, in
memory of what the Inka accomplished in the past. Another example of this use
occurred in the course of a tale of Inka times narrated by Don Donato Inka of
Livichuco, when he expressed the Inka’s dominion in terms of the “reach” of his
They say the Inkas used to produce silver and gold, I’ve heard this. They knew how to sow
it under the ground. These terraces exist right until now and it’s there that they used to
sow the silver.......and the Spaniards took his money, all of it.......they even killed the Inka.
And the Inka pleaded with them: “Don’t kill me, I’ll pay you silver and gold, up to the reach
of my hand, I’ll pay you up to there; don’t kill me,” he said (our emphasis).

T he H ands of M odern K ipus

A similar notion of circulating wealth “between hands” can be heard in the

more mundane instance of making toasts for money, when the language about
the circulation and growth of money as wealth is homologous with that about
the organic growth of food crops in the ground. This occurred in the Livichuco
weaving organization, APSU, on 1 August 1998, when they made toasts “For Bo-
livian pesos, the young girl who is wont to go from hand to hand.” Don Feliciano
Maraza explained: “This is money (phaxsima). For example, if we have ten boli-
vianos (just over a dollar) then this goes from hand to hand; we’re referring to
214 A N D E A N T E X T S A N D T H E I R I N T E R P R E TAT I O N


1 10 100 1.000 10.000
lower finger mid-finger

Fig. 21. The hand and kipu counting (from Kaufmann Doig 1969, 23).

that........It’s the ‘restless little Inka girl’ (inkit imilla) that goes.” He went on to
compare the circulation of money among hands with the biological notion of
“rooting.” First you plant something, and then “it goes on taking root, so one
thing develops and then another, expanding in this way.” His logic derives from
the idea that manual labor works the land, and produces wealth from it.13
Another clue to the manual practices of measurement can be found in Lara-
marca, Huancavelica (Peru) where kipus based on the model of the hand were
still in use in the 1960s. Studies on these kipus by the Peruvian ethnologist Froilan
Soto Flores show that the joints of the wrist and knuckles are translated in the
kipu knots, with simple digits coinciding with the wrist position, the tens in the
next joint where the thumb stems from, and the hundreds, thousands, and tens
of thousands in the following joints (figure 21). In this case, the kipu should be
held with the left hand and its knots related to the knuckles of the same hand.14
Contemporary practices of kipu reading in Oruro and northern Potosí (Bo-
livia) provide more evidence for a corporeal dynamics, including the joint action
of fingers, hands, and gestures. There, kipus are also read from right to left, with
the input of certain items on the right, and the output of other items on the left
(similar to the Xauja kipu analyzed by Murra in his classic essay of 1975, and
the way in which Guaman Poma “reads a kipu” in his descriptions of the Inka’s
clothing).15 This spatial orientation of the kipu, linked to the use of the hands,
is reiterated in contemporary sayings when the right hand is associated with in-
coming money and the left hand with releasing it (figure 22).
The same logic operates in kipu color codes concerning “loans” and “debts”
(jucha), with their symbolism of white and black, respectively.16 In present-day


manner of reading

outgoing of items entering items

Fig. 22. Left- and right-directed operations.

kipus, this color coding seems to function at the level of gender (white = male,
black = female), resulting in a comparative categorizing of threads and knots ac-
cording to the use of these two colors. It is also possible that the hierarchical cir-
culation of marriage partners between groups has been applied to a previous sit-
uation of war and conquest, in which a woman taken (mi negra, “my black one”)
demanded in response the manual labor of a son-in-law (the white) in bride ser-
vice, to break up clods of earth and open the land for sowing, and to contribute
his physical energy in favor of the ayllu group, in case of conflict. Such a system
of relations would explain the additional meanings of juchha (debt), mentioned
by Platt (1987a, 90): as “something broken down” (to “pulp”); or huchhacha uma-
chatha as “grinding someone down with beatings.”
Another possibility is that the base five of many pre-Inka kipus is founded
on the number of fingers on the hand.17 In Aymara counting, as in many Am-
erindian languages, there is evidence of a base five, probably the way of
counting on the fingers before the incorporation of Aymara confederations to
the decimalization of the Inka state (see figure 23).
If Quechua and Aymara numbering systems are compared, there is clearly
a decimal system in the former, while the latter conserves features of a base
five system. So paqallqu (seven) is composed of two terms (pa “two,” and qallqu
“five”), and the same happens with kimsaqallqu (kimsa “three,” and qallqu “five”).
In the case of llatunka (nine), Bertonio gives as the meaning of llalla “decreaser”;
therefore lla “minus one” plus tunka “ten” equals nine.18
216 A N D E A N T E X T S A N D T H E I R I N T E R P R E TAT I O N

Quechua Aymara

1 uj maya
2 iskay paya
3 kinsa kinsa
4 tawa pusi
5 phisqa pisqa (= qallqu?)
6 suxta suxta
7 qanchis paqallqu = 2+5
8 pusaj kimsaqallqu = 3+5 Fig. 23. Quechua and Aymara
9 isqon llatunka = -1 + 10 numbering compared (cf.
10 chunka tunka
Urton 1997a, 42).

These manual or body-based readings of kipus form part of a widespread

and dynamic logic in which even the reader’s body is subordinated to the com-
municative interests of the higher levels of the system.19

T he H eads of M odern K ipus

Apart from the emphasis on hands, many comunarios in the region of the
present study compare the corporeal structure of a kipu (or chinu) with a trophy
head. In certain contexts, the principal cord and the knots on its pendant cords
are called “head” (p'iqi), and the pendant cords are compared with its hair.20
This is part of a wider kipu corporeality whose homologies function in terms
of positioning. For example, in the experience of Don Domingo, the modern
kipus he handles do not just have a “head,” but all the other body parts. He also
distinguishes between an untied thread “without a head,” (jan p'iqini), and a tied
and knotted thread that “has a head” (p'iqini), implying that each tied “knot”
(chinu) is a “head” (p'iqi).
Regarding positioning, he says that the knotted “heads” cannot be in just any
position on a cord; they should be “at the top” and then “they are lucky.” In con-
trast, the knots in the middle of a cord only have “half luck” (chika surti), being
at the “heart” of the cord, while the knots at the bottom of the cord are “like the
feet.” The different knots along the cord, then, adopt the relative positions of the
different body parts. Even the head knot at the top of a cord is differentiated into
distinct parts with different functions: the eyes, the ears, the mouth where the
breath (and luck) comes out, and so on. For him, this allows the head knot tied on
the cord to “think” and “speak.”
This semiotic organization of a modern kipu, with its head, body, and feet,
is not only a figurative trope. The logical positioning of these elements accord-

ing to a familiar hierarchical language, as a necessary stage prior to the organiza-

tion of discourse, is precisely what the computer codification language of HTML
does, in order to organize infinite quantities of information flow on the World
Wide Web.

O ther T erminology of M odern K ipus

In Don Domingo’s community, present-day kipu terminology embodies

other fragmented memories of the political order, thought of in terms of homol-
ogies with the natural world (human and nonhuman, male and female, progeni-
tor and offspring), particularly when comunarios talk about the dynamics of kipu
use and meaning.
In this dynamics, a primary classificatory division occurs between kipus that
relate to human and nonhuman domains. So when Don Domingo speaks of the
cords related to animals, he names their divisions after different birds of the high-
lands (the mountain caracara or allqamari, the condor-mallku, and the humming-
bird),21 but when he speaks of kipus related to people, he names their divisions
according to the store contents of farming produce. Don Domingo also calls the
principal kipu cord allqamari (the caracara bird) when it refers to animals, and yet
“storeroom or deposit” (almasina) when it refers to people. Another classificatory
division of kipus is into gendered categories that concern the flow of masculine
and feminine substances.
Kipu organization also distinguishes a genealogy of progenitors and offspring,
rather like that found in the genealogy of written documents and weavings.
The same semantic domain refers to the proliferation of kipu “babies” (wawa),
guarded like trophy heads, according to the textual theory at the heart of onto-
logical depredation. Here, modern-day kipu terminology replicates that of the
“frames” of HTML codification on the World Wide Web, organized according to
a similar generational hierarchy of grandmother, mother, and baby units.
Other instances of present-day terminology relate kipus directly with rights
to land. The original Arabic term almacén, borrowed into Aymara, refers in gen-
eral to a “storeroom,” but its more juridical sense concerns land rights; the “gen-
eral warehouse storerooms” (almacénes generales de depósito) were establishments
where merchandise was stored against the presentation of a title issued to the de-
positors. This refers us to the same semantic domain as the deposits of commu-
nal capital mentioned in chapter 7, inscribed in the communal kipus guarded in
chests under three locks and keys.
In other cases, kipu terminology refers to the community flocks. In the kipu
samples that refer to animals, Don Domingo holds that the principal cord called
218 A N D E A N T E X T S A N D T H E I R I N T E R P R E TAT I O N


mother (tayka)

Lord sprout
(ch'illch'i mayku)

Lady sprout
(ch'illch'i t'alla)

Fig. 24. An Aymara chinu with its primary and secondary cords.

allqamari “handles everything,” and he compares it to a sling that is braided simi-

larly (somewhat like the Inka’s crown). For him, there are two kinds of principal
cord called allqamari. One, braided only in white and black, has the function of
“tying things really tight” (mat'iña); the other “must have all the colors” so as to
“bloom” like the animals, during the libations, when such kipus form the point
of reference for the toasts.
Apart from these ornithological and zoological references, the dynamics of
kipu reading appeal to the organic images of botanical proliferation in seminal
thought. For example, Don Domingo considers the thicker cords that descend
from the main head cord to be “mother cords” (tayka), while the secondary
threads that branch from them (saraqaña) are their “sprouts” (ch'illch'i). He de-
scribed how these are “like branches stemming from the main trunk of a tree”
(somewhat like Don Feliciano’s commentary about the proliferation of money
from similar organic roots). The subthreads that stem immediately from the
mother cords he names “lord shoots” (ch'illch'i mayku), as they are “male” (urqu),
while the secondary threads deriving from these he calls “lady shoots,” as they
are “female” (qachu). In another botanical allusion, he considers the threads that
descend in this way to be “blooming.” Even so, the Aymara verb ch'illch'iña, “to
sprout,” refers more specifically to the filtering of water, as if the threads had
some liquid inside them, like stems of plants. See figure 24.

The terminology concerning the liquid medium of threads/shoots implies

that a liquid medium lubricates the flow between the different levels of the higher
ordering system of which the kipu is part, in a similar way to what happens in
the rainmaking prayers and libations described in chapter 8. There we postulated
that a centralized Inka control over local production in the past was particularly
concerned with prayers to call the rains, given that if the rains did not arrive, then
there would be no production from the land (farming produce, pasture for the
flocks, fleece or weaving). Here both the terminology and the texturing of the
kipu draw on the same productive media and its proliferation (farming produce,
pastures, and fleece) connected by a substratum in common, namely water. Once
more, in regional textual practice, the “medium” of communication of the kipu
is homologous with the “message.”

K ipu C olors

The existence of determined norms in the use of color in modern kipus is

unlikely, given that each kipu reader or kipukamayuq with whom we have had
contact has his own idiosyncrasies, and this has probably always been the case.
Nonetheless, in local discourse about the colors of kipu threads, there are certain
homologies with the color coding encountered in other textual practices (weav-
ing, singing, and braiding). It is also common to hear that the practices of color
use in the kipus of the past were similar to those of today. The organizing crite-
ria behind these similarities might have to do with another pervasive idea, that
color was related both in the past and in the present to the domain of the sun and
moon (the Inka gods), and the communal obligations of lands, flocks, and war-
In Qaqachaka, they say that green threads represent pasture, red ones blood
and war, and yellow ones the sun, and that to send black threads indicates that
“whosoever would arrive at night,” while such threads mixed with white indi-
cates their arrival “with a moonlit night” and so on, just as in weavings.
For his part, Don Domingo relates the symbolism of kipu colors with the
same productive domain of land and waters, and hence with the Pachamama,
who tends the flocks, giving them pasture and waters. “The orange thread (aru-
mi) represents the Pachamama. The reddish ones represent the wet pastures,
and the green one the algae in the water. Red represents the blood of the Pa-
chamama. Cherry red is so that the food crops sprout.” For him, the color of any
determined kipu thread also replicates the color an owner “desires” for the ani-
mal denominated in that thread. This same nexus between naming the fleece or
hide color of an animal, and the desire for this color on the part of its owner is
220 A N D E A N T E X T S A N D T H E I R I N T E R P R E TAT I O N

found in the songs to the animals.22 Commonly these desired colors are named
in the libations to the animals, according to the color of the guiding thread. In
this practice, as in the songs, there are strong ties between color and voice. Given
that the llamas “have many colors in their fleece,” then more colors are named in
their threads, and more colors in their libations. As a result, the toasts to the lla-
mas “are much longer.” In comparison, the threads that refer to sheep only have
“three colors,” and so there are fewer libations to them. The same logic holds
that mixed colors (ch'ixch'i or wanakillu) “give more strength” to the animals, and
these special animals with multicolored coats are toasted apart.
This notion of multiplicity expressed by naming many colors, or mixed col-
ors, is a particular feature of Andean numbering. In contrast, Western number-
ing tends to fragment this same reality into its different elements. Perhaps the
multiplicity of fleece colors mentioned on these occasions expressed, in the past,
the potential strength of generating offspring on the basis of the greater number
of trophy heads captured, with their different hair colors.

K ipu K nots

Apart from the codification of color and its combinations, other subsystems
of kipu codification denominate the class and size of a knot, the direction of
spin of a thread (S or Z), and the tying of a knot (to the right or left). There is
also a basic difference between making knots “in reality,” as a form of “count-
ing” (wakhuña), and making knots “in the head or heart” by following a prede-
termined pathway (thakhi), as a way of “recording” (amuyaña). This difference is
fundamental to the social and cultural context of numbers, numerical practices,
and a regional art of memory.
In Don Domingo’s community, recording the “knots of memory” entails fol-
lowing the pathway of luck called chinu ch'alla, meaning literally “toasting knots,”
when the different animals are remembered. The act of mentally “tying a knot”
(chinuntaña) in a thread “makes you receive it in your heart” (chuymar katuqaña),
and not forget it afterward. (Don Domingo compares this effect with recording
the trace of an idea by writing it down.) At the same time, the very act of making
the knot “grasps luck” (surti). You should never untie the same knots afterwards
“because this can release the luck.” Although some animals are butchered, their
knot should never be untied, because “it would untie all the rest of the flock.”
The relation he makes between tying a knot in a thread and receiving luck im-
plies that something is being tied into the knot. It is as if luck were being captured
there in the knot, like a prisoner of war.
In all of this, the dynamics of making the toasting knot, while at the same

time tracing the pathway of a particular animal, seems to replicate the hierarchy
of the state. As Don Domingo puts it, in following the animal’s pathway, you fol-
low the knotted pathway of a predetermined thread (“you follow the pathway of
the cow, the llama, and all the rest”), and there “you encounter the Pachamama
and the Guardian Mountain.” This is when you finally return to the principal cord
as maximum authority of the kipu, to encounter there the Pachamama and the
Guardian Mountain, as modern stand-ins for the Coya and Inka of the past.

T ypes of K nots and T hreads

Conversations with two kipu readers, Don Domingo from northern Potosí,
and Don Mario from Oruro, revealed that different classes of knots and threads
are used for different purposes, according to the same Andean political hierarchy.
First of all, a generalized ordering of knots and threads in the libation sequences
facilitates the memorization of toasts. Here, some “go in pairs on the heel of
each other, so that the associated libations go in pairs,” while others “go further
apart, so that the libations are separated.”
As for the threads, the longest represents the biggest animals (and “a great
deal of luck,” jach'a surti) whereas the shorter ones represent the babies or off-
spring (and a “smaller degree of luck,” jisk'a surti). Similarly, the thicker threads
are said to have more strength (and their descendants will be strong too) whereas
the thinner threads are “weak” (indibli), like their descendants. Unlike Don Ma-
rio, Don Domingo insists that all the threads “have to be left straight always” and
not twisted with others, as “it would be twisting luck” and you could “wipe it
out” in this way.
In his classification of the knots, Don Domingo identifies those “of an aver-
age size” and others that are “double and bigger.” The standard knots are used as
points of reference in the sequence of libations, and they are always tied toward
the right (kupiruw chinuntañax) as this “holds onto luck” (surt katxasix). In con-
trast, a knot made toward the left “would release” the animals. These comments
confirm the same braiding logic of right and left that we encountered before.
The bigger double knots represent the “pens” (uyu) of the animals, whether
of llamas, cows, sheep, or goats. In a ritual context, when the “corner council
rite” (iskin kawiltu) is carried out, these large knots are called turnings (rutiyu) or
the still more elaborate name: “court of white and yellow gold” (chuqi kancha,
quri kancha). These large knots are tied at the ends of a thread where they act as
the “destiny” of all the middle-sized knots named in the libations sequence. They
are “pens of knots.”
Don Domingo identified similar large knots at the ends of threads, which also
222 A N D E A N T E X T S A N D T H E I R I N T E R P R E TAT I O N

libation knots
(ch'alla chinu)

pen knots
(rutiyu chinu)

Fig. 25. The movement of the libation knots toward the pen knot, according to
Don Domingo Jiménez (Pimentel 1998, 99).

had other middle-sized intermediary knots, in some historical and modern kipus,
for example in the drawings attributed to Guaman Poma, and in the book Code of
the Quipu (Ascher and Ascher 1981). This kind, that he calls “turning knot” (rutiy
chinu), is destined for the long series of libations called ch'alla chinu. (See figure
25.) This observation, at once practical and profound, has not been taken into ac-
count until now in studies of kipus.
Other classes of knot derive from gender relations. “Female knots” (qachu
chinu) destined for the female animals and “tied with the left hand” (lluq'i am-
parant chinuñawa) are looser and easier to “untie” (antutaña) “because the female
is for procreation.” The female knot is “so that many are procreated, and so that
she releases them. She may easily give birth to, and release offspring on all sides.”
In comparison, “male knots” (urqu chinu) destined for the male animals are tied
with the right hand (kupi), tighter and “very hard” (wali turu), and have “a lot of
strength” (ch'amanichijaya), so there is no way of untying them (figure 26).
These commentaries reiterate the logic of kipu directionality, in which the
right is related to masculinity and with “grasping and not letting go,” while the
left is related to a femaleness “that releases easily.” This would also seem to locate
numerical practices within a wider theory of textuality, where masculine prac-
tices concern tying strong knots for plenty of luck, and not letting them go, as
if trapping the strength and luck of prisoners of war there in the knots, whereas

male knots female knots

(urqu chinu) (qachu chinu)

Fig. 26. Female and male knots, according to Don Domingo Jiménez.

feminine practices concern the release of this trapped strength and luck, which in
turn help to proliferate offspring.
Don Domingo compares this gendered distinction between knots with fun-
damental physical differences between women and men. For him, a woman is
always “of less strength physically” and not “really hard” like a man. In particu-
lar, he attributes the slightly loose quality of a female knot to the corporeality of
women and female animals, comparing it to the mucous plug of the cervix that
“bursts first of all” when they give birth. When a woman gives birth, “this makes
the mucous plug (t'awaraku) burst easily and then the baby is born, the mucous
plug acting as the “foundation” (thaxsi).
This observation gives us clues to a more female language about kipus, in
which female knots also serve as the foundations of the world where the new
harvest of babies is generated.

T he K ipu D ocumenting of D eath and R egeneration

In the context of an Andean theory of textuality, a key relationship between

knot, number, and writing would seem to concern the multiplication of one’s
own lucky elements (as wawas) from the transformed energies and luck of a cap-
tured Other.
224 A N D E A N T E X T S A N D T H E I R I N T E R P R E TAT I O N

Striking cognitive relations in the development of writing and numerical sys-

tems suggest that this is the case not only in the Andes but also elsewhere. As
many have pointed out, the development of numbering went hand in hand with
that of writing, resorting to the same textual bases.23 This is particularly notewor-
thy in studies of the development of accounting systems. For example, the major-
ity of cuneiform ideograms in the Middle East contained temple accounts. Urton
observes that in the city of Uruk, some 90 percent of examples from the archaic
period are clerical archives, in a combination of verbal formulae (writing) and nu-
merical signs. More pertinent is that in cuneiform writing, the graphic form for
“dead” or “to die” evolved from the sign for the numeral one, superimposed with
a wedge (1997a, 31). This pattern is known in other ancient forms of writing, for
example in Linear B of ancient Greek, where numerals are often accompanied by
syllabic signs and ideograms.24 This suggests that beyond the Andes there might
have existed the same relation between the beginnings of the bureaucratic state
apparatus and the death of the Other, which in turn helped generate the prolif-
eration of one’s own objects that could be counted and deposited.25
The disadvantage of archeological sources is that we can only guess at the
corpus of knowledge that served state administration. By contrast, in the Andes,
it is still possible to unravel the threads of various discursive practices and their
changing institutional bases, in which ideas about corporeality, number, and writ-
ing proliferate.

Kipu, N u m ber, and W ri ti ng

The absence of knots might bear a relation to the finishing and perfection of the
weaving........[I]n the realization of the kipu, this feature is pregnant with meaning.

Sophie Desrosiers, “Lógicas textiles y lógicas culturales”

Now we examine the numerical and mnemonic systems that underlie the pro-
cessing of information through practices based in kipus and textiles, in the con-
text of an Andean theory of textuality. These include the learning of these prac-
tices by young people at a pragmatic level in rural communities, as well as the
more metaphysical dimensions of the discursive practices in which they are im-
mersed. Both these aspects derive in turn from knowledge systems and con-
trolled sets of numerical and mnemonic practices in the hands of authoritative
practitioners that follow certain key institutional structures in order to facilitate
their reproduction over time.
We also propose that ritual practices to do with promoting desire and good
luck in agriculture and herding are just as much part of a regional theory of tex-
tual practice as are everyday arithmetical skills in key stages of counting and mul-
tiplying the material wealth that derived originally from a captured trophy head.

Andean Numerical Practices

T eaching and L earning N umerical P ractices

Mathematical procedures are learned in the community in ways quite distinct

from those of the classroom in terms of gendered differences, counting termi-

226 A N D E A N T E X T S A N D T H E I R I N T E R P R E TAT I O N

nology, and the epistemological notions behind counting practices. This creates
a conflict between numerical practices that the educational reform has not yet
been able to resolve.
We saw in chapter 5 how certain gendered differences relating to the integra-
tion of counting with other textual practices derive from learning procedures.
Girls gradually learn numerical practices by “counting threads” (chin wakhuña)
in weaving and braiding; boys by counting threads in weaving, the knitted details
of their woolen caps, the strands of rope work, the measures (called portadas) of
rough homespun, and the rhythm of instrumental music.
Importantly, boys in the community of Don Domingo Jiménez as well as in
Livichuco learn to count (wakhuña) in parallel with learning to make libations
(with water), from five or six years old onward. They begin counting with animal
droppings, and replicate the animal pen with stones, calling them by the ritual
name rutiyu. Don Domingo commented how “They paint with charcoal on flat
stones, practicing in this way,” and how they use water in small cups for toasting,
requesting threads from their mothers to imitate the making of knots. Progress
proceeds by the regional equivalents to the techniques of series, sets, adding, sub-
tracting and multiplying, groupings in odd and even numbers (pairs), cardinals,
and numerals.
Such differences in learning are a consequence of other fundamental dis-
tinctions between Andean and universal mathematics. One factor is that An-
dean techniques of learning are not abstracted from social reality, as is universal
mathematics, and so it is not easy to insist in the classroom that children “count
for counting’s sake.” Another is that many techniques of learning are handled
in three dimensions, as when rural boys and girls have the ability to read, an-
alyze and improvise on the complex three-dimensional structures of cloth or
rope work in a more proficient manner than their urban peers. There are cases
of young weavers who, by just looking at a photograph of a weaving, have been
able to replicate the cloth in question. Such is the case with Elvira Espejo, who
copied the complex textile structures of archeological headbands (wincha) from
Paracas just by looking at their photographs in a book.
Numerical practices are also embedded in a wider set of cultural practices
that include a gendered division of labor. From the observations of Don Do-
mingo, the numerical practices of young people are a precursor to the adult prac-
tices of making libations for good luck. In the gendered division of labor of these
adult activities, men’s libations must be accompanied by the women’s singing, in
the same way as we encountered in the ritual complex of making libations and
praying. And while the men tie the knots (symbolically) to end each series of liba-
tions for luck, the women begin to dance, as a parallel activity.

In Qaqachaka, this division of labor is not so exaggerated, as the women toast

amply too; even in-marrying daughters-in-law have to learn from older women
to toast well, as a part of the “pathway of married people.” But the vocalization
of the men in making libations does complement the women’s singing.1
These fundamental differences between Andean and universal mathemat-
ics are reinforced in the terminology and practice of counting (wakhuña) that is
again subordinated to the activity in question and not directed toward universal
mathematical values as it is in the West. The same idea underlies Don Domingo’s
comparison between different classes of counting (wakhu wakhu wakhu), and the
guiding ways of textual practices, such as the multiple pathways (thakhi), or rib-
bons (sinta) in the men’s libations and the women’s songs to the animals.
However, it still remains to be seen how the terminology and practice of
counting might provide evidence for a multiplication stage in the past, which only
occurred after fulfilling a primary activity, that of capturing the cerebral strength
of the enemy.

E veryday C ounting and R itual C ounting

The conceptual difference between “making knots” and “not making knots”
seems to derive from determined moments in a cycle of warfare. In the first stage,
you make knots to tie up an enemy, whereas in the second you release his ener-
gies by “not making knots,” or “by untying them,” or “vocalizing” them, in order
to generate the new harvest of offspring.
This differentiation between “knotting” (chinuntaña) and not knotting gives
rise in turn to distinct sets of practices. The first is a daily recounting, whereby
threads are knotted as part of the numerical practice of immobilizing the ener-
gies of your property (like booty), and tying in luck there, before going on to vo-
calize the quantities in question as a multiplying act. The second is a more ritual-
ized recounting, when the fixed mental memories in knotted thread are vocalized
in a more predetermined way with the aim of releasing (untying) their contents
and directing the flow of luck in a desired manner toward certain material ends.
Although both practices are founded in the same logic, they are applied dif-
ferently. The simple daily counting is extended to the annual counting of family
and ayllu production through inventories of its elements (human, animal, and
vegetable) as information passed on to the higher ranking ayllu authorities, and
hence to the pertinent external authorities of the state. In contrast, ritual and
magical recounting has to do with the desire of owners, vocalized by appealing
to memory and the continuing jurisdiction of past authority systems. Let us ex-
amine the everyday practices first.
228 A N D E A N T E X T S A N D T H E I R I N T E R P R E TAT I O N

T he D aily C ounting of the A nimals

From Don Domingo’s description, the daily counting of the animals gener-
ates a reproductive power that does not function according to a system of abstract
numbers as in the modern West, but rather according to a system of sounds and
colors. That is to say, the quantity achieves power through vocalization and tones.
The ideal person to count the animals in their pen in his valley community is the
owner’s son-in-law (tullqa).
This system of counting controls any loss of the animals: they should be
counted every three days “to assure that foxes and other predators do not take an
animal.” They are counted in pairs, adults first and then the offspring, or they are
counted “in families.”2 They are always counted in their pens, because “if they
are walking about, it would be impossible to count them.” The work is made eas-
ier by the fact that “sheep know how to count among themselves.” They group
themselves in families in the pen and they count among themselves by the sound
of their bleating (somewhat like pupils in the classroom). As Don Domingo com-
First the mother sheep is found, and it’s counted. “Baa,” she says acknowledging the el-
der daughter, and she also says “baa” to acknowledge the younger daughter, and if there
is one missing, perhaps you don’t hear one of these, she just goes round and round. But
the mother sheep searches for them, finds them, and says “baa.” That’s how their owner
knows. You can’t know by counting because they are too many........That’s how shepherds
count........When they [the sheep] finish counting, only then does the owner count.

Some sheep give birth in the pen by night, and the extra lambs in the morning
are called jumintu (augmented); these must be counted separately. Don Domingo
considers that if the rutiyu (pen) is well toasted, and “lucky,” then you will find
more young there in the mornings.
The animals are counted periodically by the technique called chinu wakhu
(counting knot), the name for the pair of stones found in each corner of the pen.
These small stones, one black and the other white, are square shaped, but their
surfaces, called “their knots” (chinpa), are “rounded like grains.” These stones are
compared with the classroom blackboard. The whole set of stones called “coun-
cil knot” (kawiltu chinu) are put there by the animal owners, husband and wife; if
just the husband were to put them there, then “the offspring would all be males.”
First the adults are counted, males and females, and then the young, as jumintu.
On counting the young, they say: chiku chinu, chika chinu (a gendered play be-
tween the Spanish for “boys’ knot and girls’ knot,” nudo de chicos, nudo de chicas,
and the Aymara for “half and half,” chika chika). The total of young is put to the

side of that of the adults, and after counting you waft the smoke from burnt bit-
termint (peperina, Minthostachys mollis) over them.
This periodic counting is done with some round chalk (called lapisiru, “pro-
pelling pencil”) the opposite color to that of the stone: so the black stone is
marked with white chalk, and the white stone with charcoal. The white stone is
associated with males, and the black one with females. This action is much more
than a simple arithmetical operation, as the chalk is considered to have an “eter-
nal” quality, and “not even the rain can wet it.” In addition, for the shepherd or
shepherdess, the combination of colors in counting results in a “mass of knots”
with the potential to reproduce young of the same colors: “This black stone is
tied with the white, and the white with the black; doing the reverse is not worth
anything. That’s why some offspring come out white and others black.” Follow-
ing the same technique of guide (ira) and follower (arka), one person counts and
another notes or “writes” (qillqaña) the amount on the stone “in the form of rib-
bons.” In this case, the amount is called chinu, “knot.”
According to Don Domingo, when the chosen person writes on the stone
in this way, “we say that the sheep are learning to read.” For him, this whole
process of “learning to read” through writing sums on the stone/blackboard has
a creative and reproductive aspect, in that “those who don’t know how to read
well only have one offspring, they don’t procreate, while the sheep who know
how to read well, have up to ten offspring, and we say that these “know how to
read well.” This is why Don Domingo compares the everyday dialogue between
young shepherds and the sheep, of “bleating” (phawiña) among themselves, with
the form of teaching between teacher and pupils at school:
The children ask the sheep, saying “Baa, ii, aa.” They listen attentively.
“Baa baaa, baaaa.”
“Right, right, that’s it,” and now the sheep “already knows how to read,” the children
have already asked them. They say that on each vocalization [sapa lakaru] the sheep know
how to read. “Baa.” And it replies, “Baaa ii.” On each vocalization it says “baa, baa, ii,” and
the male finishes it off in this same way.

Even more specifically:

The boys say to the sheep: “Herd them up, a, b, c, that’s it,” they tell them. And the sheep
usually stand there quietly. Then in another moment they reply:
“A, b, c.”
“Now you’ve learned, so you’ll procreate now,” they tell them.

These examples warrant comment. A first point is that the sum marked on
the stone “blackboard” of the animal pen has a quality of “writing,” considered
230 A N D E A N T E X T S A N D T H E I R I N T E R P R E TAT I O N

comparable with a letter (or letters) marked on a classroom blackboard. A second

point is that the act of “making each vocalization” has more importance than
simply looking at the sum written on the stone/blackboard, as the sheep only
“learn to read” on hearing these bleating vocalizations and repeating them. As
Don Domingo says: “some cannot bleat well. These have only one or two off-
spring, and then they are old.” In this sense, everyday numerical practices coin-
cide with classroom recitative practices, in which the recitative mechanism of vo-
calizing “from memory” constitutes a generative act in itself, rather like reading
with comprehension in a European sense.
A third aspect has to do with the dialogical self-ordering of the animals ac-
cording to their positions in the flock. The female that always leads the flock is
called puntira (guide), and the males and other animals follow her. This guide
communicates with the other animals by her bleating, and communicates with
the male in the same way. She says “baa” and the male responds “baa,” and then
“all the rest hear and she guides them anew.” Don Domingo compares this every-
day dialogue between the different members of the flock with the dialogue be-
tween a husband and wife, or teacher and pupils. One of them guides (irpiri) and
the others follow (arkiri).
Another crucial aspect concerns the relationship between the sound of bleat-
ing and the successful multiplication of the animals. For Don Domingo, the bleat-
ing of the animals, pawaña, is specifically related to the dispersion of its drop-
pings, phawaña. As he says, “it is converted into its droppings.” Here, similarly
sounding verbs have mutually multiplying qualities: pawaña means bleating, pha-
waña (with an aspirated p) refers to the more general action of throwing small
things, such as grains (broadcast) or droppings, while another similarly sounding
verb pawiña, means “to wind yarn, passing it onto the hand, and from the hand to
the arm” (according to Bertonio’s Vocabulario, [1612] 1984, 804).
What precisely is the relationship between the bleating sound of the animals
(pawaña), their multiplication, and the dispersion of their droppings (phawaña)?
This seems to concern a semantic domain relating the directed dispersion of ele-
ments of sound with the desired dissemination of other small things, such as the
proliferation of droppings in the pens and of young animals in the mornings.
One factor in common is the fertilizing act of vocalization, or any other sound
(the bleating), in the multiplying context of the pen.3 Another is a logic of mul-
tiplication that depends initially on the unraveling of something (a sound, food)
into its constituent parts that then multiply. These factors imply a vital nexus be-
tween number and vocalization that functions through an ordered code of marks,
sounds, and colors; that is to say, between number and writing.

T he R itual C ounting of the A nimals

The ritual counting of the animals has the even more determined goal of
“procreating” young. In this context, the same son-in-law is obliged to count the
animals of his father-in-law, and through his efforts, assure the procreation of his
In Don Domingo’s community, on the Saturday of Carnival and other ritual
occasions, the son-in-law must first toast the lapisiru (chalk “propelling pencil”)
used to count the animals ordinarily, and then for the luck of the animals in their
reproduction. He must count “solely from memory,” “augmenting” the sum to
assure the reproduction of the flock the following year, taking into account its in-
crease in the previous year. This form of counting ritually, called “turning knot”
(rutiy chinu), is directed to the masculine guardian mountain (uywiri) that “sur-
rounds” and tends the animals in a particular pen. In this sense, ritual counting is
located in the wider context of gendered sacred sites paired as male and female,
in memory of their part in the former state hierarchy in which the Inka and Coya
were the highest authorities.
In this case, the son-in-law can count one by one, but as his goal is to “reach a
high quantity,” he counts gradually until one hundred, and then says p'aqx. Then
he reaches another hundred, p'aqx, and another hundred, p'aqx. In this case, the
sound p'aqx, or its variant ch'atax, is a memory of the archaic way of counting,
often attributed to the Pukina language, although its power here derives from
the fact that “the Inkas spoke so; ch'atax ch'atax were two hundred.”4 This echoes
the manner of counting in the Tales of the Damned One (Condenado), when the
body parts he throws after eating their meat are counted up, one by one, and last
of all, the head! It is as if, in arriving at high numbers, you were nearing the Inka
and the domain of body parts owed to him in war.
The son-in-law counts up to three hundred animals, and then doubles this
sum up to six hundred, using the animal droppings (compared to little sweets);
picking up three units of droppings represents the three hundred animals and six
for the six hundred. Each dropping represents one hundred animals and the vocal
reduplication of this number in the ritual expresses the desire on the part of the
owner to have this quantity of young by the end of the year. If he (or she) wants
just fifty young, then you take away half of the dropping to represent the fifty
one half dropping = fifty animals one dropping = one hundred animals
three droppings = three hundred animals six droppings = six hundred animals
232 A N D E A N T E X T S A N D T H E I R I N T E R P R E TAT I O N

As Don Domingo expresses it: “It is as if the dropping were a sheep, and rep-
resented 100 of them.......so if there are 150, you take away half of the second
dropping, and it’s 150; to this you then add another 150, which you count out af-
terward, and this will come to be by the end of the year.”
Alternatively, when you go beyond a hundred, you lift up a dropping of an
adult animal to represent the hundred, and then you add some droppings from
the young to represent the remaining young that are “augmenting.”
one adult dropping = one hundred adults one lamb dropping = one extra lamb

Then the little droppings of the young are hurled into the pen, together with the
same number of coca leaves, “as if the animals were going in there, and you say
jach'a jirawa (“place of many droppings”) and “next year I’ll count even more.”
In these ritual contexts, the son-in-law does the counting “only by mouth,” to
“complete the reproduction of the flocks.” After counting, the female owner (in
this case the mother-in-law) offers another toast “for chiku, chika,” and then the
son-in-law “walks to and fro in the pen’s entrance; he is not standing still in one
place.” This is to know “that his steps are good,” as a way of assuring good results
in the ritual, as if the augmenting in this case came from a certain dynamics, fo-
cused on the son-in-law’s voice.
The important role of the son-in-law in the reproduction of the flocks is ex-
pressed in his ritual name of “sitting condor” (kuntur chuku) or “rich lord” (mallku
qapaxa), nicknames that highlight his ties with the Inka and his warrior aspect,
as a vital in-marrying husband, to augment the power of the wife’s own group.
When the animals are counted in this way, the sum is augmented, the account
is closed, and then the son-in-law leaves. In this case, the warrior son-in-law, as
the Inka’s modern-day representative, first vocalizes the number of animals, then
augments this quantity through his desire, and then ties a knot ritually, resulting
in the multiplication of the animals in their generative pens (sources of the piles
of droppings and young), helped by the bleating of the animals as an expression
of their wise management.

L uck and an A ndean T e x tual T heory of C ounting

In these two examples, the relation between vocalizations from memory and
numerical practices derives from the dynamics of kipu-based writing, as the per-
tinent textual basis for later vocalizations. This relationship is even more explicit
in the annual counting of the “libations of luck,” when the vocal transmission of
the basic and larger units of a kipu form part of a wider processing of informa-
tion. In this case, the stages of processing information derive from the vital nexus

between textual practices and theories of learning, in the framework of an An-

dean textual theory.5
The annual recounting (census) of the animals takes place in the context of
the marking ceremony, in the formal series of libations, or ch'alla, when the “liba-
tion knots” (ch'alla chinu) and “libations for luck” (surti ch'alla) guide a ritual per-
formance directed toward the desired reproduction of the flocks. During this rit-
ual, knots are tied symbolically in the mind, all in sequence. This sequence traces
the textual relations between the totality and its parts.
Before describing the ritual, let us consider briefly the idea of luck (surti) in
this regional setting. “Luck” in its Andean interpretation describes a facet of the
dynamic processes concerned with stimulating through ritual the potential for fe-
cundity and regeneration. According to Don Domingo, luck is taken up “by tying
the knots” (chinuqaña) and is received into the heart. The same act of recording
the libations also “makes the person think,” as the knot itself “has the capacity to
This Andean interpretation must have emerged from colonial uses of the
term suerte, “luck,” as in “principal luck” (suerte principal) to refer to the use of
gains (interest) on a capital sum without touching the capital itself.6 However,
in the context of numerical practices and their origins, this Andean interpreta-
tion of “luck” also draws on the potent logic of sacrifice. More specifically, “luck”
designates an aspect of the fertilizing forces incarnate in the captured enemy that
has to be released. In releasing this “luck,” the captured thoughts in the head of
the Other are thereby transformed for one’s own benefit.
Cycles of luck, and the sacrifices associated with them, are performed in a
constellation of local and state-controlled ritual sites. So, in spite of the colonial
origins of the term surti, the practice of making libations for luck in the mind of
older comunarios harks back to a more distant past, to the question of rights to the
Inka’s lands. From Don Domingo’s point of view, the custom of making libations
derives from the time of the Inkas and the chullpas, long before the Conquest,
when kipus were called sinta phara phara (“dry ribbons,” a name that seems to
emphasize the dry and dead condition of the threads). For him, the great regional
authorities or mallku left this custom as an inheritance, and “Tata Riy and Mama
Riya ordered them to carry out this custom,” teaching them how to do it. Now,
in place of the Inkas, “only the mountains are remembered.”
Like the comunarios of Qaqachaka, Don Domingo is very conscious of the
function of reciting libations, not only in exercising the memory and recording
ancestral customs, but also to “make luck” and direct it to the food crops, help-
ing them grow in the ayllu lands. He insists that the pathway of the ancestors
should be followed (arkaña) and “not forgotten, as it is by current day youth,” as
234 A N D E A N T E X T S A N D T H E I R I N T E R P R E TAT I O N

the libations go together with production. By naming the products, the land is re-
membered, and by naming the animals, their pasture is remembered. “This is the
reason why young people don’t cultivate the crops well, and have to buy them
instead and bring them long distances.” In his way of thinking, spoken libations
help production by acting as a “weft” thread that directs the motion of the food
crops as they grow in the furrows.
In this sense, Don Domingo compares the generation of production by vo-
calizing the libations of luck (as a kind of text) with what children do in school
with reading and writing. Just as the animals work the land, so the rows of let-
ters in the children’s notebooks are “lists” or “furrows” that have to be worked. In
plough teams, the bull that works the land is ideally of a mixed or speckled color,
while the white bull is the land itself and a black bull the ploughshare. At school,
the black bull is the black pencil that writes in the furrows of the notebooks and
the white bull the paper. “They are couples, and they couple together,” he said.
The tiny letters are also “speckled,” because they are “of all kinds,” and the black
ink on the page is “the sweat” from physical work.
His comments underline why local education insists in the similarities be-
tween vocalized libations and the good management of the ayllu lands. They also
explain why older comunarios in rural communities grumble that schooling fo-
ments the forgetting of ancestral customs as it ignores the vital learning of the
numerical and mnemonic practices of libation making. They accept the school-
room practices of teaching reading and writing up to a point, as long as they can
continue with their custom of making libations at feasts and assure luck for the
community. But now with modernization (and the ever greater influence of a
new wave of evangelist sects) a watershed has been reached. Young people “nei-
ther learn to read and write well at school, nor do they learn to exercise well the
memory through libation making.” So they fail to direct their futures into the
production of the place.

T he L ibations at the A nimal - C ounting C eremony

In the animal-marking ceremony, the preparations before the libations begin

with the sacrifice of a white sheep (uxaruru, a virgin), which has to be eaten with-
out any salt. Various ingredients are offered on an altar, and then libations are
made all night long with alcohol and corn beer. They proclaim as they do so,
“Now let’s stay up until daybreak,” as the participants should not go to sleep dur-
ing the ritual.
The sequence of libations begins the following day. Each libation in northern
Potosí, as in Oruro, is made in a round in which the dispenser serves a cup to all

those present and each person says in turn “for such and such” (ukataki) generat-
ing the characteristic spondaic rhythm. After finishing a round of libations, a knot
is made, and with this “the luck of the animal is tied.”
The order of the libations traces the logical steps of an Andean art of mem-
ory, based in the textuality and ordering of past kipus. At present, these process-
ing operations, now uprooted from their proper textual basis, only act as the or-
ganizing basis for cognitive steps at a mental level, but Derrida would argue that
an institutionalized process of writing-like inscription still takes place. This is evi-
dent in the fact that these same ordering processes can be found in a whole gamut
of different textual practices. For example, as in the songs to the animals, the or-
der of libations begins with the names of the larger animals (beginning with the
cow) and those of the small animals follow. As always, the llama “wins out over
all the other animals” in its sheer quantity of libations, a direct consequence of its
quantity of colors. The toasts to the leading male llama (called tilantiru) are par-
ticularly important. They have to be tied well, as “they hold all the rest.”
Let us detail two sequences of libations to the animals and then analyze their
Sequence of libations to the cow (waka):
Iskin turu (offering place). An introductory theme.
niñu, niña (the strength of the animals)
Tata San Marcos, San Isidro
Capitán Awki
yunta, jila sullka yunta (greater and lesser plough teams)
Tayka, machaq qallu, nasas (nayra q'ara) (Mother, new offspring, and their drink-
ing names)
wakhullu (its sleeping place, ikiñpa)
mulluyu, rika jurtuna (rich fortune, sometimes called jach'a jira)
mulluy qupuyu (the place in the pen where something of its presence remains,
e.g., its hairs).
Sequence of libations to the llama:
Tata Päri Antuñu, Tata San Antuñu (considered to be brothers)
tilantiru and sariyu (leading males and stud males)
jumintu (augmenting young)
chuchikillu and wanakillu (a single knot)
walthañpa (its wallowing place, also called mult'isi or qhayana)
uypa (its pen)
236 A N D E A N T E X T S A N D T H E I R I N T E R P R E TAT I O N

iskina (its corner)

Pilu mayku, Pilu t'alla (Lord and Lady of Hairs)
kumpirisa (its droppings)
mulluy qupuyu (the sleeping place called “golden turning, silver turning”)
jatiyu (its load)
allqamari (the rope used to tie on its load)
wantirillu (a little flag placed on top of the load)
sinsiru (the little bell worn on its neck)

L ibation U nits

In any sequence of libations, the person who counts (whether this is the
owner of the animals or a person selected to serve food and drink to all the par-
ticipants in a ritual) is guided by a precise and predetermined order. After the for-
mal openings that name the house and its various ritual sites, the sequence for
each animal continues “one upon the other.” Don Domingo commented (using a
textile analogy) how “you have to throw them completely” like a ball of wool.
The sequence usually begins with the name of the animal’s patron saint.
There follow the names of the different kinds of animal, according to sex (male
and female), age (old and young), color (chuchikillu, wanakillu), and position in
the flock (irpiri, puntachiri, and so on). Then come different names for the ani-
mal pen, followed by the names of certain effects of the animal that character-
ize its “tracks” (giving the place its scent and presence), for example the elements
(hairs and droppings) left by the animal in its pen or wallowing place. For Don
Domingo, the ritual term for the effects in the pen, mulluyu qupuyu, alludes to its
wealth and serves to call the animal “as it sniffs there,” and then binds it to the
place. These places, known to the animal, “run in families among the llamas, so
they are known to the father, sister, and brother.” In addition, the animal wallow-
ing places serve to unite the earthly sites of the llamas with their celestial coun-
terparts in the “black lakes” of the Milky Way.7
At the termination of each libation sequence, the sequence is sometimes
formally rounded off, by saying chinu mayku, chinu t'alla (“Lord and Lady of the
Knots”), with the objective of “tying luck there,” and “grasping” it in the heart.
Alternatively, the sequence is closed with an even more exaggerated rounding off
(muruq'tata) that serves to end the toasts (ch'alla tukuñay). Taken as a whole, the
order is always a variant on the following basic order:

The animal patron saint

The different classes of animal: by sex, age, color, and position in the flock
The animal pen or wallowing place, where it leaves its scent
Lord of the Knot, Lady of the Knot, and the final rounding off

Likewise, in the case of the food crops, the libations sequence is rounded off with
some elements that “grasp” the produce, whether this is the “deposit of three
seams,” in the case of corn, or “its sleeping quarters” in the case of potatoes.
This whole sequence has different levels of meaning. In general, the pathway
(thakhi) of each animal (or food crop) takes the form of a ribbon that guides the
calling of the animals, especially when their distinctive smell is named (as the
hairs and droppings left in the pen), so as to bring luck and grasp it in its place. In
the case of the llama, the key focus on the male lead llama aims to bring luck in
the “silver” that follows behind him. Other subordinate libations that help to at-
tract luck name elements concerning the production and reproduction of money
(shining/fluttering butterfly, bank, provider, and so on).
In each case, after making the last knot in the sequence, and “finishing it off
so” (jiksuña), the last part of the thread is left loose. For Don Domingo, this is
its “pasture”; the loose end leaves the animal free to wander and maintain itself
drinking in its watering places (awara) and grazing on its pastures (pastu). “On the
contrary, if we were to tie [a knot] here, it would be as if it were in prison, that’s
how it would be.”
In its totality, the sequence of toasts trace the descent of each animal, begin-
ning with the ancestral mummy saint that often has its counterpart in a black
spot, or “black lake,” of the Great Ribbon (the Milky Way). These protect the ce-
lestial and earthly flocks as entire groups. As Don Domingo affirms, by naming
Tata San Antuñu, the patron saint of the llamas, “it is naming all the llamas as a
group.” Then you proceed to the earthly flocks, by naming individually the differ-
ent animals that constitute the flock. Then you name the pen and the elements
it contains, as the destination of all the different pathways of the sequence (both
earthly and celestial). Finally, the sequence is closed by naming other protectors
of the pathways, the Lord and Lady of the Knots (chinu mayku and chinu t'alla),
with an especially elaborate rounding off (muruq'tata), but leaving the end of the
thread open (and so the destination of the animals).

T o wa r d a M at h e m at i c s I n c a r n at e

This whole sequence of libations has an analytical dimension. The elemen-

tary particles, as the letters and numbers of the kipu text, begin with the se-
238 A N D E A N T E X T S A N D T H E I R I N T E R P R E TAT I O N

Heading close to Thread Knot

the principal cord


THE PATHWAY of animals

going in file
Naming EACH
of the pathway
of the PEN



Fig. 27. A libation sequence on threads and knots.

quence heading that names the ancestral mummy-saints of the animals closest
to the principal cord. Then, the sequence of each pendant thread alternates be-
tween the general “thread-pathway” (thakhi, sinta) of the different animals going
in file, and the more specific “knot-pens.” So you always pass from a funicular
(threadlike) element to a more knotlike one, and a homologous configuration is
reiterated by naming the funicular hairs and the germinal droppings (as the per-
sonal elements) and the protectors of the thread pathway (mallku and t'alla) be-
fore completing the sequence with a rounding off (muruq'tata) and the loose end-
ing (pastures and water). See figure 27.
In mathematical terms, first the animals are denominated as a group (by
naming their saint), then the count passes on to the units (by naming each ani-
mal), and then to the constituent elements of the units (hairs and droppings). In
the final loose ending, the elements of the parts are denominated (the pastures
and waters on which the animals originally grazed, so as to incorporate these
nutritive elements into the development of their long fleeces and rich meat, and
disperse the remains in the droppings). So the order of the sequence goes from
the totality to the units, then to their constituent parts, and finally to the parts of
each component. It is worth noting that the sequence of libations replicates the
ontological pathway of creation of the flock, from the ancestral celestial flocks to
the earthly ones, and gradually proceeds to the origins of its very elements in the
elementary particles of ayllu pastures and waters from the surrounding moun-
tains. So it proceeds from their celestial creation to their material existence in
ayllu territory (and the elemental resources of its lands and pasturing places).

In terms of corporeality, libation dynamics derives its power from kipu struc-
tures in which male elements (seed) are combined with female elements (blood)
in a fertilizing mixture, where spirit (ispiritu) generates new life. Similarly, as you
“tie luck” in the kipu knots, you grasp the same information in the memory, and
the knot like a ball of wool which is the heart.
This Andean form of analysis does not depend on alphabetic writing or math-
ematical formulae as in the West. Rather the systematic interaction between
the human voice and the kipu threads and knots generates an open kipu text,
through which libations can serve as a medium of analysis, if we take the Greek
term analusis in its literal sense of “releasing,” and in a wider sense as “undoing,”
“determine the elements of,” “decompose into its constituent parts.”
We might compare the configuration of levels and strata and the unraveling
of ideas in space during Andean libation making with the deconstructivist ten-
dency of the French philosopher Jacques Derrida (1976, 88) as a stage prior to the
act of reconstruction. At a theoretical level, this would be a philosophical reading
of the process of undoing the forces of the enemy in order to convert them into
one’s own, a logic of sacrifice in which the deconstruction of death is what gen-
erates the necessary elements for initiating another cycle of construction.
Finally, the attention given to the direction of spin of the wool, and the for-
mation of knots (especially the exaggerated knot muruq'tata), imply that kipu
reading might record a dynamic process marking the termination of one an-
nual life cycle and the initiation of another. In this sense, kipu structures might
deal with systemic processes rather like the “dissipative structures” described by
Prigogine and Stengers (1984), which often have the same spiral configuration. In
the state of flow (thermodynamics) of these dynamic systems, the whole system
can only continue if the less useful forms of energy are dissipated (by augment-
ing entropy), and other more useful forms of energy incorporated.
Put another way, reading kipu knots in libation sequences quantifies the en-
ergies of a dynamic life system through an analytical discourse that is eminently
“spoken,” and understood by all the members of a community present in the rit-
ual. In a manner of self-awareness, it also guarantees the transmission of under-
standing of the metabolism of the system through a kind of “autopoiesis” (from
the Greek roots auto “self,” and poiein “make,” as in “poetry”), in the terms of the
Chilean biologists Maturana and Varela (1981). “Autopoeisis” refers to life’s con-
tinuous production of itself. The importance of this sense of self-awareness for
concentrating the order of things was something that Don Santiago of Livichuco
intuited in his preoccupation with the very survival of this practice.
240 A N D E A N T E X T S A N D T H E I R I N T E R P R E TAT I O N

T he P ath w ay of L uck

We should not pass over the fact that the dynamics of the pathway of luck
derives from kipu textuality, and that the point of reference for the flow of all li-
bations is a hollow yarn. It is as if the hybrid pathway enacted by the comunarios
with reference to the kipu structure of knots and threads lubricates the possibility
of ayllu production and reproduction, in a similar way to their prayers for rain. At
the same time, the textile and farming language of libations, in the hands of men,
reinforces the weave with the seeds they are sowing in their fields.
This perception of the kipu has a long trajectory that dates back much fur-
ther than the time of the Inka to the mythic world of Andean tales. Once, on see-
ing a photo of an Inka kipu with a wooden handle, engraved with two figures (a
man grasping a drinking cup to one side and a water bird at the other; see figure
28), Don Domingo reflected on how the kipu thread entering the drinking cup
was hollow (phusa) “like a leaf of grass,” in which the sap rose and fell, acting as
the water reservoir of the animals, plants, and people (similar to the tuft of wool
in the Inka rituals mentioned at the beginning of this chapter). He compared
the extended hollow thread with the beak of the hummingbird (q'intiluli). This
meant that the kipukamayuq could “drink water from far away,” as he “sucks and
the water just comes out and then enters the mouth,” the same as the humming-
bird that “sucks nectar from the flowers from far away.” For Don Domingo, the
warrior hummingbird is “he who ordains,” since he brought the woman down
from the condor’s nest (another reference to the dynamics of ordering threads
as if they depended on the movement of women in marriage). It was perhaps
this kind of kipu textuality, incarnating the liquid medium of communication
par excellence, which permitted the kipukamayuq, through the flow of liquid in
his body, to communicate from the center of the Inka state to any corner of the
The dynamics of this liquid medium of communication also reiterates our
previous comparison between the kipu knots and the complex of “head-babies.”
Remember that when a sequence of luck is carried out for the food crops or the
animals, you must tie a knot (mentally) and in this way “tie luck” (surt chinu-
ntaña). This is done “just by drinking,” and as a result “it’s there that the head is
given” (ukan p'iq churtnaxa), in the sense of giving ideas. At the level of the indi-
vidual and the home, it is as if, in the past, knotting the sequence of libations gave
the necessary impulse (or luck) to redirect the energies (ispiritu) of the captured
head, and so put the head taker (and his family) on top mentally, able to fulfill his
own productive process, and so proliferate his own offspring.
At a communal level, the same configuration of ritual practices would give a
ig. 28. An Inka kipu with a wooden handle (from Ascher and Ascher 1981, 91).
242 A N D E A N T E X T S A N D T H E I R I N T E R P R E TAT I O N

similar impulse to all the local resources, given that participation in libations for
good luck helps generate the cosmological circulation of all the elements in the
ayllu (waters, pastures, animals, and so on). In this context, communal toasting by
sprinkling alcohol helps “concentrate in quantity” the liquid from the libations.
To reinforce this possibility, each participant must begin and end the cycle of luck,
mentioning the archetypical personages of the food product (the various mama-
las) or the patron saints of the animal that have the power to gather together
(concentrate) all the elements under their charge. Similarly, the last toast in the
sequence of luck, which has the function in this hydraulic dynamics of concen-
trating all the liquid from the libations, is enacted in a homologous way, by tying
(mentally) the final knot. As Don Domingo comments, to make the knot at this
moment is “as if you were giving heads, and heads, and heads [ukat uka p'iqiñchxi,
p'iqinchxi, cabezani].” As a result, there are more babies/products (wawa) in the
family storerooms.

G u a r d i n g t h e K i p u a n d t h e P r o l i f e r at i o n o f H e a d s

The symbolism of heads is alluded to again at the end of the pathway of luck,
when the kipu is finally well guarded in a little box (kajita), rather like that used
for the written documents in the custody of a family. The little box is located on
a small board called tinta (“shop”), or alternately tipu (perhaps a play between the
Spanish tipo, “type,” and the Quechua term kipu).
In its rolled up form, a kipu is called a head (p'iqi). It is guarded in the little
box with some of the ropes called allqamari, together with llama fat (qarwa untu)
and a llama fetus (qarwa sullu) “so that it endures and never ends.” Before putting
the kipu in there, the owner has to pass smoking muña or bittermint over it, and
lubricate it with llama fat; if he does not, then it is in vain, and nothing will be
produced, neither food crops nor animals. As Don Domingo affirms, those who
omit this “are not able to produce potatoes or corn; they are wandering about
poor and borrowing things.”
There out of sight, the kipu, as a kind of head, still has the power to regen-
erate things. According to Don Domingo, the very box acts as the “foundation”
(simintchixay) of all the elements guarded there, of which the kipu/head is the re-
generative trigger. So, there is a final toast before introducing the threads inside,
beseeching them to produce offspring of various kinds: “All kinds, good kinds,
kinds of coffee-colored hair (corn), kinds with twelve eyes (potatoes), kinds of
beans, kinds of peas, turning kinds, a delicious hour, a good hour, now it’s lucky,
hit the mark.” The kipu, having become like a procreative “head” through the
process of “making luck” (focusing on the knots), now has the capacity to “give

heads” to the babies/products (wawas). “Having a head” in this context describes

the growing condition of the food crops, before the full production of the har-
vest. Don Domingo explained that all the products “have heads”: “they have
hair,” and so “they look like heads.” Potatoes have heads (from where the points
of growth sprout), corn has a head (from where its long hair grows) and wheat
has a head (in the ears with their hairy “beards”). For Don Domingo, then, tying
knots in the kipu threads “is as if they were tying the hairs from a head, but with
llama yarns.” Having a head is “lucky,” it produces ideas and generates another
cycle of production. Even the babies/products in turn produce heads that sprout
hair ready for the harvest.
These final comments clarify other aspects of the language of the knots. Each
knot, as a head/baby, has the fertilizing power of proliferating the words of the
libations, all according to a certain logical order and a predetermined pathway of
memory. So vocalizing the libation is to participate in a sort of “vocal writing,” as
an actor who plays his (or her) determined role. The words are learned in the fies-
tas, on entering as feast sponsors, and the words are repeated on seeing the actual
knots or recording the mnemonic knots.
This reasoning explains why the community members still prefer the elegant
simplicity of an Andean deconstructive logic instead of the “confusion of letters”
of alphabetic writing. In the former, the Other is destroyed to reconstitute the
Self; in the latter, you attempt to construct the Self, but the same process results
in confusion.
We should not forget that these numerical practices concerning the prolif-
eration of babies and heads derives from regional notions of corporeality, and
functions around the flow of an original liquid element between neck and head,
in order to achieve the luck of the sequence. Could this flowing element be
the breath-and-blood mixture of Andean corporeality that rises from the heart
through the “great vein” toward the mouth, to then become externalized with
the voice? And could it be that, in some sense, we are still listening in the libations
to sonorous echoes of the Inka’s voice?

T e x t ua l Log ic i n t he A ndes

The conservation of certain technical features is probably related to the existence of a

symbolic substratum that the Spanish were far from imagining........None of them thought
that a few technical elements would have an ideological weight of their own and play a
subversive role in this way.

Sophie Desrosiers, “Lógicas textiles y lógicas culturales”

Let us return now to the question of Andean weaving as writing, in the context
of regional ideas about metaphysics and the ontology of being. Our challenge
is how we might add, from a perspective centered in Andean texts, to the Euro-
pean debate inspired by Derrida about the status in theory and practice of dis-
tinct definitions of writing, and the more Amerindian-centered debate, inspired
by the Brazilian anthropologist Eduardo Viveiros de Castro, concerning ontologi-
cal depredation and its cultural consequences.
We saw how, faced with the history of European texts, Derrida’s deconstruc-
tivist stance takes as its point of departure the articulation between the Self and
the Other as mediated by the practices of alphabetic writing. His position is based
on the fundamental role of the voice (which he argues has not been recognized)
in the origins of European writing practices since the ancient Greeks. Taking into
account these origins, he redefines the nature of writing as such, and recognizes
in the different writings of the world (drawings on ceramics, rocks, glyphs) their
own form of grammar (gramma), based on the use of the voice and the rhythm
of the breath.


Yet, from the point of view of Andean rural communities, there are certain
inconsistencies in his argument. For example, Derrida deals with each culture’s
texts as if they were invariant, but we have already seen how a gendered division
of labor in the Andes affects teaching and learning, construction and organiza-
tion, theory and practice. Using the regional textual theory developed so far, we
explore the dialectics of Andean knowledge according to gender, and the way
this shapes notions of text, voice, and corporeality, the Self (naya) and the Other
(jupa). Then we reply to the question about the validity of Derrida’s grammatol-
ogy, or Viveiros de Castro’s notion of ontological depredation, from an Andean

T e x t i l e L o g i c a n d C u lt u r a l L o g i c

In her essay of 1997, Sophie Desrosiers identifies a series of correspondences

between textile and cultural practice that she calls “textile logic and cultural
logic.” This means that gendered differences in ways of learning, the corporeal
meaning of different garments, and approaches toward the relation of writing to
weaving, are reinforced at the level of textile structure. According to Desrosiers,
these are founded in two critical aspects: the direction of the threads when they
are worn (whether they are predominantly horizontal or vertical), and the treat-
ment of borders (whether they are closed or open).
In our previous studies of the textile logic of the mantle (a feminine garment)
and poncho (a masculine garment) in Qaqachaka, we noted a difference between
the horizontal and more figurative designs of female ties and the vertical and
more abstract striped designs of male ones. The corporeality of the cultural logic
at play here has to do with the way that horizontal female ties trace connections
of maternal blood (wila) between social groups, whereas male ties trace connec-
tions of male seed (muju), or of masculine blood in warfare.1
The textile structures that correspond to this cultural logic are twofold. Ver-
tical warps are homologous with the long-standing patrilineal groupings located
in one particular hamlet through the generations, with their common surnames
(Choque, Quispe, Condori) derived from a putative ancestor in common. Even
in the garments that women weave for men, such as the poncho, they express
through the directionality of their warp-faced stripes in use (hanging downward),
the vertical nature of masculine relations, whereby members of a paternal line
always stay in one place.2 In contrast, the horizontal wefts that interlace between
the vertical warps are homologous with the crisscrossing web of female ties (es-
pecially after women marry) between paternal lineages, which then stretch across
the ayllu territory and beyond, through the generations.
246 A N D E A N T E X T S A N D T H E I R I N T E R P R E TAT I O N

The constant female interweaving of consanguine and affinal ties between

ayllu members gives rise to a living web of interrelations, a textile “discourse”
that comments on the modes of production and reproduction within, and be-
yond, the ayllu limits, all on the basis of fleece.
Textile structuring in which vertical warplike elements predominate can be
found in other items woven by men, for example in the rope work of slings or
kipus. In the case of a sling, its “holder” loop at one end and loose strands at the
other are homologous with the pendant thread of a kipu, which has a similar
open loop at one end where it is attached to the principal cord, and loose strands
at the other.
This directional logic in contemporary textiles from Qaqachaka is homolo-
gous with that observed by Desrosiers in pre-Columbian weavings. In many an-
cient textiles, the principal items of masculine attire are characterized by the “ver-
tical” direction of the threads that form the adornment (in the highlands), or the
openings for the head and arms (on the coast). The counterpart female garments
have the same features, but set horizontally (1997, 330).
For Desrosiers, this regional textile logic characterizes both modern and ar-
cheological weavings. In addition, “it puts in parallel the vertical/horizontal op-
position of various features related to textile techniques, with two schemes of
organization in Andean societies: between highlands and coast, on the one hand,
and the contrast between masculine and feminine, on the other”:
highlands : coast :: hanan : hurin
hanan : hurin :: masculine : feminine (ibid., 331).

In the Southern Andes, these cultural and gendered differences are also present in
the logic of textile borders. In their handling, open borders concern the mascu-
line and closed borders the feminine, another feature that is confirmed in a num-
ber of contemporary studies in the region.

F emale C losed T e x tile B orders

Women’s garments are always woven with four closed borders, and some-
times with additional edgings, a characteristic feature used as much in garments
of the past as in the present. Textile and cultural logic demands that female gar-
ments “are never cut,” but rather worn just as they are when the weaving is
finished. If a garment of this type is cut, it is as if you were “cutting your own
hands.” Desrosiers cites a weaver from northern Potosí: “Cutting a weaving is
killing it” (1997, 327, 332).
In Qaqachaka, the very terminology of these borders is saturated with mean-

the border called k'illpha

the 2 thick threads

of the pulu

the ch'ukurkata thread that ties the

pulu back to the loom pole, and so
acts as the weaving “foundation.”

Fig. 29. The pulu and k'illpha borders of a woman’s weavings.

ing. In the mantle, the vertical border of the weft (in the first row of warps) is
called in Aymara k'illpha, and the horizontal border of the warps (in the first row
of the weft) is called pulu. (See figure 29.) In practice, the pulu border has the func-
tion of holding the warp threads at the start of weaving, reinforced by the cross-
piece of the loom. When a woman sets up her loom (the horizontal loom is used
at present), she first drives the stakes into the ground, and then locates the cross
poles in their initial position. Then she begins to warp (tilaña) the loom, a task al-
ways carried out with another person. They pass the ball of wool from one to an-
other to lay out the warp threads (chinu) in the form of a figure 8, starting at one
crosspiece and ending at the other. On finishing the warp, two thick threads are
inserted inside close to the crosspiece to form the first pulu border of the warp. A
second pole is placed there, and the pulu (together with the whole set of warps) is
attached to this with another thick thread that acts as holder (called ch'ukurkata,
a name that recurs throughout the present chapter). Then the original pole is
taken out, and the whole pulu complex is readjusted by beating it firmly with a
llama bone (wich'uña). The set of threads beaten in this way (called pulu thawkata)
acts as the “foundation of the weaving.” As Elvira commented, “This is the tight-
ened pulu; only then do you begin.” (See figure 30.)
ig. 30. The pulu border tied by the holding thread (ch'ukurkata) to the loom pole.

As she proceeds, the pulu border marks the termination of the warp at the
two horizontal ends. If the pulu is well finished, then weavers comment that the
weaving is “bordered” and “secured” (sawurkipata) and the threads can’t come
loose. Analogously, food produce can’t get loose from the woven storeroom, nor
from a woman’s belly.
The lengthwise or vertical border, called k'illpha, is formed as the weft ad-
vances in the form of boustrophedon (qipantaña). This border must go “re-
ally straight”; if it curves inwards for having woven too tightly, weavers remark
“k'illphjata,” as the weaving lacks a small piece, like the animals in the marking
ceremony of the same name, k'illpha, which have had their ears notched.
These closed borders communicate information homologous with that com-
municated by the horizontality of feminine garments, namely an “internal dis-
course” for local consumption concerning both fixed wealth (in produce/babies
that the lands generate annually) and mobile wealth (in persons, animals, and
food produce), as well as the expansion of their dominion, but inside the closed
borders of the ayllu. In a feminine, bodily language giving importance to the
menstrual cycle and the gestational process, the aesthetic expression of women’s
weavings, as placental wrappings, expands to embrace the whole ayllu territory.
Various studies on Andean textiles demonstrate that borders and their ico-
nography symbolically define and reinforce the uses of the space inside and out-
side these borders. We already mentioned a study in the community of Chuki-
ñapi (Omasuyos Department, Bolivia) where a weaver compared the activities of
weaving and furrowing the land, and explained how both textiles and fields are
converted into living beings (jaqi) at the moment when their borders are finished
off. On finishing a textile border, she said “it’s like finishing the edge of a field
with a hoe,” and then added: “the boundary of the field has its function in the
production process, because it doesn’t let the food produce out from the field. It
prevents them.”3 Her comments implied that some force inside a field stimulates
the food produce to leave, but that the field borders somehow prevent them.
Exploring the nature of this force, Anne Paul (2000b, 162–63) refers to the
wayñu song and dance performance at the feast of Carnival in Qaqachaka. In a
critical moment, when young people are dancing in a circle with their hands in-
terlaced, the circle is broken deliberately, with the object of forcing the “warrior
spirits” of the dead to return to the “black lakes” of the Milky Way, whence they
came. Significantly, this happens in the place of piled-up stones called taqawa,
where the bones of the ancestors and the enemy dead were kept and displayed in
the past (whether in the ayllu as a whole or the school compound).4 In the wayñu
rainy season song and dance, the circle of interlaced hands has the function of
trapping the devilish forces of the warrior spirits inside the dance until the mo-
250 A N D E A N T E X T S A N D T H E I R I N T E R P R E TAT I O N

ment when their energies are spent, when the ring of braided hands as an exter-
nal boundary is broken on purpose. This implies that the energies of these war-
rior spirits are also trapped in new weavings, and prevented from leaving by the
reinforced textile borders until they have fulfilled their creative function inside
the cloth.
Other contemporary evidence for this trapping of forces inside textiles is
found in the community of Coroma (Potosí), described by Bubba (1997), where
the inherited textile bundles (q'ipi) kept in family archives are considered living
ancestral beings, incarnations of human substance. There, community members
comment on the “speech” of the weavings, and the dynamic movement of the
beings that dwell in its interior.
Archeological evidence for this same idea is found in some pre-Columbian ex-
amples. Paul (2000b, 163) calls attention to the use of images of twisted and inter-
laced threads in the iconography of various sacred sites, for example within the
archaeological complex of Chavín de Huántar. She asks if these same textile im-
ages might have a sacred power that would impart spirituality to the images and
objects in which they appear. As an ancient visual expression of the same energy
that present-day textile borders keep in, this woven iconography in Chavín seems
to impart a sense of spirituality by trapping this spirit within the confines of the
sacred precincts. Another modern example that reiterates this idea is when Don
Domingo Jiménez says “you should not loosen the threads on a woman’s weav-
ing borders,” but rather locate there the pulu and k'illpha borders.

M en ’ s O pen T e x tile B orders

Opposing tendencies in the pieces men weave emphasize verticality and open
borders. For example, in the case of the rough homespun or cordillate that men
have made on the Spanish upright loom since the Conquest, the warps are cut
on loosening the weaving from the loom, something never done in the garments
that women weave.
This vertical tendency is also evident in the case of garments women weave
for men, for example the poncho, where the verticality of the stripes is empha-
sized in their wear, although their openness is expressed in other ways. Instead
of the woven selvedge and closed borders of female garments, this masculine
garment has a slit or “mouth” (laka) to pass the head through, rather like the tu-
nics (unku) men wore in the past (and that the Uru-Chipaya still use) with slits for
both head and arms.
In this textile and cultural logic, masculine garments communicate another
kind of information. Our previous studies of the poncho (Arnold 1997c) have

shown how the verticality of masculine ties expresses a more open discourse re-
lated to production and reproduction, but at a more inclusive level, in relations
between the ayllu and the world outside its immediate limits. This information
is codified in a more abstract and less figurative manner than that of female gar-
ments. The case of the kipu is the same. The codification by color and the width
of threads seem to express information about production from the lands and
flocks in the annual recounting demanded by the higher levels of the state.5 Evi-
dently the open borders of male garments communicate something homologous
to that communicated by the verticality of masculine garments.
There are other implications. Female garments are not cut from the loom or
in use; it would be “like cutting your own hands.” However, in male garments
the weaver can cut the warp threads or leave slits in the seams, both for the head
and for the hands. This difference implies a gendered distinction in the attitude
toward the forces within the textile. In the case of women, the forces of the war-
rior spirits within the textile (or belly) are closed in and prevented from escaping
by the reinforced borders, perhaps because these captured forces help in textile
and farming production. But in the case of men, any similar force in their gar-
ments is free to leave. Besides, while the valorization of manual labor of women
automatically values the role of their hands, it is as if male manual labor, and
even the heads of men, were more dispensable, and could be removed if neces-
sary. This leads us to ask if male manual labor and heads served for other matters,
say those of the state at a higher level.

C omparative D ata on the T e x tile B orders : k ' i l l p h a and p u l u

Some comparative points about the two kinds of textile borders, that of the
weft (k'illpha) and warp (pulu), are examined by Crickmay (1997). For her, these
two borders form a pair, with feminine connotations in the case of the k'illpha,
and masculine ones in the case of the pulu. Crickmay affirms that the k'illpha
weft border is related to the horizontal ties woven by a woman throughout the
ayllu via her menstrual blood. In contrast, pulu, the term denominating the two
first weft threads, generally of a red color, has a semantic field related to seeds
or semen, and with male blood (1997, 542). For Crickmay, pulu also has to do
with matters of war. In her reading of the seventeenth-century Vocabulario by the
Jesuit González Holguín ([1608] 1952, 298), pulu as an allophone of puru might
also belong to the same semantic domain as phurur awqa: “stone bullets hurled
from the top of a fortress to defend it.” With the literal meaning “staunch sol-
diers (awqa) of seed,” phurur awqa were round stones, like bullets.6
This striking gendered imagery of the two textile borders reiterates the con-
252 A N D E A N T E X T S A N D T H E I R I N T E R P R E TAT I O N

cerns of ideal couple of weaver and warrior examined in chapter 5. The mean-
ings of the feminine border allude to female fecundity and the potential of the
land to produce a harvest of babies, whereas those of the masculine border al-
lude to masculine strength in agricultural labor (and war): “The recurrent images
related to the terms for the borders refer to feminine fertility as potential—the
land—and to masculine fertility as a motivating influence—breaking the land,
plowing, irrigating” (Crickmay 1997, 532).
Citing the seventeenth-century Aymara Vocabulario by Ludovico Bertonio,
Crickmay also points out that the combined forces of these two borders result in
“a healthy life” and a “harmonious speaking or reasoning” (ibid., 542):
Killponi poloni arusitha. To speak harmoniously, as a man who is immersed in
what he is talking about.
Killpani poloni hacama. To live healthily, or that God gives you health.
Poloni, killpani arusitha. To reason harmoniously.7

In this instance, it is as if the combined textile borders were speaking with the
voice of the new being formed on completing the weaving, and as a result of the
ideal marriage of weaver and warrior.8
Other observations by Crickmay about the meanings of these two kinds of
border (pulu and k'illpha) refer to the more masculine domain of libations, in
the context of the relationship community-state. For example, Crickmay points
out how the themes of the male blood of sacrifice and male seed are united in
the meaning of pulu or puru as a calabash or gourd used for keeping or serving
liquids. Polo polo, according to Bertonio’s Vocabulario, is “a bowl or dried gourd
for water” ([1612] 1984, pt. 2, p. 273). Nowadays in the town of Bolívar (Cocha-
bamba Department, Bolivia), where Crickmay did her fieldwork, the gourds used
for the ritual serving of chicha (corn beer) are called pulupulu; in the valleys of
Qaqachaka they are simply called pululu. The calabashes contain not only large
quantities of corn seeds but also water, another symbol of masculine fertilizing
According to Crickmay (1997, 544), another Quechua term for a calabash or
receptacle in the form of a gourd or jug, kaka, has a semantic domain similar to
that of pulu. Other Quechua terms in González Holguin’s Vocabulario associate
the root term kaka with kakacuna: “those of the pueblo who contribute,” where
the verb kakacuni means “to contribute all the things to eat for the provisioning
of the tambo waystation for payment, or owed to the priest or encomendero of
foods” ([1608] 1952, 127). In this sense, Crickmay relates the semantic domain of
kaka-pulu with community members as tributaries to the state.
If this is so, then the semantic range of the textile term pulu not only echoes

the domain of the ideal couple of weaver and warrior, but also their obligations
to the state, mediated through libations.

A n d e a n N u m e r i c a l P r a c t i c e s at a T e x t i l e L e v e l

Desrosiers also discusses the question of gender and its woven expression
through the handling of the warp threads. In this context, she notes another con-
sistent gendered difference between men and women concerning the “knotting”
of the warp threads, as men make knots whereas women do not, whether in con-
temporary Andean weavings or those of pre-Columbian times. Speaking about
women’s textiles, she comments: “always and whenever it was possible, knots
were avoided.......practices [that] coincided with the pre-Hispanic period” (1997,
329). Her observation begs the question about the nature of this difference, its
reason for being, and the explanations given by the weavers themselves.
In chapter 10, we argued that this difference concerned the notion of “luck”
(surti), and the way it was handled by men and women as groups, at both local
and regional levels. There were contexts in which knots should be tied, and oth-
ers when they should not. The explanation given was that the tying of a knot on
these occasions makes the luck disappear, and as a consequence, the hands be-
come tied, the feet and the body too, reaching a situation in which “you cannot
do anything.” However, in other contexts, for example in the annual counting of
the animals, you must tie a knot to mark the end of one annual cycle and the be-
ginning of another. In this case, the action of tying the knot helps close one cycle
and open another.
In this operational logic, women take advantage of the forces of the warrior
spirits captured and immobilized in a textile, which give them energy for their
manual tasks and direct them how to achieve the desired dissemination and cir-
culation of their weavings, and the objects prefigured in the weavings, through-
out the ayllu. Men count with the same strength of the captured warrior spirits
in their local tasks, but their obligations in regional and state matters seem to de-
mand that they keep the knots tied, but “open their pathway,” releasing any luck
from them. If the luck were to be tied on these occasions, then hands, feet, and
bodies would also become immobilized. This implies that the proper functioning
of the regional system demands the circulation of elements not only within the
ayllu but also outside, toward the state center. The flow of masculine voices in li-
bations for luck seems to have the same power to impel the movement of these
other elements, too.
254 A N D E A N T E X T S A N D T H E I R I N T E R P R E TAT I O N

C ounting in T e x tile M athematics

The differential power of men’s voices in libation making and women’s voices
in weaving derives from the meaning of counting and its relationship with the
voice. This, in turn, has to do with gendered differences in the corporeal nature
of weavings that stimulate the flow of energies that will liberate the voice. For
both men and women, this incarnate and embedded vocal power and the dis-
course it embodies, whether in libation making or weaving, are fertilizing and
In chapter 5 we showed how both women and men learn the basics of count-
ing practice in weaving and in pretend libation making, from the age of five or
six until “they become accustomed to it.” As Don Domingo comments: “No one
teaches them, they just learn to count (wakhuña) in their hearts; they don’t even
ask anyone. First they select the heddles and then they weave........They really
learn through weaving.”
Women’s arithmetical ability advances with the elaboration of each weaving.
Their three-dimensional ability entails conceptualizing some eighty operations at
the layout stage before weaving even begins.9 These operations shape textile enu-
meration on warping the loom, by measuring out the length of the warps, and
then counting the warp threads in fives and tens, as they are tied back to the ini-
tial loom pole. As the weaving advances they count each pass of the weft to select
(p'itaña) the designs, their number, size, and proportion, and the heddle changes
to increase the depth and complexity of the cloth, from plain (inaki) to selected
(apsu), with up to seven heddle sets (illawa).10
Their weaving techniques differentiate between counting the “selected”
threads of an apsu design, where the two faces of the cloth are distinct, as apsu
wakhu, and the “plain” part of the cloth, called the pampa, where the textiles
faces are identical, as llanu wakhu. In addition, Qaqachaka terminology differenti-
ates between apsu, when the two faces have the same number of threads (for ex-
ample two and two), and ch'ulla when they are “odd and even” (two and one, or
four and one).
Each garment that women weave demands a distinct way of counting the
threads, so there is the counting of the mantle (away wakhu), the overskirt (urkhu
wakhu), the poncho (punchu wakhu), the belt (t'isnu wakhu) and belt tie (watu
wakhu), and different braids (mullit'imnt wakhu). In addition, each design, whether
of 6, 10, 15, 20, or 200 threads, has its own form of counting that women know
from experience, simply from the name of the figure (for example, sun, condor,
tipa leaf, dog paw, or winding path).
Men work mainly with homespun and cordillate in which the techniques of

measuring, counting and selecting threads are less complex. Designs in home-
spun only go in twos, with up to four heddles. Cordillate, the thickest and most
difficult cloth, “is apsu and goes in odds and evens,” and entails counting the por-
tada width units in dozens and tens.
In each case, the operative counting techniques shape the framework of an
insistent replicative practice that first makes objects material in cloth, and then
directs their dissemination into the real world by a process of wishing them into
being, accompanied by the weaver’s own vocalized discourse of creation.11 In this
sense, counted threads act as the foundational support through which a harvest
of new beings (condors, dogs, potatoes, llamas, children) proliferates through
various discourses.
Each of the different counting techniques embedded in everyday weaving
practice is at the service of a wider set of gendered obligations that derive from
the complex of ontological depredation. Here, the counting techniques of weav-
ing practice served in the past to bind the forces of the dead enemy Other, and
then multiply them in favor of the own group.
This more metaphysical dimension of counting techniques is centered now-
adays in the practices around “luck,” with their gendered differences. Women,
with their greater ability in counting, tend to focus luck (and in this way the idea
of multiplication) in weaving and song, whereas men, with less ability in count-
ing, tend to focus luck (and multiplicity) in their libations. In the corresponding
corporeal differences according to gender, women center their abilities in the
generation and tending of numerous babies (wawa), whereas men center their
energies in seminal strength for procreation, and physical strength in warfare, for
obtaining trophy heads or their modern equivalents (the consumer goods of the
modern marketplace), and manual labor for producing and providing nourish-
ment for the different kinds of baby (animals, plants, and people).
Men, in the vocalization of memory during libation making, follow a mental
pattern homologous with that of the kipu structure to “knot threads” in order
to “grasp luck.” Women have other techniques of “grasping luck” without mak-
ing knots. This difference derives from the idea that making knots is an aspect
of male physical strength (ch'ama). Men are considered to have “a great deal of
physical strength” (wali ch'amani), which their ability at knotting reinforces, while
women have “lesser physical strength” (pisi ch'amani).
In this sense, handling textiles and textile structures has different meanings
for women and men. In the numerical practices of their everyday weaving tasks,
women order the structures of textile borders in a more immediate way, while
men order their numerical practices of making libations by drawing on the fertil-
izing powers of these female-centered textile structures.
256 A N D E A N T E X T S A N D T H E I R I N T E R P R E TAT I O N

Thus men allude to the structure of textile edges, especially the thick thread
of the first selvedge, called pulu ch'ukurkata, and the way of “tightening it by beat-
ing” (thawkaña) to make their libations “hit the mark.” For example, Don Do-
mingo mentioned the setting out of the pulu holding thread (pulu ch'ukarkata)
in his libations to the animals (oxen, sheep, goats, and so on), comparing it to
making furrows in the earth. For him, the function of making a toast “for this
holding thread,” and then for the action of beating down the pulu selvedge, as
if you were covering the furrows with earth, is “in order to make luck reach its
mark” (surt asirtu). In this instance, the naming of this edge, and the action of
beating it down during the toasting, reinforces the male labor of making furrows,
whether in the earth during agricultural production, or in opening the furrows
of the memory, in the ritual labor of making libations, or implicitly in the male
role in reproduction.
These homologies highlight a sequential nexus between number and vocal
practice. First is the generalized notion of counting (wakhuña) as an initial immo-
bilizing practice, and the homologous way that the holding thread (or principal
kipu thread) counts, orders, and immobilizes the textile under its control. Then
the more discursive ways of memorizing knots in toasts (or in weaving or kipu
reading), as a directing practice laid out in pathways, finally generate a whole dis-
course of multiplying words concerning material objects and their dissemination
in the world.
Don Domingo proposed that the libations to the llama were “as if we were
counting llamas.” Then he compared the libation pathway with a ribbon that re-
constructed the whole of llama being: “It is to reconstitute all that the llama is,
through the toasts, their saints, their colors, their loads, all in the form of a rib-
bon. It’s as if we were going along their pathway.” As we saw, llamas gain over all
the other herd animals in the quantity of toasts, and the length of kipu threads
dedicated to them, because of their sheer quantity of possible colors. “Then the
libation ends, making luck is ended, and the earth on the furrow-thread is rear-
ranged” (through the action called thawkataña in the sense of rearranging the
Just as women count colored threads in weaving, so men count colored
threads derived from llama fleece in making toasts; both aim to multiply the
flocks. We suggested earlier that ideas concerning the fecundity of a variety of
colors, as well as their multiplying power, may have derived in the past from the
quantity of hair a warrior might have captured in the form of trophy heads from
defeated enemies. As we saw in chapter 4, pendant kipu threads are still com-
pared with the hair of the head, and the laying out of colored warp threads on a
loom is compared to tending the hair of the Other.

In practice, the potency of llama hair to create new animals for the flocks is
released in the toasts through the intervention of the human voice. The same
happened in the case of the human hair of captured trophy heads, which were
prayed over by the women of the capturer’s household, before being woven into
new beings. Women also accompany weaving by a spoken discourse that “ex-
ternalizes” the captured being in the cloth, converting it into a baby of her own
group. This suggests that the gradual reconstitution and multiplication of being
(whether llamas or other animals, even the food produce, or human babies) fol-
lows the initial destruction of the Other, according to a logic of sacrifice and its
accompanying vocalization.
For Don Domingo, the very act of voicing (arsusxa) in the toasts is equivalent
to the act of counting (wakhuña):
Qawqha ch'allanxay arsuxa, In how many toasts will I voice them?
ukay wakhux. This is what counting is.

Even more specifically, “a voiced toast” is equivalent to “tying a knot” of words:

Mä ch'allax qawqhay arsuxa, In a libation, how many times will I vocalize
may akan arsuxa, when I vocalize them just once,
uka mäya chinuraraktsä. I’ve tied this [thread] once too.

The use of the verb arsuña here for “voicing” is particularly apt. The root (aru-)
means “word” and the directional suffix (-su) indicates “taking out words from in-
side,” that is, externalizing them by voicing them. In parallel with women’s mul-
tiplying power in weaving, male strength in libation making concerns external-
izing into a knot of words, the words embedded in a knot, while the tone of the
knot “colors” the male voice.

T ying L uck and G endered P rocreation

through C ounting

These present-day, gendered differences in textual practice imply that count-

ing with kipus in the past might have been a more male activity, in parallel with
women’s counting in weaving. Even now, in the region of the present study, men
resort to a mental image of the mnemonic logic of a kipu when they recite the
libations of wakhu chinu (“counting knot”) and ch'alla chinu or surti chinu (“liba-
tions knot” or “knot for luck”).
In practice, both men’s and women’s counting practices concern gendered as-
pects of reproductive power at the service of their respective tasks. For their part,
258 A N D E A N T E X T S A N D T H E I R I N T E R P R E TAT I O N

men draw on images from their manual labor in the fields to direct luck to the
household and its produce through libation making.
Don Domingo described how this was done. When toasts are made to ani-
mals such as the oxen, you drink first for the “holding thread” and “its action of
covering the furrows with earth.” This is done with reference to the principal
cord of a kipu (real or imaginary). So men name the textile border called pulu (or
pulu ch'ukurkata), which is homologous to the principal cord of the kipu, men-
tioning the fact that it is tied up and beaten to seat it in its location, as the proper
basis for beginning the toasting sequence. He states, “It is its foundation” (si-
mintu); the pulu beater is what “finds the luck” (surti asirtu). Making these first li-
bations is “as if we were counting” the animals. “First the llamas are counted, the
libations are like this, I tie it here, here, here.......isn’t it so?”
Following the mental pattern of a knotted kipu, the sequence of libations fol-
lows, as we saw, with the name of the animal’s patron saint (Päri Antuñu, “Father
Antonio,” in the case of the llama), as if you were counting all the llamas in their
totality, “their pathway goes from there; if it weren’t so, the llamas would die
out.” Then, in the rest of the sequence, the threads “always go in pairs,” male and
female, in a reproductive sense.
Luck is “hit” twice, once at the beginning of the sequence, by naming the
principal cord, and then on ending the sequence “luck is grasped” again by say-
ing “chinu mayku, chinu t'alla” (Lord and Lady of the Knots), and by vocalizing the
final “rounding off ” as a llama pen of words. In this sense, making luck as a mul-
tiplying act derives its power from replicating the same braided structure of the
kipu, as a guiding principle that is confirmed vocally.
The local discourse about mnemonic practices during libation and knot mak-
ing (even mentally) not only relates these multiplying acts with the generative
configuration of knots and threads, but also with the flow of gendered bodily
substances within them. Different threads and knots, perceived as male and fe-
male, unite to reproduce in quantity, and the vocalizing of this reproductive po-
tential contributes to the generative power of Andean orality. However, impor-
tant gendered differences in discursive priorities skew attitudes toward the nature
of this power and its agency in the world.
For men, the dynamics of recording libations and making knots, whether real
or imaginary, replicates a masculine germinating force and motivating influence
of the same kind as was needed for breaking the ground (or an enemy), and then
plowing and watering it. In this sense, the very act of a man tying a knot provides
Don Domingo held that when the knot is made well, the animals “go up-

Fig. 31. Detail of a kipu pendant thread, with the larger quantities above. Based on Locke (1978).

ward,” meaning that they augment and do not diminish in quantity. He was al-
luding to a mode of quantifying in which the animals are imagined to be going
in an upward direction in a pendant kipu cord. This is in keeping with the classic
studies of counting with kipus, for example the Aschers’ book Code of the Quipu,
which show that the larger quantities are always found higher up, closer to the
main cord. See figure 31.
The male strength applied on tying the knot is a masculine animating energy.
Talking of its nature, Don Domingo commented that “it is as if marrow (parpa)
had been introduced there, so that the animals don’t grow thin (tuxuntaña).” He
continued: “if the animals have good marrow, they have lots of fat and are really
plump. Although the animal is thin, it will be fine with some good marrow. Even
a thin animal with marrow can gain over a fat animal without much marrow,
knocking it over like a mere shrub.” This implies that on voicing the libations,
and making a knot, male energy gives marrow to the bones and fat to an animal’s
body. And by making the knot “really hard” (suma turu) with the words of the
toast, a lot of strength is given to the animals.
The ceremonial words used during toasting transmit this particularly mascu-
line and procreative force to the animals themselves. When an animal does not
260 A N D E A N T E X T S A N D T H E I R I N T E R P R E TAT I O N

have much marrow, it is weak, but when it comes to have marrow through the
libations, luck making, and tying of knots, it will come to have male sexual ener-
gies and the capacity to generate young. By the same logic, the more libations
there are, the more offspring there will be:
When they don’t have marrow, the bone is hollow (phusa) and thin (tuxa). Although it’s
a young man or woman, it’s hollow. The animals are the same. But when we come to
make luck and tie [knots], it’s as if we were putting marrow there. Now this one will have
marrow; it will mount once or many times like the bull on the cow, and then it will have
young, one and another. When there are many knots, it’s so that they multiply with many

Don Domingo included females in this same logic. “The one that’s not tied
well is hollow and thin, because she doesn’t have any blood, so what blood will
she receive [male seed] in? She’s empty, isn’t she? She will become barren.” But,
through libations and the tying of knots, this female comes to have blood and
then procreates: “This other one has good marrow. There’s blood in each thread.
Each knot has blood, and they’ll procreate in this way. It has lots of fat.”
In Don Domingo’s opinion, the male strength that results from the combined
actions of tying the knots well and voicing well the words of the libations has the
power to generate both male and female bodily substances (semen and blood, re-
spectively), first in the threads and knots, then in the animals. As a consequence,
the animals will come to have energy, procreative strength, and offspring.
In an Amazonian context, the idea that male semen generates female blood,
and so influences women’s fertility, has been called “seminal nourishment.”12 This
same idea is evident in the Andes, at least in male discourse. But another underly-
ing discourse recognizes the feminine part in regeneration. This emerged when
Don Domingo, at the end of his clarification, confirmed that each thread is, in
fact, a juncture of two (or three) strands of different colors and sex. In the black
(or red) strand “runs the blood of the female” and in the black one “runs the se-
men of the male.” (See figure 32.) For Don Domingo, the different strands of the
thread communicate among themselves through the sexual substances that flow
in their interiors to then unite in greater quantity in the knots, where the fertiliz-
ing mixture produces the spiritual energy (ispiritu) that generates young: “Blood
runs in one strand and semen in the other (or it might be a combination of three).
With his mixture, the blood [in the black or red strand = female] receives, the se-
men [the white strand = male] gives the seed, and the knot receives the ispiritu
that results from this mixture. The blood runs in one and the seed in the other.”13
For him, the two (or three) strands of each thread are like a couple (chachawarmi);
one is the man (chacha) and the other is the woman (warmi). “If there were only

deposit (almasina)
= a woman’s belly with the
seed-baby inside

white strand = male = flow of seed

black strand = female = flow of blood

knot = wawa = mixture of seed and blood

Fig. 32. Kipu with baby knots and the flow of sexual substances.

the man, nothing, or if there were only the women, nothing........It’s only [pos-
sible with] a ‘pair of doves,’ because one of them alone would do nothing.”
Don Domingo explained in this way the textual corporeality of the kipus that
he works with. For him, the knot that grasps the mixture of seed and blood is
like the entrance to a woman’s belly (the birth canal or vagina, jaqiwa) where the
seed is received and mixed with blood during the couple’s sexual activity (at the
dark and full moon) and converted into a person. A woman’s belly, like the main
cord of the kipu, is a “storeroom” or deposit. And the male voice has the seminal
power to generate the baby knots in the kipus, and through the kipus into lived
After comparing the two strands of the kipu cord with the form of voicing li-
bations, Don Domingo went on to compare the different ways that toasts should
be served to the participants present. “This is why two cups are always offered
in the libations, one for the male and the other for the female. The toast to the
female goes “toward the left” (lluq'iläru) and that for the male goes “toward the
right” (kupituqi), like the two strands of the kipu cord. In other words, the gen-
dered pattern of braiding the strands is homologous with the patterned move-
ment of the toasts. His remarks reveal the way that weaving activities persistently
structure everyday practice.
For him, voicing libations through masculine words regenerates life at many
262 A N D E A N T E X T S A N D T H E I R I N T E R P R E TAT I O N

levels. In a shared logical pattern this practice of voicing contributes to the sexual
life of the strands, the coupling of the libations, and the sexual life of couples.
The result is the fertilizing of the world and the production of a new generation
of babies (human, animal, and food crops).
In a similar way, vocal power passes to the knots so that these, too, “speak.”
Then the same power of speaking passes on to animals that the libations are di-
rected to, such as the llama, especially when you speak “with many words” (walja
palabraw). “It is as if there were many llamas there, and these must speak, the
male and female young, and the male and female adults.” Within the same logic,
if there are more libations with more words, then there will be more offspring in
the world. The power of the voice acts as the incentive (agency) for all this cre-
Another aspect of the same idea is centered in the large double knots (called
“turning knots”) of some kipu threads, which replicate the animal pens. As we
saw in the case of the guiding thread for toasting the llamas, a variety of middle-
sized knots serve as mnemonic resources for the libations to which they pertain.
Then, at the end of the sequence of libations, a large double knot at the end of
the thread “receives all the other knots” in its interior, “like a pen that receives
all the llamas each night to release them again the following day.” According to
Don Domingo, the very size of this knot influences the procreation of the ani-
mals named there. These procreate until they enter the “turning.” Then, as the
middle-sized knots enter the large knot, “it is as if all the words that belonged to
each middle-sized knot were to group themselves into the whole discourse of
the large knot.” So, in order to increase the flocks, you must do a really long liba-
tion, mentioning many names “so that all of them feel included,” and then you
achieve the filling up of the large knot.
On ending the sequence, the kipu’s main cord, in its function as the border
(pulu), “will adjust” (thawkatxanixa) the entrance of the middle-sized knots into
the large knot, to complete the series of toasts there. As Don Domingo says,
“The pulu will adjust it there.” Then the principal cord (pulu) has to “give leave”
(licencia) to the llamas so that they may go out to graze in another moment, via
the loose end of the cord.

G endered D ifferences in T e x tile C reation

These gendered attitudes toward textile borders, counting, discursive prac-

tices about counting techniques, and the tying of knots, all in turn color the at-
titudes of men and women toward both the metaphysics and corporeality of
textile practices. Men evidently count kipu knots according to the criteria of the

ritual practice called surti chinu, “lucky knots,” to generate male procreative force
in the world. But is there any female equivalent to this which generates female
procreative force in the world?
For Don Domingo, the large textiles (mantle or poncho) that women weave
do have their knots, but the equivalent female task of “knotting” is their tying
of the thick pulu threads back to the loom with the holding thread to order the
warp set at the start of weaving. From his point of view, this female task has two
slightly different aspects. One is that of “counting,” when the holding thread is
“that which knows.” Here counting is a way of knowledge, as the holding thread
has to “know” the entire process of weaving in advance, since the laying out and
counting of the warp determines the desired figures and plain background that
can be woven. In Don Domingo’s words, “We ourselves say that the holding
thread has a good mind.”
The other aspect of the holding thread is that of “grasping.” When a woman
warps the loom, once the thick pulu threads are in place, she wraps (muyu-
kipayaña) the holding thread around the initial loom pole and each few warp
threads in a continuous helix. (See again figure 30.) In this way, the holding thread
“grasps” the textile at the start of weaving, as well as “undoing” it (jararaña) at
the end, since the completed weaving can only be untied from the loom by un-
winding the holding thread, without the need for cutting anything.
As a man, Don Domingo perceives in the function of the holding thread as
“that which orders” the action of weaving in advance, the very “foundation” of
weaving activity, the basis from which all else derives. He recognizes that the
holding thread must be special, made of the thick rough wool (called wala wala)
from the llama’s neck, as “this wool endures and can’t be pulled out.” If the hold-
ing thread has luck, then a woman will finish her loom quickly, according to how
much energy it gives her. In the opposite case, if she has no energy, she will delay
in weaving. This is why, before beginning to weave, women chew coca, contem-
plating the strength of the holding thread. For Don Domingo, this whole tex-
tile configuration is not just a matter of technique; it has a profound religious
sense. It constitutes a foundation related to the trinity of God, so when three
thick threads are used for the pulu, he calls them, after Andean seminal thought,
“Father, Son, and Holy Spirit” (Yus yuq ispiritu).
In contrast to Don Domingo, the young weaver Elvira Espejo, although she
agrees that the holding thread “has thoughts,” gives more importance to the role
of the heddles (illawa), both in counting and in textile thinking. For her, “The
heddles count and know about the threads.” The role of the heddles in the weav-
ing process also continues from start to finish, as each moment they have to
“communicate” to the weaver if there is some thread that is not in its right place.
264 A N D E A N T E X T S A N D T H E I R I N T E R P R E TAT I O N

As she says, “The heddles tell you about a thread that is not grasped properly, as
they have thoughts.” The heddles with their scissor movement (like the zigzag
of lightning) order the threads on opening and closing the textile mouth (laka).
In this case, the “mouth” is the open fell (end of a web of cloth, called khachi) be-
tween the different levels of cloth, which opens and closes in each passing of the
weft. As Elvira says, the heddles “go into the mouth. You take them down to the
The weaver Elvira also has a broader sense of the textile voice and its compo-
sition from various elements. For her, the action of weaving is accompanied by
the vocalic sound of the llama bone (wich'xata) that beats down on the thick pulu
threads, and then in each pass of the weft, with its beating voice. But the principal
voice of the textile “is inside the mouth [khachi],” where “it is eating the threads
of the weft as its food.” She concluded her analysis by comparing textiles and
books, as both have voices. For her, the textile “has a mouth; it’s like a friend or a
woman friend. It attracts me; it’s a friend like a book, that’s how it is. It talks like a
book, the textile talks in this way.”
Elvira’s commentary helps clarify how the female act of weaving, with its
foundation knot and heddle voice-maker, provides in and of itself the procreative
dynamics for forming a new being, through the stages of baby, wawa, and then
the complete person, jaqi, examined in chapters 4 and 5. Both Elvira’s and Don
Domingo’s analyses also imply that female procreative power, with its multiplica-
tive and multivocal dimensions, serves a parallel function to what men accom-
plish through the use of the kipu and the male voice. This would suggest that a
woman’s task of weaving of this new being constitutes the complementary fe-
male aspect of the trophy head complex, when the spirit of the trophy head of
the enemy Other, originally captured by the warrior husband, is finally revivified
as a new baby Self of her own family group. It also explains why the final part
of the cycle of metamorphosis, in female hands, is founded in a woven body of
knowledge and textual transmission.
One aspect of this corpus is how regional discursive practice around textual
genealogy emphasizes female presence. This discursive tradition traces the for-
mation of the new woven being through quasi-biological stages. In this woven
ontology of being, the weaver first makes an offering to the earth and then sets
out the cage of loom poles. In parallel with male shamanic discourse concerning
the “sprouting” of offerings to the mountains, the hard organic loom pole is held
to have the power of initiating textile growth, in the same way as the small fam-
ily chest with its wooden board base, where the kipus were and in some cases still
are guarded.14 Remember that Don Domingo compared these elements, in the

female case, to the “mucous plug” in the mouth of the uterus, from which the
baby is thought to grow.
The weaver then tends the hair of the Other, incarnate in the colored warp
threads, to be converted gradually into a new baby through the creative act of
weaving. One of the textile voices is incarnate in the tended hair. For example,
Juan de Dios Yapita has heard it said, “The hair is wont to speak and count.”
The inserting of the thick red pulu threads and the tying of them back to the
initial weaving pole with the holding thread constitutes the foundation of the
procreative process, a rack for reworking the captured being there. In the sub-
sequent weaving dynamics, the fell-space that opens and closes with the heddle
movement forms a textile mouth, which is held to be breathing (samsji) and eat-
ing (maq'asji) with each passing of the weft from outside to inside. As Don Do-
mingo explained: “It breathes from here, and then it goes over that side and is
made to come back again from there to here, and it ends here. It goes in and
breathes too from here.” In the corporeality of cloth, each passing of the weft
is thought of simultaneously as the breathing in and out, and digestive process
of the new being. The weft is “that which personalizes” the weaving; its motion
“converts it into a person” (akay jaqichixa).
The weaving is constantly tightened after each passing of the weft by push-
ing down the new weft threads onto the advancing cloth with a pointed llama
bone called wich'uña. The weaver reinforces this action with additional punches,
as if she were asserting her dominance over the captured being incarnate in the
Finally, the textile is usually finished with a section without designs but “with
steps” (patapatani), which gives it its name. This has to be executed with a long
needle (yawri), as the tight warp no longer allows the weft to enter. The part fin-
ished by needle is considered to be the “crown” (sunaqi) of the textile, that is to
say the “top of its head.” Women consider that with each passing of the weft,
“the textile will be completed up to the crown of the head” (sunaqiru p'itsxani).
As another weaver, Emiliana Ylaya, says, “it’s as if we had created a whole being,
from the feet to the head.”
Once the weaving is complete, the holding threads are untied and the new
being is loosened into the world from its loom prison, as another baby product
to be harvested.

Sp e e c h i n W o m e n ’ s W e a v i n g s a n d M e n ’ s K i p u s

In a situation of ongoing conflict, a broader textual theory about the poten-

tiality of the death of the enemy Other, and its later revivification into a new be-
266 A N D E A N T E X T S A N D T H E I R I N T E R P R E TAT I O N

ing, should include a comparative account of the agency of speech and voice in a
woman’s weaving and a man’s kipu. Once again, a regional discursive formation
founded in corporeality asserts a body of knowledge whose textual authority is
founded in weaving production, and articulated through woven voices.
A vital part of the development of the new textile being is its emergent ability
to speak. In both Qaqachaka and northern Potosí, the combined motion of weft
and heddles is considered to introduce breath, light, and food into the textile, and
also speech. A weaver explained how, with the motion of the weft, “from the one
side, it externalizes the voice (arsu) and then it ends, and so the voice ends, and
then from the other side it begins again.”
Don Domingo confirmed that the movement of the weft as “that which
breathed” (samsutpakay) made the textile speak, comparing this motion of tex-
tile speech with a human head (or perhaps the original trophy head) that talks:
“it finishes saying something and then it says something again. It’s as if, on mak-
ing each pass of the weft, you were letting a head talk.” For him, “Speech comes
from the head and ends again in the head.” He also asserted that each face of the
different colored sets of thread had the power of speech, having been nourished
with the veinlike threads at each passing of the weft.15
Such theories and practices of Andean corporeality (with their parallels in
classical and medieval Europe) are founded in the homologies posited between
the growth and development of a textile being and a human being. We saw how
human life is thought to develop from its foundation, the mucous plug that closes
the uterine opening during pregnancy, permitting the fetus to develop. A human
baby is first nourished by its parents with food produced in the furrows of the
earth, then this nourishment passes to the stomach and on to the heart, with the
circulation of the blood. The blood then flows from the heart to the head via the
“large vein” (jach'a wina), the archaic name for the windpipe-aorta complex. Fi-
nally, this combination of energies is externalized through the mouth, via the hu-
man voice. As the being grows, it is gradually nourished more in food and knowl-
edge in a constant articulation between outside and inside.
Similarly a textile body develops from its foundation (pulu) complex through
the tying of the holding threads. Gradually as the weaver passes the weft to and
fro, she introduces nourishment through the two textile mouths at the external
openings of the k'illpha borders on both sides of the advancing cloth. The weft
thread is compared to blood or veins that go in one direction and then the other,
nourishing the heart of the cloth. In nourishing the textile body, the heddle move-
ments open yet another mouth in the fell space that articulates the different warp
sets, allowing the new textile being to breathe and speak.
As Crickmay (1997, 541–42) pointed out, the combination of the two textile

borders (k'illpha and pulu) produces “reasoning,” a “healthy life,” and the possibil-
ity of “speaking harmoniously.” It is as if, with the completion of these borders
in the advancing cloth, the living being inside the textile were to begin to reason
and talk, as Bertonio wrote in his Vocabulario, “as a man immersed in what he is
talking about.”
As the weaving advances, the colored weft threads, like blood, gradually fill
the textile body up to the crown of the textile head, where the trapped ener-
gies of the captured being, having reached their limit, become externalized again
through the new textile voice. In particular, the female ability to manage the color
red in weaving determines the efficacy of this process of revivification. Don Do-
mingo confirmed how “Respiration and color, the play of colors, are women’s. A
man is just white and tan. Otherwise, a man would be able to weave.”
In this female-centered discourse, the dynamics of cloth is centered in the
revivification of an ancestral force that is ambiguously dangerous yet fertile, a
being captured and incarnate in the warrior spirit trapped within the textile bor-
ders, tied back to the loom pole, and beaten once in a while to control it effec-
tively. This discursive tradition locates its authority in a woven space of knowl-
edge, and the genealogical role played by women in rebirthing this force through
the blood red textile belly, in order to contribute their share toward the house-
hold harvest of babies.
The dynamics of kipu making and kipu reading, in the hands of men, oper-
ates according to the more logocentric assumptions of male procreation. In the
kipu structure, the pendant threads that descend from the main pulu-like cord are
read from above to below. Here the knots constitute the focus of speech, a visual
and tactile element that stimulates the powers of the male voice to record the
many words of a complex of ideas. Kipu corporeality concerns the flow of bodily
substances (blood and breath/seed) via the hollow threads that generate suste-
nance for the voice. These flow in a similar way to the movement of such sub-
stances in the human body, from the heart via the “large vein” to the head. In re-
lation to the trophy head complex, the kipu knots incarnate the memories from
the heads of enemy Others taken in battle. Through the logocentric process of
male vocalization, these are transformed into a harvest of own babies.
In the practice of vocalizing the memories held in these knots during libation
making, luck is achieved by making mental furrows (or pathways) via the threads.
Male counting practices are centered in this ability, related to their superior physi-
cal strength, to multiply babies through the voice and the quantification of words.
When there are more libations and more words, so there are more babies.
In contrastive terms, male or female procreative powers begin from a com-
mon basis (the pulu), whether as the main kipu cord or the first thick threads of
268 A N D E A N T E X T S A N D T H E I R I N T E R P R E TAT I O N


threads k'illpha

The warp-like The weft-like

male voice female-voice

vertical emphasis horizontal emphasis

Fig. 33. Direction and borders in the textiles of men and women.

woven cloth. Similarly, the verticality of the pendant kipu threads is homologous
with the verticality of the textile warp. Likewise the horizontal female dynamics
of cloth propelled by the movement of the weft, and the opening and closing of
the khachi textile mouth, has its equivalent in the male voicing of the kipu knots.
A fundamental difference, though, is that the male reading of kipus follows the
vertical rhythm of open-ended borders, so that the information contained there
can flow beyond the immediate confines of the ayllu, whereas the female read-
ing of cloth follows a horizontal rhythm of closed borders, in a more contained,
ayllucentric information flow. See figure 33.
In textual terms, these observations relate this regional notion of text with
ideas from other parts of the world. A key notion is that Andean textuality only
lives through the vocalization of the knowledge incarnate in the threads and
knots of the kipu, or the warp and weft of cloth. Without vocalization, these
texts simply remain as dead and mute beings. In essence, it is as if the vertical
threads of male kipus or the vertical warp of female cloth were to constitute the
elementary support of an Andean form of writing that only a posterior vocaliz-
ing, in the case of the kipu, or the to and fro of the weft in the case of cloth, can
complete into a fully functioning text.
A confirmation of the nature of voice in Andean textiles emerged in a chance
conversation with Don Domingo, when we showed him a photo of a trophy head

captured by the famous headhunters, the Achuar (or Jívaro) of Ecuador. The head
was already shrunk and the mouth secured with slivers of the black palmwood
called chonta. Wound around the pieces of chonta were threads whose pendant
ends dangled from the mouth. He commented in the following way: “This has
a head with lots of luck, that’s why it breathes so much luck [laughs]........This is
luck, this is luck, and this is luck [indicating the pendant threads]........The head is
really lucky, that’s how it is.”
We went on to ask Don Domingo about trophy heads (which he called p'iqi
t'uxlu) in an Andean context. He clarified the nexus between the voice of the con-
queror (in this case the headhunter) and what came out of the mouth of his vic-
tim, the trophy head. Then he compared the speech of the victim with the “luck”
of the threads that came out of his mouth, as if it were his own “breath coming
out” (samsuña) after exhaling his last gasps of life. He also compared the lucky
threads in the mouth of the trophy head with the Inka practice of tying knots af-
ter tiring out each cycle of luck, as a way of completing the exhalation process to
guarantee new wealth in the future:
It’s just as if I myself spoke out “This is so, and that is so,” isn’t it? It came out of my
mouth here. It’s the same. It’s as if it were one of my family.......[and I could consult it]
about everything. [And it would reply] “This is so, and that is so.” Then it comes out of
here [the voice].
“This is so, and this is so,” one knot, another knot, and here’s another knot. Even the
Inka did the same. Here it shows us all kinds of luck exhaling from the knot (samsuxa), it
exhales again, and you have to tie it. This is wealth (piña) of all kinds.

He concluded that the enemy himself “brought luck” in his own head.
Although the situation has changed in recent decades, and cases of decapi-
tated enemies are now extremely rare, nevertheless Don Domingo remembered
the warfare of the past. He records how they only took the heads “of the most
important people who knew how to talk well” (sumay parlt'iri). “They used to
take the heads of these people, and handle them,” as they had the power to bring
luck to the person who held them, so that he had more babies, more food crops,
and more animals. “That’s why he was called amuy p'iqi [thinking head].”
These guiding principles of textual practice in weaving cloth and braiding
kipus enable Andean rural populations to handle the relations between Self and
Other, voice and memory, and number and writing. The same guiding principles
have enabled these populations to deal with the similar issues in alphabetic read-
ing and writing for many centuries, when one of the principal institutional sites
for the appropriation of these distinct numerical and writing practices was none
other than the school precincts. There the community had to relate to the state
270 A N D E A N T E X T S A N D T H E I R I N T E R P R E TAT I O N

and, more specifically, discover how they might preserve the authoritative textual
polity of former Andean states in the face of the emerging demands of the mod-
ern nation-state. This was evidently done through the skillful subversion of cur-
rent hierarchy through an alternative apparatus of pathway (thakhi), corporeality,
and accounting, founded in social memory.
In order to make luck hit its mark, a series of libations had to be tied “in the
head.” On the contrary, if they were tied in reality, “it would be like tying up the
hands, the feet and the whole body.” As Don Domingo explained, you have to
leave the furrows of memory open as a series of pathways (“pathway, pathway,
pathway”). Here, the dynamic relation between body parts, voice, and memory
draws on a structuring image from local production (the furrows) to express the
ideal relations of communal service to an alternative state, in another writing-like
mental inscription.
In the structural dynamics of a kipu, you read the threads “from above down-
ward,” but at the same time the quantities must augment “from the bottom up-
ward.” When the knots are tied from above downward, luck is tied up, but there
is the danger of tying up the body parts, too. However, when the pathway of a
thread is recorded well, according to the furrow of memory from above down-
ward, then the animals augment their young, and this quantity goes from the
bottom upward. Also, there is no danger for the body parts. The question then is
to open the furrow of memory, and then assure that the vocal power associated
with it speaks from the top downward.
Comparatively speaking, in the human body, it is the upward movement of
the blood and breath from the heart via the great vein complex (of windpipe and
aorta) to the head, which generates the power for the voice to externalize itself.
Could the upward motion of these substances in the hollow threads stimulate
the same vocal power to spill over into a much wider domain?
This seems to occur in the context of communal-state relations, as expressed
in the liquid medium of making libations. We saw how the concentration of
voices of all the participants making toasts throughout the Andes impel a pro-
creative movement upward in the kipu pendant threads, not only for the animals
and their young, but for all the body parts. In this sense, by opening the furrows
of social memory, the participants in libation making give over their drunken
voices to the service of the state, in their role as tributaries of the community
(commune) of “all the things to eat for the provision of ” the state. The name pulu
for the main kipu cord is resonant with these meanings. And, in terms of hierar-
chy, it is only through its power over a whole collectivity of voices that the state
system, in its totality, can function.

A recent study of textual practices in ancient Greece by Svenbro (1998)

reaches a similar conclusion. He redefines “text” as the dynamic relation between
vocalization (in the sense of the to and fro of the weft) and the vertical elements
of the textile warp (comparable to the kipu’s pendant threads), as the “writing”
Playing with the etymology of the word “text” (from the Latin textus, “textile”), I have the
impression that everything happened as if the text were formed by a written warp and
a vocal weft that was handled in reading and later undone........Within this conception,
that I don’t think is faithful to the classical experience of reading, the text would not only
be therefore a static object, but also the term given to the dynamic relationship between
what is written and the voice, between the writer and the reader. So, the text would be
converted in the sonorous realization of the written, a writing that could not be distrib-
uted or spoken about without the intervention of the reader’s voice. (1988, 69)

There, “writing” in itself was incomplete, and could only be completed through
the voice.
For Svenbro, the essence of reading in ancient Greece was its vocal nature, as
it is in the Andes. There was the same instrumental character of the reader and
the reading voice. In addition, the reading voice was clearly at the service of what
was written by the state. That is to say, the reader gave up his (or her) voice to the
service of the state. It is as if the voice “did not belong to the reader during the
reading: he had ceded it over.” Even more importantly, “the writer that achieved
being read, acted on the vocal apparatus of the other, who served him, even after
his death, as an instrumentum vocale, that is to say as someone or something at his
service, like a slave” (ibid., 70).
These observations, transferred to an Andean context, clarify the specific
character of its textual polity, and the primary source of textual authority and
textual domination. From the regional history we have examined in these chap-
ters, who made writing in the Andes? None other than the Inka. And, at whose
service was the whole chorus of drunken voices in the recitation of libations?
Surely the Inka’s, in his role as the head and hands of all, he who incarnated the
kipu’s main cord and to whom the pathways of memory were directed. As they
say in the communal assemblies of Qaqachaka:
Juma Inka thakhip sarirïta, You are wont to go on the pathway of the Inka,
liyt'amay...... so please read it......

With the memory of the Inka in mind, you are obliged to give yourself up, blood
and bone, manual labor, and trophy head in war.
272 A N D E A N T E X T S A N D T H E I R I N T E R P R E TAT I O N

To answer the questions posed at the beginning of this chapter, we must con-
clude with Derrida that in the Andes, too, voice underlies writing, whether this
writing takes the form of weaving, kipu, or any other woven or braided vocal
support. But even more primordial than voice is sacrifice, something intuited by
Derrida but never expressed openly in his work. In this sense, Viveiros de Cas-
tro’s notion of ontological depredation, where the play of reconstructing the Self
from an enemy Other in a situation of imminent violence, comes closer to the
Andean experience.
In addition, we must conclude that vocal reading in the Andes is still at its
peak, not only in all the rural communities of our study but in all the rural
schools. Whether in voicing the libations of Andean toasting, or the letters of
Western reading and writing, the memory of the Inka is being recorded long af-
ter his death.
Guaman Poma’s description of the handling of kipus by the Inka state, in
chapter 1, implies that the content of the accounts carried by the secretarios of the
period were “the very words of the Inka,” or else of the Inka with those of royal
lords of Tawantinsuyu. All of this suggests that the textual basis on which the
words of the Inka are written has evolved to include not only the toasts of the co-
munarios, based in any kipu and textile supports still in use, but also the activities
of the nation centered in the school compound.
In this sense, the multivocality of Andean writing expresses a conception of
territoriality in which ongoing echoes of the Inka’s voice reverberate through
each stone/warp and letter/design. Therefore Andean writing in its many tex-
tual forms does not easily obey the processes of colonization (from the Spain of
Nebrija) that saw in writing a mode of vocal control and a new construction of
territoriality. Perhaps this is why the comprehension of writing has been ignored
in the educational reform proposals until now.

Towa r d a n Andean T extual Th eory

The regional theory of textuality we develop in this book, on the basis of a criti-
cal reading of the work of Jacques Derrida, seeks to clarify the nature of textual
struggle (and the struggle for sovereignty) in the Andes since the Spanish Con-
quest, as founded in different ideas about writing and its status in the world. Der-
rida’s work was useful for its exploration of universal concepts about writing that
include both European and Andean variants. This allowed us to compare Andean
ideas based on cloth about being, knowing, and personhood with European ideas
about these same notions based on writing, at theoretical, methodological, and
practical levels.
Rejecting Saussure’s idea that alphabetic writing is a representation of the
voice, Derrida proposed that writing is rather a textual support onto which vocal-
ization and the play of meanings are developed as a later stage. For Derrida, Eu-
ropean writing is something inert onto which comprehension and interpretation
are “grafted” in the subsequent dynamics of vocalization. Svenbro, too, provides
a conceptual bridge for exploring the same question in the Andes in his reference
to the historical moment in ancient Greece when consciousness of the dynamic
play between inert writing and its vocalization still drew on weaving practice as a
point of reference, giving rise to the original Greek notion of “text,” whose ety-
mology derives from this woven and vocal dynamics.

The Origins of Writing in Europe and in the Andes

Throughout, we have been acutely aware that Derrida, Rousseau, and others
intuited the violent origins of European writing (for Derrida, the very imposition
of writing is a form of violence), although none of them clarified its nature. Now

274 A N D E A N T E X T S A N D T H E I R I N T E R P R E TAT I O N

we can suggest some criteria for understanding these violent origins of writing
from an Andean perspective.
We propose that the “written support” (the warp) of writing in the Andes is
something dead and inert, more specifically that the whole weaving apparatus
used at present derived originally from the act of tending the long hair of a cap-
tive trophy head over wooden stakes. We heard weavers themselves confirm that
the act of weaving is at once warlike and fertilizing. First the defeated enemy is
laid out, and then, through the act of weaving, the weaver breathes new life into
the textile body. The goal of weaving is that of creating a new being (jaqi) within
the loom rack, a baby (wawa) that begins to speak. The very act of weaving is a
constant dialogue with this new being.
The Andean textual theory we posit concerns a complex that passes from a
destructive cycle to a constructive one, in which the suffering of death gives rise
to the creativity of weaving and new life. At the roots of this theory is a trophy
head complex, a regional variant of the wider Amerindian phenomena of “on-
tological depredation” described by Viveiros de Castro. In essence, this theory
concerns the appropriation of the forces of the Other in order to generate the
creativity and continuity of the Self and one’s own. In the highland region of
Qaqachaka, this warfaring complex was a lived reality in the recent past, in the
fights over land and other resources between ayllu groups. As a result, comunarios
of the region still apply the social memory of capturing trophy heads in war and
converting them into a new harvest of babies to other creative tasks in a wider re-
gional symbolic economy.
We argue that this complex of ideas underlies local cultural practice in the
foundation of the household by the ideal couple of warrior and weaver, in mar-
riage and death customs and rites of birth and childhood, as well as in regional
cultural theories about being and knowing, personhood, the body, and so on.
The reproduction of this regional body of knowledge has ramifications in no-
tions of childhood, schooling, interculturality, and educational practice, whether
at a communal level, or administered by the state.
In this symbolic economy, a male warrior must fight enemies and bring home
trophy heads to his spouse. He must also transform the energies and luck em-
bodied in the seedlike head to generate wealth for his family. In terms of textual
practice, he has the creative task of toasting to attract luck through the didactic
guide of the kipu structure, whose knots are homologous to trophy heads and
threads to its flowing hair. In this case, his releasing of enemy strength coincides
with a flow of tributary relations between the community and the state, in which
each participant is subjected (head, hands, and body) to state interests in the per-
petuation of its own universe.
TO WA R D A N A N D E A N T E X T UA L T H E O R Y 275

Women as weavers are more concerned with the revivification of the dead
enemy. The weaver first immobilizes the alien forces of the captured trophy head
within a loom framework, reinforcing its imprisonment through the textile struc-
ture of borders and additional edgings. Subsequent weaving processes reiterate
the submission of the captured enemy, with a series of blows with a llama bone,
reinforced by additional punches. Then gradually, the weaver appropriates these
alien forces, converting them into fertilizing energies for herself and her family,
through the creation of a new being (jaqi or wawa) and its subsequent harvest of
new babies.
This same configuration of ideas orders how vocalization is understood in
an Andean context. In various rituals we described, a sacrifice during the prepa-
rations provides a kind of written support with which later vocalizations inter-
act. We mentioned in passing some comparative studies from other parts of the
world that postulate a similar link between voice and sacrifice, through dialogue.
It is still premature to confirm this point, but it would seem that communal rit-
ual activity reiterates most clearly the social memory of these violent acts of the
While these ideas confirm certain aspects of Derrida’s proposition, his ideas
must be adapted to take into account the local dynamics of Andean textuality, es-
pecially through gender relations. In present-day textiles in women’s hands, the
warp structure serves as the written support (an inert thing), while the dynamics
of reading occurs through a woven “vocalization” when the horizontal move-
ment of the weft opens “mouths” at the two sides of the advancing cloth. The
discursive practices of weavers, founded in the institutional authority of the Inka
“pathways of knowledge,” confirm this relationship when they refer to the verti-
cal warp (like the pendant kipu cords) as threads of hair tended over the loom.
The motion of the heddles forms another textile mouth that provides the circu-
lation of light and breath, food and blood, to the emerging textile body. In this
multivocal way, the textile gives up (transmits) its message. In present-day kipu
reading practices, in the hands of men, the same vocal dynamics of reading kipu
contents concerns the proper circulation of corporeal substances through its
threads and knots, propelled by male agency, resulting in the verbal transmission
of its message.
This regional constellation of ideas also sheds light on some of Derrida’s
metaphysical preoccupations. Derrida found at the basis of alphabetical writing
a nexus between European notions of being, Logos, and voice. Similarly, in the
Andes, we find at the basis of weaving a nexus between person and personhood
(jaqi, jaqichaña), the word (from Augustinian seminal thought), and voice (aru).
Derrida finds in European writing two opposed criteria: Being and Self (related
276 A N D E A N T E X T S A N D T H E I R I N T E R P R E TAT I O N

to the notion of totality), and the Other (related to the infinite). Similarly, in An-
dean “writing,” we find the same kind of duality. At the basis of textile genealogy
is the infiniteness of the Other (jupa, jupanaka) that draws on the social mem-
ory of multiple ancestral weavers (family members or enemies) incarnate in the
“written” support of the warp threads, originally the hair of these deceased, or
the writing-like “ancestral words” that the hair communicates. In contrast, the
new totality of the loom space provides a framework for the creation of a new
being (wawa or jaqi) that is part of one’s own (nanakana).

T e x tual A esthetics , V oice , and T erritory

In practice, these metaphysical bases of writing underlie the aesthetics of re-

gional textual and numerical experiences as something both simple and elegant.
Its very simplicity is what produces the sense of a textual clash when Andean
populations encounter alphabetic writing. This emerged in conversation when
Don Santiago Inka asserted that “libations unfold more easily, while the same
quantity of letters produces a whole lot of confusion.” This stems in part from
the nature of Andean texts, where the dynamics between an inert written sup-
port (whether textile or kipu threads, real or imaginary) and vocalization unrav-
els according to an iconic and institutionalized codification of information. For
example, by first naming the tambo way station in a series of toasts, you go on to
generate a maximum of interpretation from one single object.
Another key difference between current Andean and European writing is that
rural Andean textual dynamics begins with icons (in the kipu knots or weaving
figures) that express a unification of meaning, and then proceeds toward their
fragmentation (as Kusch intuited). In contrast, modern European writing prac-
tices tend toward an insistent linearity and fragmentation of reality as the domi-
nant tendencies of European thought. In the example of libations, they unfold in
a sequence of iconic knots, in which the first knots express totality, and then the
following iconic knots differentiate their parts. The sequence ends by rounding
off in another icon whose location at the end of the thread marks the transition
between ending and reconstruction, and then passes on to another thread to initi-
ate another series of toasts. This regional form of discourse is something similar
to Derrida’s idea of deconstruction as a necessary prior stage to reconstruction.
We postulate that its origins lie in the logic of sacrifice.
Evidently, the teaching and learning of Andean writing practices proceeded
and still proceed by logical steps, from lesser to greater units, in order to manage
sophisticated levels of discourse and analysis at an oral-textual level in the play be-
tween deconstruction and reconstruction. One of the repercussions of these fac-
TO WA R D A N A N D E A N T E X T UA L T H E O R Y 277

tors in the practice of the current Bolivian educational reform is the necessity to
reconsider the didactic sequences in the new teaching modules. Instead of experi-
menting with didactic sequences on the basis of communicative models of lan-
guage or constructivist pedagogy in vogue in other parts of the world, modified
reform materials might take into account the pupil’s cultural and textual back-
ground, as well as local didactic needs, in order to develop more acceptable and
effective didactic techniques in rural communities.1
This Andean theory of textuality, with its axiomatic explanation of the re-
gional nature of “text” (through the interlacing of voice with a written support)
also clarifies the elusive nexus between voice and territory. Andean metaphori-
cal language, as a vital part of linguistic and textual fields of knowledge, derives
directly from regional relations of production and reproduction, where textile
warp or kipu pendant threads are perceived in terms of furrows in the earth, and
letters in terms of seeds.
A vital function of children as pupils in these cycles of production and re-
production concerns the interlacing of the interstitial roots between the mascu-
line warp and the female weft to achieve the generation of letter-like seeds. This
function would seem to explain the preoccupation with a continuous vocaliza-
tion of recitation and memorization, under the children’s charge, practiced on an
everyday basis in almost all the schools of the Bolivian highlands. This is at odds
with written expression, which demands distance between the written voice and
the original speech, and the decontextualization of its own conditions of produc-
An Andean textual theory has other resonances in regional epistemology and
ontology, especially in notions regarding being and knowledge. We mentioned
lexical evidence from research by Itier, in which the root yä- (or yäna-), held in
common by the more general Andean terms for knowing, expresses a primor-
dial dynamics of occupying the limits of a given space in a creative and fertilizing
way, originally associated with the activities of the Andean God Viracocha. This
is corroborating evidence for a textual theory of managing knowledge about the
Other in a determined space (primordially in the loom or kipu layout), as a prior
stage to appropriating this knowledge in the generation of new beings of one’s
own group. In this context, it is surely not a coincidence that Viracocha was the
Andean God of waters and weaving.
We propose that the dynamics of child development occurs within this same
space of knowledge. Here, knowledge is developed within an enclosed space, and
then gradually extended outward beyond the body, the home, and the ayllu, into
the world beyond, all under the direction of the authoritative institutional setting
of the pathways of children’s learning. The initial containment of knowledge
278 A N D E A N T E X T S A N D T H E I R I N T E R P R E TAT I O N

within certain limits (of growth and abilities) happens in an internalized stage
during the construction of the natural person. Then these limits are purposely
broken in various rites of passage to attain a more externalized stage in which at-
tention is given to the development of the social person. In each case, the mode
of conceiving this process is homologous with the structuring of textile, borders,
and loom, and reiterates the process of forming a new being.
The archetypical stage of forming a social and juridical person (jaqi) occurs at
marriage, in the conformation of the ideal couple of weaver and warrior. There,
advice giving and sayings, with the authority of commandments, mold the spiri-
tual rebirth of the couple, in which they, as babies (wawa), appropriate the forces
of the Other (incarnate in the past in the baby metamorphosed from a captured
trophy head) and convert them into the new harvest of household babies as fam-
ily property, goods, and chattels.
All we have said about Andean writing practices, centered in the social sphere
and in the body, is equally pertinent in the context of numerical practices. An-
dean mathematics equally unfold according to the logical relations between a to-
tality and its parts. And implied in its present practice is some prior moment con-
cerning the counting of enemy trophy heads and their innumerable hairs, and of
appropriating their forces in a multiplying way, to replicate a new harvest of ba-
bies/goods through the powers and color of the voice.

T o w ard an A ndean I nterculturality

This same regional theory of textuality helps clarify the difference between
the universalist and mentalist stance at the basis of the interculturality proposed
by the current educational reform, and the more corporeal and social approach
of Andean interculturality.
We found that the school has always been conceived as an intercultural place
of encounter between Self and Other and between the community and the state.
In the past, the school formed part of the tributary networks between the com-
munity and the state, when children were sent to school as a part of an annual
sacrifice on the part of the community, an organic tribute still reiterated in the
ritual and ceremonial apparatus of social memory around the history of this phe-
nomenon, whose recompense is rights to land.
Paradoxically, each new configuration of interculturality draws on the roots
of this historical encounter. For its part, the modern state apparatus incorporates
rural children (through state education, official history, reading, and writing, and
the inscriptional practices of the nation) by symbolically transforming commu-
nal babies into trophy heads. For their part, the community reincorporates the
TO WA R D A N A N D E A N T E X T UA L T H E O R Y 279

present-day state into the culture of the place (in memories, symbols, and the
monumental time of former Andean states) by reappropriating the energies of
the ancestral dead, buried below the school rostrum, into a new fertilizing har-
vest of communal babies. This is done through their own interpretation of read-
ing and writing, and the criteria of regionally constituted textual practices. This
complex of ideas, centered in ritual practices based in the liquid medium of liba-
tions to the school, names the mountain peaks where the heads and blood of lo-
cal animals are offered, proportioning strength and health to the community and
its flocks, and insistently “making them grasp” these memories in the head. This
intense rituality reiterates historical precedence, when the wider jurisdiction of
ritual authority depended on the school’s insertion into the political, ritual, and
military hierarchy of former Andean states.
Within this ritual-semantic complex, the pupils, through school rituals and
daily classroom practice, help the spirit of the letters to sprout, to play their part
in regenerating the ayllu’s farming and herding production. Through their inser-
tion into the cycle of ontological depredation aimed at the metamorphosis of
heads, the children appropriate the generative strength of the Other (incarnate
in the ancestral skulls and bones of the school rostrum, called taqawa), and make
them bloom with new babies/flowers. All this came to be throughout centuries
of teaching Christian doctrine, in which Andean communities reinterpreted the
letters of the Sacred Scriptures as if they were enemies against which they had to
From this perspective, the rites of the nation in the hands of the commu-
nity (and schoolchildren) that we examined serve rather to celebrate their victory
over this enemy. From the community’s point of view, the site of the school, even
school rituals, form part of an Andean interculturality in which the comunarios
(and to a certain degree the teachers) still participate in the reproduction of the
nation, through the defeat of the enemy and its incorporation into their own tex-
tual basis.
These conclusions demand a review of the ethnography of schooling else-
where to look for similar trends relating to tribute and recompense. This revision
should include not only other parts of the Americas (especially the lowlands) but
a more universal sweep, since our study indicates semantic domains with much
more widespread resonances. One of these is the ritual ability and potency of
children (in particular of adolescent girls) in calling the rains, in a kind of men-
strual hut. Another is the function of schoolchildren as warrior apprentices suf-
fering the hard steps of a rite of passage in a kind of men’s house.
As a whole, the book raises more questions about the application of the new
reform interculturality than it can answer. This is because the local and global re-
280 A N D E A N T E X T S A N D T H E I R I N T E R P R E TAT I O N

lations of what we call Andean interculturality articulate community-state rela-

tions around the issue of rights to land. Communal oral history locates regional
interculturality in the social memory of the Inka period (whether in a real or ar-
chetypical way), and in the origins of the so-called Andean pact, a set of relations
of obligation and tribute on the part of communities, and recompense on the
part of the state, on the basis of usufruct to what are still regarded on ritual occa-
sions as the Inka’s lands.
Historically, schooling served as a key mediator of community-state relations,
and comunarios had the additional burden (cargo) of serving as duty holders in the
reproduction of the relations between the community and the state. We argued
that the social memory of this historical pact has compelled Andean rural popu-
lations to continue to view relations with the state in terms of obligation and trib-
ute, even with the modern mestizo-criollo state with its neoliberal Educational
Reform and Popular Participation laws.
We showed how some of the reasons for the present rejection of teaching
in native languages can be found at the heart of the community-state relations
founded in the Andean pact for lands. If schooling, in Andean history, has always
been a place for teaching and learning practices promulgated by the state (for ex-
ample, in the past, of Inka textuality and statecraft, and Quechua language), then
we should not be surprised that comunarios as parents transfer this perception to
the present-day Bolivian state, in their demand to learn Spanish and the textual
practices of the nation, namely reading and writing based on paper and ink. Only
in this way do they have the confidence of handling the state’s textual medium
and the language of the national bureaucracy, as a means of dialoguing (when-
ever necessary) with state functionaries in favor of their land rights or whatever.

C olonial Wrappings

New directions of research that might permit us to unravel further the vio-
lent origins of writing, in the Andes and the West, include the psychological re-
search about “external writing” (tattoos and so on), or the examples in the school
precincts when rites of inscription in the new writing practices are accompanied
by punishments centered on the skin (as the limit of the body), and in clothing.
Notwithstanding, this same complex of ideas clarifies a fundamental dynam-
ics of the colonizing process, which concerns absorbing the social body of the
Other in the textual wrapping of the Self. In this iconography of colonization, the
widespread practices of ontological depredation give us clues for understanding
this phenomenon at a tribal or ethnic level in the many cannibalistic rites around
identity and the Other. We also know that the Inkas expressed the submission of
TO WA R D A N A N D E A N T E X T UA L T H E O R Y 281

the multiethnic groups under their more imperial dominion through a mesh of
woven wrappings. In Europe, too, the proposals by Nebrija achieved domination
and submission at an imperial level through the wrapping of the New World ter-
ritories in the body of Christ and the language of the Spanish empire. The more
familiar works by Anderson, and Marvin and Ingle, provide clues for understand-
ing the iconography of colonization at the level of the nation-state since the nine-
teenth century, when the map of the world was covered with the pale pink “skin”
of the British empire.
The Spanish insistence, faced with Andean textual practices, on the quasi-
divine status of European reading and writing as the Word of God, was simply a
variant on the same colonizing theme, even more hypocritical for having forgot-
ten its origins in forms of violence in the Old World.
In this colonizing context, the Andean textual theory we posit might even of-
fer new ways of perceiving the set of theories and practices that we call “culture.”
We assert that a vital part of regional cultural activity, even of language itself, is
centered in the dynamics of deconstructing cultural artifacts (textiles, kipus) into
their constituent parts (fleece and lands, water and pastures), only to reconsti-
tute them, in the context of a communal consciousness of these origins, in the
way of sacrifice. Outside the immediate ayllu, at an intercultural level, this in-
volves the ongoing process of appropriating aspects of the Other and incorporat-
ing them into oneself. So decolonizing cultural practice would equally concern
the shedding of foreign wrappings and their collective reappropriation into new,
but barely imagined forms.

T h e P h e n o m e n o n o f C o n ta c t b e t w e e n T e x t ua l P r a c t i c e s

Such decolonizing acts would undo the driving impulse of the present edu-
cational reform and the design of the current teaching modules (in which new
ideas about being, knowing, and personhood are subordinated to reading and
writing and the book), which simply replicate philosophical and ideological no-
tions whose predecessors we encounter in the colonial context of indoctrination
on the basis of Scripture (beginning with the fatal encounter in Cajamarca), and
that were implemented in the promulgations (about acceptable texts and their
modes of standardization) of the Third Lima Council of 1583–1584.
We locate the reception of alphabetical writing in the Andes in the frame-
work of contact between textual practices with the European Conquest, in a field
of struggle between different texts and their meanings, between distinct modes
of interpretation and comprehension. An additional influence in the understand-
ing of Andean populations about writing emerged in the techniques of indoctri-
282 A N D E A N T E X T S A N D T H E I R I N T E R P R E TAT I O N

nation based on the seminal thought of St. Augustine. This phenomenon of con-
tact between textual practices is analogous to that of contact between languages.
We find there instances of transferences between textual practices analogous to
the phonological, morphosyntactic, and semantic transferences that occur be-
tween languages in contact.
From this theoretical basis, we identified a phenomenon of bi-textualism
on a par with bilingualism, in which maternal textual practices (TP1), whether
weaving, braiding, or kipu knotting, intervene in the interpretation of reading
and writing as a secondary textual practice (TP2), as maternal languages (L1) in-
tervene in learning a second language (L2). This is evident in rural communities
where traditional pedagogical norms promote at an early age a high level of de-
velopment in the abilities of spinning, weaving, and braiding, that then serves as
the textual bases for a whole set of other derivational practices (song, music, cho-
reography, and so on). In weaving as a maternal textual practice (TP1), we find
a level of manipulation, interpretation, and comprehension developed according
to the norms of different stages of learning that permits girls to achieve control
of the complexity of textual and numerical units. This knowledge is then applied
in the learning and interpretation of alphabetic writing and associated activities
(such as Western math) as TP2.
It is not sufficient to approach these levels of interpretation and comprehen-
sion according to Piaget’s universalist criteria at the level of the development of
the individual. Rather, the advanced development of abilities in maternal textual
practices that children have on entering the basic cycle of primary schooling may
clash with the official state norms for learning reading and writing. We underline
as an obvious factor to take into account the age at which both sets of practices
are put into effect (from age three onward in the case of weaving and from six
onward in the case of reading and writing), and that children’s preexistent textual
repertoire of weavings, as alternative writing practices, are among the most com-
plex in the world.
One way of analyzing the distinct levels of transference between textual prac-
tices in contact would be to identify equivalents of the textual units in each set
of practices. At a written level, these would be on the basis of letters and words,
and then in the morphosyntactic relations of the larger units of discourse (lines,
sentences, and paragraphs). At an oral level, these would be on the basis of the
gramma of orality (sound quality, rhythm, and versification). Transferences exist
also in oral and written discursive practices (for example, the comparative orga-
nization of rhetoric and oratory, the use of persuasion in arguing, gendered char-
acteristics of discourse). Other transferences occur between communication me-
dia (weaving on the basis of fleece, or reading and writing on the basis of paper
TO WA R D A N A N D E A N T E X T UA L T H E O R Y 283

and ink), in the writing conventions of handling and guarding resources and stor-
ing and accessing information, and in the semantic domains of meanings in con-
flict. This book examines only some of these levels of transference, and it would
require another study to follow the theme with greater depth, so the following
points are merely suggestive.

A ndean G r a m m a and Western G rammar

Starting from the arguments of Derrida, Brotherston, Mignolo, and others,

we postulated that the gramma of Andean textual practices would be quite dis-
tinct from formal grammars based in written registers.
For example, the influences and transferences that emerge from the phenom-
enon of contact between textual practices reveal a comparatively low level of
practical manipulation by Andean populations of the units of reading and writing
in relation with their own textual practices, but a relatively high level of concep-
tual manipulation, especially in the religious nexus litra-lutrina (“letter-doctrine”).
The regional textual community of apoderados and their secretarios operates in
this context, with their reading of written documents according to the local prac-
tices of guide and follower. Their “written support” for reading and writing is
still founded in orality and regional writing practices (above all those of the kipu)
rather than in alphabetic writing. This also happens in the conventions of inter-
generational transmission, when the contrapuntal voices in dialogue tracing the
ancestral genealogy of a text are compared to the sowing of letters in the local
lands, which these voices water to make them sprout, resulting in plentiful pro-
These points are significant for understanding the revival after 1952 of An-
dean recitative and memorization practices by rural teachers, when an initial
struggle for lands was transformed into a new educational struggle. This histori-
cal conjuncture explains how rural classrooms continue to follow regional didac-
tic practices centered in an oral gramma instead of a formal grammar based on
the written register. Therefore, it is not simply a question of modernizing these
practices in the classroom, but of understanding them in their context and dis-
cussing with their practitioners about the ways of adapting them, when neces-
sary, to present-day criteria concerning reading and writing.
284 A N D E A N T E X T S A N D T H E I R I N T E R P R E TAT I O N

L etters , S ound Quality , and the O vervocali z ation

of R eform A ymara

The phonological transferences that pass from an oral to a written level pro-
ducing “interferences” at the level of orthographic systems are well known, as are
the syntactic transferences in the comparative grammars of Aymara and Spanish
(or Quechua and Spanish) at oral and written levels. Less known are the transfer-
ences between the written register of the dominant Hispanic culture (modes of
punctuation, norms of writing and orthography, and so on) to oral tradition and
the writing practices in Andean languages as a result of using the materials of
educational reform.
In practice, on the basis of seminal thought, comunarios give great value to
the subsyllabic level of words, particularly the semantic aspects of the sonority
of letters. This phenomenon might have emerged from the colonial modes of
teaching reading and writing, but it also coincides with regional textual practices
of emphasizing vocalization.
Unfortunately now, the dominant written register adopted by the educational
reform for Andean languages has given rise to “written pronunciations” (when
pupils pronounce all the vowels in written Reform Aymara, which would nor-
mally be elided in speech) because of the exaggerated distance between the writ-
ten practices of the dominant culture and those of rural communities.2 We saw
how, in rural communities, the act of vocalizing sounds from a written support
has not only a practical but also a metaphysical function, since this very act re-
lates the speaker with a genealogy of antecedents, including the Inka himself. In
addition, each sound of “writing,” from an Andean perspective (reinforced in the
colonial period by Augustinian seminal thought), is thought of as a seed within
a wider productive sequence, whereby it has to be planted in its furrow, made to
grow, nourished, harvested, eaten, and finally digested, and hence incorporated
as the Other. Now if the act of pronouncing seedlike letters irrigates them and
makes them sprout, then if there were more letters, we might expect even more
effort to pronounce them all.
The particular manner in which a local reading of the new texts of the edu-
cational reform has emerged should not surprise us. The form of standardiza-
tion adopted for the writing of “Reform Aymara” has the vowels included at the
end of all words (including those vowels normally elided in speech). So, the tex-
tual practice adopted in highland schools has been to pronounce all the letters,
including the final vowels, in order to generate better the next harvest, in spite
of losing the morphosyntactic sense of discourse. As a result, present-day read-
ing of Aymara in the classroom has no sense, not even for the pupils themselves.
TO WA R D A N A N D E A N T E X T UA L T H E O R Y 285

We found the same phenomenon in the Teacher Training Institute in Caracollo

(Oruro), where students and teachers alike were beginning to read the new Ay-
mara textbooks in this way, pronouncing all the letters, in spite of losing the
wider aural comprehension of the text.
This metaphysical background to Andean pronunciations, centered in the
power of vocalizing words, reinforces the evidence presented by various schol-
ars that anterior forms of Andean writing (like other world writings, including
Arabic and Semitic scripts) only marked the consonants of words and not the
vowels.3 So ancient techniques of recitation, like present-day ones, in vocalizing
words “made them live.”
All of this demands rethinking the attitude of the educational reform techni-
cians toward the comprehension of what is read. From their perspective, it is as if
their alien literate culture is able to comprehend written texts while Andean pop-
ulations do not. But if the regional objective of recitation and the memorization
of letters was that of “requesting rains” to fertilize the earth, then the local com-
prehension of reading matter involves a different gamut of concerns—sounds,
seeds, and sprouting and so generating a vegetative cover.

T ransferences bet w een T e x tual C onventions :

T he R a w hide C hest and the R eading C orners

For historical reasons, other levels of transference occur in the confrontation

between regional conventions for handling and guarding local textual resources,
and of guarding written “texts” under state administration. This phenomenon,
equally centered in the question of lands, is in the hands of the local textual com-
munity of title bearers and their descendants and their conventions for handling
and storing information concerning local resources, and then passing this infor-
mation on to higher authorities in the state hierarchy. Evidently, these practices
based long ago in the handling of kipus and other Andean texts became adapted
when the Toledan Reforms put into effect new criteria for the management of lo-
cal resources. The new state objectives demanded that communities have in their
possession a communal chest locked by three keys.
In the past, the handling of this local documentation gave importance to the
genealogical ties of the sacred documents of parchment and paper. They were
not perceived as inert things (as in the West) but as living beings with strong
family ties. As ancestral remains, they were guarded (as weavings were) in fam-
ily chests of rawhide or wood, attitudes to which were molded according to the
same genealogical ideas, in spite of the state requirements to centralize them in
the communal chest.
286 A N D E A N T E X T S A N D T H E I R I N T E R P R E TAT I O N

Through the historical conjuncture between the title bearers and Andean ed-
ucators, these customs passed into the school precincts with the educational ex-
periments at the beginning of the twentieth century, and they have continued
up till now. This historical perspective explains problems in the implementation
of the reform’s reading corners or school libraries due to the regional authori-
ties’ reluctance to open the communal space and permit access for pupils to the
new written documents. It is not simply a matter of adapting the epistemologi-
cal space of schooling to include libraries and reading corners as the reform pro-
poses, but rather of understanding the history of these ideas, and then working
with communities to develop new and yet historically pertinent communal mod-
els of information interchange.
In this sense, present-day restrictions in the administration of the reform
book corners can be understood both in terms of ongoing acts of communal
resistance (of maintaining an ancient conception of the appropriate form of au-
thorized knowledge), and of ongoing fears of an alien presence, still not domes-
ticated enough under the more familiar practices of the place.4 The modern text-
books of the educational reform have not been incorporated sufficiently into the
curricular system of public instruction nor into the universe associated with na-
tionalism and the concept of citizenship. They are still relegated to the former
system of venerable manuscripts that belonged to a much more ancient social,
political, and intellectual order.
The same can be said of the regional textual dynamics in which the genealo-
gies of texts are based in families of mothers and offspring, or in hybrid charac-
ters that emerge from the juxtaposition of personages in colonial processes of
documentation, when the copying of texts is always carried out taking into ac-
count the transference of living forces from one text to another (through the use
of the verb waraqaña).
In each case, these already assimilated reference points could serve the devel-
opment of more appropriate writing terms for the school textbooks of the fu-
ture. New lexicon in the classroom could draw on the familiar terrain concerning
the transferences between communal and state communication media. And the
dynamics of regional textuality, with its focus on local familiar and historical ties
more than on documents per se, and its appreciation of the life, genealogy, and
corporeality of writing, could provide clues as to how to proceed with any new
translations of terms.
TO WA R D A N A N D E A N T E X T UA L T H E O R Y 287

New Struggles in the Andes

It was not by chance that the winds of change in the Bolivian October Revo-
lution of 2003 began in the rural teacher training college of Warisata. So, how
might community-state relations change in the future and what consequences in
educational and textual practice might ensue?
Our arguments underline how contemporary educational practices in rural
Andean communities derive not so much from a set of subaltern nostalgic grum-
bles from periodic peripheral engagements with a distant state center, but rather
from the social memory of a parallel alternative state. Certainly they draw in-
cessantly on the nostalgic memories of the last Andean state (the Inka), but the
intervening centuries have also witnessed dynamic processes of change and the
modernization of these practices within their own terms of reference.
With a majority indigenous population in Bolivia, many Andean practices
have penetrated the state apparatus in hybrid, intertextual, and intercultural ways.
The implementation of changes in the current educational reform is so difficult
to carry forward precisely because the present state educational system is pen-
etrated at all levels by regional textual practices, legitimate textual transmission,
and textual genealogy. The current result in practice is a system of cover-ups to
disguise the fact that Bolivia does not read, but the appropriation of educational
reform in indigenous hands might have other consequences.
We would ask if, before introducing new didactic and pedagogical techniques
from elsewhere, it would not be better to prepare beforehand a new “mental ar-
chitecture” at all levels (in those who finance the reform, its experts, the Minis-
try of Education, and teaching professionals in teacher training institutes and the
classroom) in order to develop other methods of teaching and learning according
to the regional historical experience of education.
This new mental architecture demands a radical change of focus. Instead of
the current debit-based models, which reject aspects of Andean rural languages
and cultures for their supposed limitations of expression and organization (a re-
jection founded in the inherently racist notion that ties language to race), greater
attention should be given to regional conventions in numerical, vocal, and or-
atorical writing and textual practices. This would be in line with current prac-
tice in the rest of the world, where a universalist focus toward the acquisition of
speech and writing in second languages has been replaced by attention to social
perspectives on literacy and “multiliteracies.” These new approaches take into ac-
count the mutual influences between cultural, cognitive, discursive, and commu-
nication practices, including the organization of written form (for example the
288 A N D E A N T E X T S A N D T H E I R I N T E R P R E TAT I O N

influence of the traditional Japanese rhetorical style of ki-shoo-ten-ketsu on Japa-

nese writing, or of Taoism and Confucianism on Chinese writing).5
As a first step, the design of the present reform modules might accommodate
the criteria already mentioned in the context of a situated learning that takes into
account multiple literary forms. Instead of using the teaching modules (and An-
dean languages) as vehicles for the imposition of new cultural, moral, and writ-
ing transferences from the dominant urban culture toward rural populations, the
modules could incorporate the more positive aspects of Andean languages and
cultures that have much to teach the dominant Hispanic culture of Bolivia. As
for languages, it is necessary to reconsider urgently the many advantages of the
verbal dynamics of Andean languages instead of blindly nominalizing them, as
well as the intertextual ways of understanding the expression of the textual and
esthetic modes of Andean languages, instead of subordinating all to alphabetic
reading and writing. Regarding cultural aspects, it is necessary to reconsider ap-
proaches to the corporeality of knowledge, forms of the socialization of learn-
ing, and local relations between age groups and genders.
As for gender relations in this complex of ideas, we emphasize two points.
First, the importance of women in handling textile language at a regional level
derives from a gendered division of labor, founded in the institutional author-
ity of the traditional pathways of knowledge, and the experts charged with their
implementation. Women have always been concerned with the production and
management of a community’s textual resources (above all cloth) and hence the
reproduction of its textual base, grounded in territory. So women were excluded
from the state educational system until recent decades in order to guarantee their
part in the reproduction of the community’s own textual resources, through their
labor in herding and hence their control over the use of fleece in weaving.
Secondly, blind modernization on the model of Western culture as promoted
by reform ideology and current materials design, with its trends toward what
Illich (1983b) has called “unisex,” seeks to change traditional gender relations
and with them, textual practice, while ignoring the institutional authority under-
lying both of these. The strong pressure on young women to choose between
the “two textual pathways” (of weaving and alphabetical reading and writing),
within the context of the present reform, is not making their lives any easier. The
tendency in reform materials to focus on reading and writing, and simply incor-
porate Andean textual practices into an alphabetic framework seems particularly
mistaken, considering the limited work possibilities and access to written texts of
these young women in the future.
From the point of view of the community, the emerging problems of school-
TO WA R D A N A N D E A N T E X T UA L T H E O R Y 289

ing concern gaining ground in educational spaces. For historical reasons, com-
munities have not wanted to intervene much in the day-to-day running of the
schools and have left these activities to teachers and the departmental and state
hierarchies of the Ministry of Education. But the demands of the recent educa-
tional reform, as well as the more formal recognition of popular participation in
schooling as a result of the Popular Participation Law, have compelled parents
and local authorities alike to inform themselves of other schooling activities, par-
ticularly the curricular contents.
The current juridical framework of schooling, formerly limited to the level
of local knowledge as the basis for historical land claims, is currently expanding
in an ever-accelerating process of political change. In the framework of globaliza-
tion, the new widening of boundaries and debates is creating more far-reaching
claims for cultural and linguistic identity, even of territorial autonomy, through
the medium of international documents such as Convention 169 of the Interna-
tional Labor Organization (ratified by Bolivia in 1991), the Draft Universal Dec-
laration of Linguistic Rights (1996), the Aymara Manifesto of Jach'ak'achi (2000),
the Declaration of Lima (Peru, 2002), the Use of Multilinguism in Cyberspace
(2000), and various draft resolutions of the United Nations that include proposals
for the educational rights of indigenous peoples.6 However, these new directions
oriented toward a more authentic regional education program, together with
the naming of positions responsible for its implementation, have still to be estab-
lished legally at a national level.
Many new proposals from Andean communities and their leaders seek to
explore lived reality as opposed to subordinating their textual practices to the
criteria of Western reading and writing and the normative writing practices of
bureaucratic translation. This lived orality emerges from Andean territory and
textual practice such that textile and song, braiding and music, narrative and oral
history, are all part of a common pathway, with its own norms of learning, ex-
pression, and performance. We mention here ways of appreciating Andean tales
(Jukumari the Bear, and The Fox and the Condor) as didactic resources to be re-
cited at the proper time in the year.
Other curricular proposals from the Aymara leader, the Mallku Felipe Quispe,
demand that the existing textual repertory of schooling expands to include infor-
mation technology and communications (ITC). This would involve the reincor-
poration into the curriculum of the visual and semiotic languages of weavings
and kipus, whose codes of organization and conceptualization facilitate concep-
tual bridges with the languages of the World Wide Web, for example HTML
(Hyper Text Markup Language). From this perspective, a regional lexicon deal-
290 A N D E A N T E X T S A N D T H E I R I N T E R P R E TAT I O N

ing with the dynamic processes of creation can be developed, whether on rustic
looms in rural areas, or the computerized Jacquard looms of Andean urban cen-
Databases could be designed around three-dimensional weaving structures
and games could be developed to teach programming to Andean children based
in the logical steps of weaving.7 This alternative approach to regional textual
practices also demands the development of new spaces of research and debate
(about philosophy, metaphysics, mathematics, religion, psychology, pedagogy,
linguistics, philology, ecology) in a regional rethinking that rejects the positivist
and mechanical models of the nineteenth century in favor of processual dynam-
ics. For those who are part of this radical rethinking, rather than facilitating the
instruments of an alien modernization, we offer first of all a critical reassessment
of the Andean textual past.
N ot e s

1. On the linguistic model of languages in contact, see the classic studies of Lado (1957),
Weinrich (1968), and James (1980).
2. See Selinker (1992) for definitions of “interlanguage,” Escobar (1978, 30) on “interlects,”
and Yraola (1995, 185–88) on the “semilanguage” of rural schoolteachers. See also Appel and
Muyksen (1987, 83) on the continuum hypothesis.
3. Cf. Ong (1988), Mignolo (1994a), and Kersenboom (1995). Mignolo (1994a, 236) notes that,
historically, the term “text” was adopted by Roman rhetoricians to describe the texture of written
4. See chapter 2 of Brotherston’s Book of the Fourth World (1992) for other implications of this
5. Cf. Platt (1992).
6. On Red English, see Leap (1993).
7. On Andean Spanish, see Zavala C. (1996) and Yapita (1996).
8. See for example Hornberger (1997) and Arnold, Yapita, and Sachdev (2001).
9. Cf. Street (1993, 1999).
10. See, for example, Cook-Gumpertz and Keller-Cohen (1993), Street (1999, 9).
11. See, for example, Kress and van Leeuwen (1996, 2001), Heller and Pomeroy (1997).
12. For example, see Villavicencio (1990) and Urton (1997a).
13. On the parameters of this debate, see, for example, Rappaport (1992a), Stock (1983,
1990), and Street (1984).
14. See, for example, Tedlock and Tylor (1987), Rösing (1995).
15. See Arnold (1992a, 1997e), and Arnold and Yapita (1999b).
16. Albó (1995, vol. 1, p. 20).
17. See, for example, Arnold (1988, 1992c, 1993), Arnold and Yapita (1996a and b, 2001), Aber-
crombie (1998), CEDPAN (1995, 1996, 1997a and b).
18. See Albó (1995, vol. 3, maps).
19. For a more detailed analysis of Aymara and Spanish terminology, consult Arnold and
Yapita 2000. See Hardman et al. (1988) on Aymara grammar in general, and Briggs (1993) on so-
cial and regional variations in Aymara vocabulary and pronunciation.

Chapter 1: Andean Textual Polity

1. For a general overview, see Murra (1965, 1989).
2. See for example Palacios Ríos (1977, 157).
3. We borrow the term “network production” from Hardt and Negri (2001, 294–97).
4. See for example Zorn (1987), Torrico (1989).
5. See Arnold (1997c, 114).

292 N O T E S T O PA G E S 2 2 – 3 0

6. Cf. Frazer (1994).

7. The relation between textile and local production is described in Silverman Proust (1994),
Arnold (1997c), and Arnold, Yapita, and Apaza M. (1996); between textile and Andean ecology,
and the social organization of space and time, in Harris (1985); and between textile and flora,
fauna, and avifauna, in Peters (1991) and Arnold (1997c), for example.
8. On weavings as contemporary maps, see Arnold (1997c); on the radiating pathways called
ceques around Inkaic Cusco, see Bauer (1998).
9. On the pathways of the gods, see Arnold (1993, 67–71). On the relation between textile,
choreography, and music, see for example Van Kessel (1992), Arnold (1992b).
10. On the ties between weaving and the celestial pathways in the Southern Andes, see Ar-
nold and Yapita (2001, chaps. 5 and 6); see also Silverman Proust (1994) for the region of Q'eros
11. On the contemporary division of labor in weaving and herding activities, in Qaqachaka
in the Southern Andes, see River of Fleece, River of Song (Arnold and Yapita 2001), also see Arnold
and Yapita (1996a).
12. For the Andes, see Cereceda (1978), Desrosiers (1982), and Arnold and Yapita (2001, chap.
2), among others; on similar criteria in painted designs on cloth in Amazonia see Gow (1999).
13. See Arnold, Yapita, and Tito (1999, 50–51).
14. See Rowe’s essay on Chavín art (1980, 313–17). See again Frame (1986) and the commen-
tary on Frame’s essay in Arnold (1994).
15. On the use of tripartition in Aymara, see Arnold and Yapita (1992, 34) and Huanca (1989,
132). On indirect reported speech in Aymara, Quechua, and Mapuche, see Adelaar (1990). On
“intertwined sentences” (k'anata in Aymara) see Huanca (1989, 141); on “hidden sentences” (taypi
ch'anka), see Yapita (1992, 52).
16. On the dialogic nature of Andean discourse, see López G. (1998, 39–42) and Arnold
(forthcoming a); on evidentiality in Aymara tales, see Dedenbach-Salazar (1998, 1999); on “open-
ing and closing illustrative topics,” see Huanca (1989, 136–37); on woven connectors between sto-
ries, see Arnold and Yapita (1992, 9).
17. See Arnold and Yapita (1998, 553).
18. See for example Conklin (1990).
19. On Inka kipus as counting systems in state administration, see the early studies by Locke
(1978), Ascher and Ascher (1981), and the essays in Ravines (1978), Radicati Di Primeglio (1979),
and Mackey et al. (1990). On the direction of spin and lay of knots, see Urton (1994); on kipu or-
dering, see Murra (1975); for a phonetic reading of kipus, see Pärssinen (1992, chap. 1), Arnold
and Yapita (1992), and Urton (1997b); on the connection between voice and territory, see Salo-
mon (1997) and Pimentel (1998).
20. See Silverman Proust (1994, 176).
21. On sphota, see Kristeva (1989, 85); on chora, see Kristeva (1984, 25–26).
22. See for example Ong (1988), Goody (1985), and Goody and Watt (1963), and the ample
criticism of these by Halverson (1992).
23. On the social and cultural context of their production, see the essays in Howard (1997a);
on the major and lesser units of their organization in libations poetics see Abercrombie (1998);
and in song, see Mannheim (1986), Arnold (1995), and Arnold and Yapita (2001, chaps. 4 and 7;
24. An expansion of this analysis can be found in Arnold (1997c) and Arnold and Yapita (2001,
chap. 6).
25. Arnold and Yapita (2001, chap. 5).
26. See especially Cereceda (1978) and Torrico (1989).
27. For more details see Bubba (1997).
28. On cellular automata, see for example Burks (1968), Frazer and Connor (1979).
29. On the animating character of breath and seed in Andean songmaking, see Arnold, Ji-
ménez, and Yapita (1991, 143–45, 164–65), and Arnold and Yapita (2001, 162–63). For comparative
N O T E S T O PA G E S 3 1 – 4 3 293

material on the power of words, see the classic texts of Malinowski (1935) and Tambiah (1968);
on the animating character of breath, see Butt Colson (1956) and Guss (1986, 423) among others.
30. The Chachapoyas comment is personal communication from R. T. Zui