The manuscript as testament

The Huarochirí manuscript alone of all colonial sources records a prehispanic religious tradition of the Andes in an Andean language. It tells us of a remote age when cannibal deities preyed on otherwise immortal humans, of the mountain deity Paria Caca who emerged to expel the fire deities of antiquity, of the human groups that traced their victories from Paria Caca's five simultaneous avatars, of Paria Caca's brotherhood with the fivefold female power Chaupi Ñamca, and of the society ritually organized in their names around a grand complementarity of male and female superhumans. It unfolds the splendor of ceremonies that prehispanic priests devoted to a landscape alive with the diverse sacred beings called huacas. It recalls memories of Inca rule and of how unknown invaders, the Spanish, brought new gods to displace the children of Paria Caca and of Chaupi Ñamca. Nothing else in all the sources from which we seek the Andean "vision of the vanquished" (Wachtel 1971) rivals it for immediacy, strangeness, and beauty.

But the voices we hear in its pages do not relay to us a verbatim record of what was said and believed before the Spanish invasion. It is true that when Father Francisco de Avila reworked part of the same or similar testimony to make his 1608 Treatise on the False Gods ... (hereafter referred to as Tratado), he judged that the narrative "does not refer to the present but to history" (Arguedas and Duviols 1966: 198). Yet the way people recalled their ancient tradition and the occasion of their recalling it were themselves facets of a colonial situation the tellers had already endured throughout their whole lives. The telling could not but be influenced by the seventy preceding years of colonial turbulence, during which one potent innovation was the art of writing itself. Andean peoples used no writing before the Spanish "Vira Cochas" arrived. So the process of capturing their culture as text in the alphabet of the padres and bureaucrats was inextricably bound up with forced conversion and persecution, even when the actual authors were themselves Andean and the actual narrators at least partly faithful to the old huacas.

The manuscript is a complex composite testimony of these changes as well as a compendium of ancient memories. The research it contains was apparently sponsored by a clerical persecutor, Father Francisco de Avila, who seems to have used it as secret intelligence for his assault on American deities from 1608 onward. The text does contain opportune denunciations of "idols" (as the Spanish called the sacred beings of the Andes) and of those who steadfastly fed and served them in secret long after official conversion. Yet at least one of the actual makers of the text seems to have thought of the task as one of historical remembrance. The untitled preface to the manuscript looks to a future in which the ancient deities would be remembered with pride, promising a monument of Andean greatness to match Spanish chronicles:

If the ancestors of the people called Indians had known writing in earlier times, then the lives they lived would not have faded from view until now.

As the mighty past of the Spanish Vira Cochas is visible until now, so, too, would theirs be.

But since things are as they are, and since nothing has been written until now,

I set forth here the lives of the ancestors of the Huaro Cheri people, who all descend from one forefather;

What faith they held, how they live until now, those things and more.

Village by village it will all be written down: how they lived from their dawning age onward.

One gets a strong impression that the creator of these lines was engaged in reconceptualizing the Andean mythic tradition rather than destroying its memory.

The exact process of composition is unknown, but this passage differs from the wholeheartedly anti-Andean viewpoint that Avila expressed in other writings. It may contain the words of a native writer or editor to whom Avila gave some leeway in compiling the text. A measure of unselfconscious candor would have increased its intelligence value. Whether or not he was present at its composition, Avila did read and annotate at least part of it; his devastating subsequent attacks on the deities mentioned suggest that the stratagem of leaving the witnesses some freedom of expression succeeded. But the text's partly intra-Andean genesis also had a paradoxical long-term effect: because it was composed in relative independence from Spanish preconceptions about native religion, it has in the end provided a uniquely authentic monument of the very beliefs Avila meant to destroy.

Whoever composed the untitled preface thought of the manuscript as a totalizing book about inherited tradition, custom, and lifeways that would give Andean memory, like Spanish literate memory, immortal visibility. The usual genre term by which the text identifies the separately remembered and narrated traditions could hardly be more oral; it is simi, which the greatest Quechua lexicographer of the age glossed as "mouth, language, commandment, law, mouthful, news, the word and its answer" (Gonçález Holguín [1608] 1952: 326). Clearly the testimonies are products of a culture in which orality encompassed the weightiest functions of language. But the book is not conceived simply as a body of speech on paper. It partakes of the assumption that written language, and specifically book language, should subsume and subordinate orality. The conception of a totalizing book that underlies the manuscript seems to be influenced at one or more levels by the Hebrew or Old Testament Bible and to some extent by the New Testament. Of course, few Indians studied the Vulgate. But the Huarochirí area had been missionized with special intensity by Jesuits in 1570-1571, with the conscious intent of popularizing Christian lore in Quechua. In the late sixteenth century, both officially promulgated catechetical summaries and popularized summaries of Bible stories called historias sagradas were widely read by or read to laypeople. Literate Indians c. 1600 usually knew traditions from both Hebrew antiquity and the New Testament through publications of the Third Council of Lima. Those who had access to churchmen's libraries or discussions could learn much more.

Although the manifest content of the manuscript only rarely syncretizes biblical material with Andean, the text as a whole has an "astonishingly Biblical" overall architecture (Turner 1988: 249). Like the Bible, the manuscript begins with myths that contrast the human condition with an imagined alternative, a time when the relations between humans and deity were radically different (chaps. 1, 2). A flood myth (chap. 3) signals the end of this era. Like the Bible, the manuscript pictures antiquity as the story of hero-ancestors who share a common descent and a covenantlike relation to an ethnic deity (including an episode resembling Abraham's averted sacrifice of Isaac; chap. 8, secs. 99-103). Its collective subject is a set of groups, each of which considered itself the progeny of a focalized ancestor. As with the biblical tribes, these groups relate to each other, at least in ideology, approximately as a phratry. As in the biblical redactions, their disparate traditions of origin and separate cults have been welded ex post facto onto the unifying argument of kinship and imperfectly articulated with apical priestly cults. Their story, like that of the biblical tribes, is intensely concerned with control over specific resources in a sacralized landscape; many of its myths encode political struggles with surrounding peoples and even internecine struggles as mythic combats with superhuman intervention. Also like the Bible, the manuscript is greatly concerned with the relation between the local sacra and the leaders and priests of immense invading empires—first the Inca, later the Spanish. The manuscript shares the biblical tendency to accrete genre within genre. Texts about priesthood, sacrifice, ritual law, and prophecy jostle with vernacular myth, claims concerning land and water, and mythicized remembrance of historic events. Bits of oracular response, religious formulas, and perhaps songs have become embedded in the text, too. And finally, as in the Bible, particularly the Deuteronomic books and later prophets, one clearly senses the pressure of contemporary political defeat on religious testimony.

We do not know all the reasons for the resemblances. One possibility is that Father Avila imposed European opinions on the text itself or indirectly on others who processed it (for example, by preparing a questionnaire or by overseeing the editing). Much European opinion of the time held that pagan myths, Andean ones included, reflected ancient traces of "true" (that is, biblical) religious and historical knowledge, which Satan's deceptions had distorted in intervening centuries. The tendency to force non-Christian testimony into patterns congruent with "universal history" and a unified Bible-based chronology is conspicuous in many Peruvian chronicles, both indigenous and Spanish, and Avila's Tratado shows that he partook of it (Arguedas and Duviols 1966: 206-208). The person who arranged or edited the myths expressed frustration at the difficulty of arranging episodes into one scheme of chronology (e.g., chap. 14, sec. 189; chap. 15, sec. 199), a step required for correlation with Bible-centered history. This literal-minded historicist reading of myth, which seems misleading to modern readers, was then thought to be a correct way of restoring American data to their "true" place in a unified world scheme of salvation history.

But above and beyond this exogenous process, the myths themselves seem Bible-like in their style of mythifying. Like the biblical writers, and unlike some myth-tellers from more "tribal" Amazonian societies, the Huarochirí narrators tend to intertwine mythic (miraculous) processes with social causation, rather than locating them in a primordial age before the world began to be as it is. Perhaps the likeness is multilayered or overdetermined: it may result in part from intea-Andean facts distinguishable from the European influences that also affected it. Such forms of synthesis may arise endogenously in societies of a certain scale, setting, and organizational form. Terence Turner (1988) offers a complex argument that a generically rather than locally biblical type of mythology occurs in societies whose status is intermediate between autonomy and complete subsumption in larger states; certainly this was the condition of Huarochirí-area societies for many centuries before the European invasion.

There is a third possibility for explaining the biblical parallel. After seven decades of exposure to European culture, Andean people had consciously or unconsciously gone far in reconceptualizing their mythology as a systematic response to imposed belief. By 1600 this reconceptualization seems to have coalesced into a distinctive ideology. Andean literati of the first generation born after conquest—Felipe Guaman Poma de Ayala and Joan de Santa Cruz Pachacuti Yamqui Salcamaygua—were not simple nativists; they partook of a Renaissance consensus in arguing that Andean people had already passed through ages of antiquity strictly parallel to those of pre-Christian biblical antiquity. But they dissented from the Spanish in their evaluation of the Andean achievement as a part of it. Where writers like Avila, Cabello Valboa, or Antonio de Calancha saw in Andean myth only a deteriorated and diabolically confused memory of original connections with biblical humanity, and therefore a culture worthy of being forgotten, some native intellectuals believed their history and its memory to be not only parallel with that of the Spanish, but equal in value. The theories of these bicultural "native chroniclers" shored up waning hopes of Andean privilege under Christian rule and appear characteristic of Andean natives descended from noble families but deprived of colonial power. Although we do not know the identity of the persons) who selected the oral material for inclusion and/or wrote the ethnographic and editorial material in the manuscript, the final redaction of the text does seem to partake of this mentality.

And what of those who actually told the myths to the book creators? What revisions of religious thought had occurred among less bicultural natives during these decades? It is important to remember that by the date when the manuscript was written the cults of the huacas had coexisted with Christianity for a whole lifetime. If huaca priests had retained the loyalty of people officially bound to Christianity, it was in all likelihood because they had succeeded, under the adverse conditions of clandestinity and church hegemony, in presenting huaca religion as comparable in cogency with the church's teachings. It is possible that by 1600 local thinkers and perhaps priests had been engaged (consciously or not) in remobilizing and reconceptualizing the inheritance of huaca religion so as to construe it as a religion, a "faith" (as the preface to the manuscript says, using the Spanish word) whose overall claims and dimensions could bear comparison with those of the imposed church. It is not beyond possibility that the welding of the Andean deities into a unified kindred partook of post-1532 efforts. Individual huaca myths seem to accord the huaca cults many of the same attributes as Christian religion: for example, a covenantal concept of obligation, an image of superhuman action as law giving, a notion of history as the continuing interaction of deity and society, and a tendency to express "moral economy" norms in terms of prophetic action. As has been suggested, it is likely that any or all of these may be overdetermined facts, arising, from preexisting, and now remobilized, prototypes in aboriginal culture as well as from European models. Perhaps there would have been an Andean story like the rescue of Isaac even if Spaniards had never invaded. Nonetheless, it would be unrealistic not to consider apologetic processes arising in huaca priests' efforts to match Catholic priests in the breadth of their claims while at the same time maintaining distinctness from Catholicism. The importance of such interactive processes in sustaining huaca religion is attested by Father Avila's remark that the huaca cults thrived most not in the areas where they had remained unmolested, but in the villages where Catholic priests had been most zealous (Arguedas and Duviols 1966: 205).

Despite the importance of all these factors, in the end nothing could be more wrong than to think of the manuscript as merely an Andean counter-Bible. For one thing, obviously, the mythic material overall is radically foreign to Europe; few books in the world give the Western reader such a powerful sense of encountering a cultural unknown. Another and more fundamental reason is that the structuring of myth—the formal architecture of event and process that gives each story internal regularity and resolution—owes everything to Andean patterns and resembles biblical ones little if at all. The dominant model in the stories is that of passage from mere difference (for example, the juxtaposition of antagonistic deities strange to each other) to complementary difference (for example, a revised juxtaposition in which the deities become male and female spouses or siblings embodying opposite ecological principles). This pattern occurs at the greatest and smallest levels of the mythology, in domains from the cults of apical deities Paria Caca and Chaupi Ñamca to the household relationship between inlaws. To imagine this pattern consistently applied to the battles of biblical Adonai is difficult. Comparison with non-Andean South American material (a task scarcely begun) may offer another path to the isolation of underlying prehispanic content.

R. Tom Zuidema (1977: 44-47) argues that some Huarochirí myths share specific structures with a myth of the Brazilian Bororo, presumably because the two mythologies share roots much older than Spanish domination.

Andean religion and "Inca religion"

Much of what is published (especially in English) about prehispanic and colonial Andean religion treats the terms "Inca" and "Andean" as near synonyms. Vira Cocha, popularized as the invisible "creator god of the Incas" (Demarest 1981; Pease 1986; Rowe 1960; Szeminski 1985b; Urbano 1981) fascinated colonial Spaniards who thought they detected in him a possible intuition via "natural religion" (MacCormack 1985) of Christendom's supreme deity. This idea still absorbs modern scholars captivated by the sophistication of Tawantinsuyu, the Inca Empire. The Incas partly persuaded nonInca Andean people, too. As we learn from chapters 18, 19, 20, 22, and 23, the Incas tried to reorganize local cults into a hierarchy capped by Inca numina, and partly succeeded. Indeed, their persuasions lasted longer than their sovereignty. By the midcolonial era, when Inca rule had receded into the golden mists of ideological nostalgia, many Peruvian Indians themselves came to recall the deified Sun and his incarnation the Inca god-king as compelling symbols of native identity and native glory (Flores Galindo 1987).

But in order to interpret the Huarochirí manuscript one must appreciate that the equation between Inca religion and Andean religion is an ideological sleight. The invisible Vira Cocha relates to Andean religious life somewhat as the Prime Mover Unmoved does to Mediterranean saint cults. At the summit of priestly and imperial society, prayers like Pachacuti Yamqui's stirring Inca invocations ([1613] 1968: 287-288, 292, 294) to the unseen source of order and beauty may have voiced metaphysical questions that we define as preeminently religious. But the religious life of most of the people who made up Tawantinsuyu's innumerable subject "nations" had little to do with abstract or universalizing expressions. Worship usually focused on sacred beings peculiar to particular kin groups, villages, mountains, canals, and so forth. In fact, religious particularism, expressed in terms of place and descent, lies at the heart of much Andean myth. No doubt discourse of this sort can embody philosophical concerns, no less readily than overtly metaphysical expressions do. But the content is only available by a route that leads through the study of what particular places or mummies (etc.) meant.

This village-based, particularistic version of Andean religious thinking saturates the Huarochirí text. The world that the Huarochirí myth-tellers imagined was structured in terms of grass-roots geography and of genealogy—their pastures and valley lands, their mythicized family tree. The unity of the text, such as it is, is achieved by an attempt (perhaps on the part of the priests called yancas) to locate the historically diverse huacas and their cults in Paria Caca's and Chaupi Ñamca's regional hierarchy and genealogy. Other traditions—the lordly priesthood of the Incas, the onset of Catholicism—are seen through the filter of such local and regional concerns. For long stretches the viewpoint belongs to one group, one collective ego: a group called the Checa, devotees of the ceremonial center Llacsa Tambo, resident (at least nominally) in and around the Spanish resettlement village of San Damián, who considered themselves children of Paria Caca while retaining origin myths apparently separate from his cult (chap. 24). Some of the ritual complexes and myths attached to local features—especially to springs, lakes, and canals—have survived with great vitality into modern times and have been studied by ethnographers (Gelles 1984; Ortiz Rescaniere 1977). For example, the modern descendants of the Concha, who live in an outlying hamlet of San Damián de Checa, still maintain today both the myth and the ritual attached to the lake that feeds their irrigation canals (see chap. 31).

The durability of these myths reflects the Huarochirí people's tenacious attachment to the local resources on which they depend. But it is possible after all to exaggerate the local quality of the myths. Paria Caca was not uniquely the deity of the peoples who speak here; he and his sanctuary were renowned throughout a wide swath of the central and southern Andes. It is likely that, when worshipers from Llacsa Tambo went on pilgrimage to him, they met worshipers from many other places and that their own practice had something in common with that of different kinds of "people called Indians." The same applies to the great coastal shrine Pacha Camac and to Cuni Raya, sometimes called Vira Cocha. Even when the names, episodes, and personalities seem peculiar to Huarochirí, the tellers' general religious concepts (e.g., classes of shrines, types of action attributed to deities and heroes, duties of humans to huacas) are shared among a wider spectrum of Andean societies, including, for example, groups in the Arequipa and Cuzco areas. In these limited senses, while it is mistaken to take the Huarochirí myths as expressions of a pan-Andean religion, one may take them as representative of broader Andean cultural premises and tendencies that are manifest even in apical Inca cults. The sharing of underlying concepts makes possible ethnographic comparison with societies beyond the bounds of Huarochirí Province, both as seen in past times (for example, via the "extirpation of idolatry" trials that postdate the manuscript; see Duviols 1986) and as witnessed by modern ethnographers (see, for example, Valderrama and Escalante 1988, who have studied in distant Arequipa a complex resembling the water cults in chapter 31).

General outline of the Huarochirí manuscript

The Huarochirí manuscript appears not to be the product of polished editing, but neither is it a jumble. With the possible exception of the two unnumbered chapters here called supplements I and II, it seems to be a fair copy edited and in the process of further editing for coherence as a unified narrative. The unification is, however, in many respects incomplete and imperfect. The editor's original intent seems to have been to treat ancient matters earlier in the text and recent ones later, but sometimes when turning to a new source (e.g., at the beginnings of chaps. 13, 24, and 31) he is forced to return to a different origin story. Recollections of past ritual practice and interpolated bits of current ethnographic observation further complicate the text by introducing into many narrations, with specious smoothness, references to times other than the time of the main narrated story. Moreover, the manuscript is full of second thoughts (cross-outs and interlineations, marginalia), tangents, overlaps, cross-references, marginal queries (probably by Father Avila), and cryptic allusions. For all these reasons, to appreciate the coherence of a theme one often must pull together partial accounts from disparate chapters. For this purpose the index supplied by the translators may be useful.

Early times and peoples

The preface (untitled in the original) promises to tell the achievements and beliefs of "the people called Indians" from their "dawning age" up to the present, village by village.

The first chapter sketches the world as it was before the present human race appeared. People lived forever (after a five-day temporary death), at the price of sacrificing half their children to the fire-monster Huallallo Caruincho. At that time, too, the subtropical abundance of the lower valleys extended far up into the heights. This whole rich and cruel world order would be destroyed with the advent of Huaroiri's great deity Paria Caca.

Huallallo's dominion is grouped with other stories of remote antiquity: chapter 2 tells how the Trickster-demiurge Cuni Raya, who "almost matches" the figure of Vira Cocha, passed through the landscape and through the lives of the female deities he seduced. In a Christian-influenced interlude (chap. 15), Cuni Raya, as Vira Cocha, is credited, almost parenthetically, with originally creating nature in an empty universe. Chapter 3, telling the myth of the deluge, and chapter 4, telling of the Sun's disappearance, end the section dedicated to remote antiquity.

The Paria Caca cycle and the myths of group identity

The mythic cycle that forms the unifying core of the text tells the apparition of Paria Caca (chap. 1, sec. 6; chap. 5, secs. 72-73), the fivefold deity who symbolized inclusive ethnic unity among the tellers' various residential and kinship groups. Like many great Andean deities, he is a mountain, a majestic double-peaked snowcap visible on the eastward horizon from the heights of Huarochirí. Many groups venerated him. We hear most about the Checa who gathered at Llacsa Tambo, but chapter 13 concentrates on a relatively distant cluster of villages, the Mama region in the lower Rímac valley. Chapter 30 is a myth of the Allauca, and chapter 31 copiously recounts the viewpoint of the Checa's neighbors, the Concha.

Paria Caca first appeared as five eggs that became five falcons that became five men, the founders of the human groups who appear as the main collective protagonists. These groups are sometimes spoken of in ways suggesting clanlike or sib-like qualities and together, within Paria Caca's cultic organization, are seen as a phratrylike collectivity of groups who "all descend from one forefather" (pref., sec. 2). But before calling them clans one should be careful to note that the text uses no term clearly translatable as 'clan' and that, although the manuscript uses the unilineal idiom for certain limited purposes, other principles of descent, marriage, and residence appear to play at least as important a role in defining groups.

Paria Caca was a stormy being of the heights. He made himself known to humans by favoring a poor man with power to defeat the rich (chap. 5), or, in another tradition, by saving a Huallallo worshiper from the obligation to sacrifice his child (chap. 8). He first appeared on a mountain in what was formerly the domain of the lowland aborigines called Yunca, which was dominated by the (possibly Huanca) deity Huallallo (chap. 5, secs. 38, 55). Once his bond with some of these people was formed, Paria Caca ascended to Huallallo's seat on the high cordillera. There he attacked and expelled the ancient cannibal deity in a world-shaking combat between storm water and volcanic fire (chap. 6, sec. 74; chap. 8; chap. 9, sec. 110; chap. 16, secs. 203-209; chap. 17, secs. 214-219). He carved out in this struggle the titanic landscape of snowcaps and lakes that is his seat and his likeness. It became the sanctuary on the icy heights to which his pilgrims brought their llamas (chap. 9, secs. 119, 127, 138-139; chap. 24, secs. 309-310; chap. 28, sec. 364). Huallallo, exiled, was left to the care of a rival ethnic group called the "dog-eating Huanca" (chap. 9, secs. 110-111; chap. 16, sec. 209). The victorious Paria Caca in his multiple incarnations swept down from the windy heights through the various fertile valleys of the Pacific slope (chap. 8). As he went, he subjected the Yunca to his own people, expelling many of them, reorganizing their lands, creating a cultic order in which both victors and vanquished would participate, and winning the Yunca wealth (e.g., chaps. 9, 25).

What human movements does the Paria Caca mythology allegorize? In a series of highly original studies María Rostworowski de Diez Canseco (1978: 31-147) has interpreted the narratives as reflections of a large and gradual prehistoric movement in which pre-Incaic highlanders of the ethnic group called Yauyo worked their way downward and southwestward, from their early home on the high tundras at the Cañete River headwaters, through various warm irrigated valleys (including the Mala, Lurín, and Rímac valleys, which form the heartland of the mythology), toward the Pacific shore and its rich deltas. Using independent and ostensibly nonmythical bureaucratic records she has been able to document Yauyo populations all over the territory of the manuscript and has mapped their major territorial divisions as understood in the mid- to late sixteenth century (Rostworowski 1988: 56-57). One of these Yauyo populations (that of Chaclla) attested a folk history closely resembling stories in the manuscript (Rostworowski 1988: 54-57).

Linguistic indices are less clear but also suggest some association between Paria Caca's mythology and a Yauyo expansion. Traces of Aymara-like lexicon and phonology in the Huarochirí manuscript indicate that the informants knew or were influenced by the same ethnic language whose modern forms persist as residual "islands" in old Yauyo territory (Gentile Lafaille 1976: 14).

In Inca and early colonial times those natives who classed themselves as Yauyo regarded the focal area of the manuscript, especially what later became the parish of Santa María Jesús de Huarochirí, as the very core of Yauyo political space. The early colonial Yauyo do not seem to have had a king or center of political command, but Yauyo witnesses said that in conducting intranative diplomacy their ancestors had recognized the Ninavilca lords of Huarochirí village as paramount authorities for seeral generations. In 15 5 8, Cristóval Malcachagua of Huarochirí gave unambiguous testimony that the Ninavilcas meant to rally "all the Yauyos" in defending Yauyo coca claims in the middle Chillón valley against Huamantangas and Spaniards (Rostworowski 1988: 94). "Yauyo" is the only term the manuscript uses in categorical contrast to the ethnic terms for foreign neighbors (the various Yunca groups, the Huaman Tanca, and the Huanca; supp. II, sec. 485). All this evidence seems a prima facie reason for treating the manuscript as substantially an artifact of Yauyo culture.

On first inspection Paria Caca does appear to be simply the chief deity of these Yauyo populations, ennobled by Inca patronage but still very much an ethnic symbol. Felipe Guaman Poma de Ayala thought of him so ([ 1615 ] 1980: 1: 241). And when Maca Uisa himself, one of Paria Caca's "sons," asked the Inca to dance a priestly dance essential to the Paria Caca cult, he told the Inca to dance it "the way our children from the Yauyo do" (chap. 23, sec. 297).

Nonetheless, one must account for some striking slippages in the association between Paria Caca and Yauyo identity. The predominant tellers of the Huarochirí myths, the Checa, thought the ancient founders of their leading kindreds were not Yauyo but Yunca. They scorned the Yauyo kindreds in Checa as half-wild nomads and barely tolerated their presence in the pilgrimage to Paria Caca (chap. 24, sees. 305-309) until Paria Caca himself taught them to respect the newcomers. In neighboring Concha, "Yauyo country" was a byword for a remote and fruitless backwater (chap. 31, sees. 391, 408). Perhaps such scorn reflects the fact that by 1586 the self-identified Yauyo had shrunk to a tiny ethnic minority comprising only about 5% of the 7,000 tributary households in the huge province to which they gave their name (Dávila Brizeño [1586] 1965: 155).

How should one read these seeming contradictions? One clue is that although the Checa narrator applied the word "Yauyo" to immigrants whose exact place of origin in Yauyos Province was still remembered, and who were considered nomads only recently attached to Checa, he also credited higher-ranking kindreds in Checa with invader origins. The difference was that the higher-ranking kindreds had arrived of old and had inserted themselves in the agricultural social structure created by the ancient Yunca founders in the days of Huallallo Caruincho. A Concha teller gave much the same account of his group in chapter 31.

Apparently the Checa and Concha used the term "Yauyo" to refer to recent herder migrants, while regarding their own ancestry, which might well have been historically no less Yauyo, as quasiautochthonous because it had been grafted by ritual and marriage onto the regimen that included "village-owning," valley-oriented agricultural huacas. Perhaps they saw no paradox in scorning people of the very sort that Paria Caca favors in his myths—impoverished wanderers from the heights—because they regarded their own ancestors as more powerful "children of Paria Caca." The proof of their superiority was that these ancestors had won dominion over Yuncas and the right to aggregate Yunca huacas into their religion while their compatriots still wandered the heights in monoethnic pastoral groups.

As a working hypothesis, one may imagine that Paria Caca's cult as recorded here—that is, in a state of bipolar coordination with Chaupi Ñamca's cult and with Yunca components generally—is the precipitate of, and a commentary on, a long and apparently still continuing series of migrations or incursions from the southerly highland fringes of the manuscript's territory. As in other world areas where pastoralists penetrated the edges and eventually the centers of agricultural societies (e.g., China under Mongol rule), the assimilated descendants of early invaders came to inhabit preexisting social and ritual forms and champion them against later invaders.

The crux of Checa and Concha religion appears to be a systematic priestly synthesis that exalts early invader groups by placing their origin myths, eloquent of a mostly pastoral highland way of life and a kinship-based social ideology (that is, the myths of Paria Caca's "children"), into a dyadic relation with ancient macroregional cults rooted in coastal agropastoral society (those of Chaupi Ñamca, Pacha Camac, etc.). It does so by interpreting the deity who was taken to represent the sum total of invader origins, namely, Paria Caca, as Chaupi Ñamca's brother and wife-giver to Pacha Camac. In this way Paria Caca was imbued like Pacha Camac and Chaupi Ñamca with a power and an identity transcending immediate ethnicity and locality.

The spokesmen of the system, perhaps Yunca priests belonging to Caca Sica ayllu or their followers, defined Checa cultural identity as descent from the creators of one local instance of such highlandvalley fusion. To be an ancestrally entitled worshiper of both Chaupi Ñamca and Paria Caca, and of their local affiliated cults, was the crux of belonging. The term "Yauyo" in Checa and Concha accordingly connoted neither 'foreigner' nor 'compatriot' but 'parvenu': immigrants not yet fully inserted into the regional-scale cultic order and its local apparatus.

This view helps us understand a problem of belief, namely, why outsiders like Guaman Poma saw the Paria Caca religion as a Yauyo cult par excellence, while the people of Checa used the word "Yauyo" to label a group barely admitted to Paria Caca's pilgrimage. To characterize the historical connection between Yauyos and Yuncas objectively is much harder. Whether the early highland aggressors whom the tellers identified with the mummified heroes called "children of Paria Caca" were in fact exclusively Yauyo, or whether such groups can be dated and archaeologically linked to Yauyo populations at all, remains unknown. For these reasons in the introduction we shall simply call the mythic protagonists "invaders" except where there is a specific warrant for using the term "Yauyo."

Rostworowski identifies the Yunca or coastal groups at whose expense the children of Paria Caca purportedly expanded as the two large politically unified collectivities (señoríos) closest to modern Lima. One, occupying the lower Rimac and Lurín valleys, was called Ychma locally and Pacha Camac in Inca usage; it housed the mighty shrine of Pacha Camac and enjoyed great religious prestige even after Yauyo and Inca depredations reduced its political reach. The other was the domain of the lord called Colli Capac, whose people the Huarochirí narrators called the Colli and whose seat the Spanish called Collique. It was based in the lower Chillón valley (Rostworowski 1988: 60-62). The Checa thought the Colli had founded the highland settlements their own ancestors conquered (chap. 24, sec. 341; chap. 25).

Chapters 11, 12, 17, 24, 25, and 26 contain what appear to be "charter" myths of specific invader groups that identified their founders as "children" of Paria Caca. If they have a unified thrust, it is that Paria Caca's fivefold self, ambiguously developed as a union of brother-huacas (chap. 8, secs. 99, 105; chap. 16, sec. 202) or of Paria Caca's human "children" (chap. 9, sec. 113), through many victories created a regional order embracing both victors and vanquished. By a combination of warfare and courtship at superhuman and human levels, Paria Caca turned his relation with the rich aborigines of the valleys, the Yunca, from one of strangeness and enmity into one of coordinated worship, in-law kinship, and interdependency (though certainly not without an abiding tension between groups). The Paria Caca heroes' victories over the Yunca, whom the myths picture as wealthy but immoral, fill several chapters. Chapter 31, the charter of the lineages of the Concha (neighbors to the Checa; see especially secs. 388-403) pictures the invaders as trying to assimilate to the aboriginal norms of their in-laws. Paria Caca's complementarity of form and function with the local female (possibly Yunca) huaca Chaupi Ñamca suggests that, overall, the mythology of Huarochirí construes a folk memory of conquest as an ideology of affinal interdependence.

How does this happen? The memorable fifth chapter is the locus classicus of a theme repeated many times in the manuscript: the poor ragamuffin who, because he is privy to a superhuman secret, carries within him a future power that his rich and splendid contemporaries cannot see. In chapter 5 it is the Baked Potato Gleaner, a byword for poverty, who makes Paria Caca's potential power real by a twofold action: he overturns the extant order's hierarchy (curing but simultaneously humbling a Yunca lord) and at the same time he becomes literally wedded to it (marrying the humbled lord's daughter). Empowered, he introduces Paria Caca's cult to a society transformed by combat and courtship.

The manuscript's ideological image of the invader-aborigine interaction as a passage from hostility to symbiosis may be seen as a one-sided rendering of, rather than a mere fiction about, the politics of coexistence between highlanders and coastal peoples. The dynamic it mythically expresses seems to have been driven by the highlanders' need for cultivable land, which they sought by downward invasion and establishment of "vertical" outliers in the mid-altitudes. Lowlanders in turn sought to capture more water sources by extending their canals upward and exerting political power over the lakes and streams of the heights (Torero 1974: 73-79). The 1558 lawsuit involving the Yauyo of Chaclla and the Yunca of Collique (among others) shows how in prehispanic times the downward penetration of Yauyo "vertical archipelagos" (Murra [1972] 1975a) led to conflicts so mutually costly that Yunca and highlanders reached a modus vivendi including cooperation on shared irrigation work and ritual reciprocity between leaders. The Huarochirí text glorifies some roughly comparable modus vivendi in mythic idiom. But the lawsuit also shows what the Huarochirí manuscript deemphasizes: that such arrangements were delicate and unstable, liable to lapse into prolonged violence when politically stressed. Even under Inca pacification, the Yauyo of Chaclla, their non-Yauyo Canta rivals, and the Colli Yunca incessantly tried to cheat and coerce each other. When the Spanish invasion unleashed the resulting tensions, the fights and suits that followed did much to drive several polities into helpless poverty (Rostworowski 1988: 83-291).

Chaupi Ñamca and the mythology of gender

Several passages detail the priesthoods, games, sacrifices, huaca-impersonating dances, and oracles that ritually organized the interlaced aboriginalinvader societies that the tellers saw as the product of their ancestors' victories. Although the relationships among groups and deities are complex and often far from obvious, one unifying motif is clear: the union between invaders and aborigines is ideologized in terms of a fraternal tie between the highest male deity of the invaders and the highest female deity of the aborigines. The link is taken to warrant marriage alliance between their respective human progenies. In this scheme female deities play an enormous role.

Chaupi Ñamca, supreme among female huacas, was a land and river deity of the lower Rimac whose great temple at Mama symbolized her ancient standing as Pacha Camac's wife (Dávila Brizeño [1586] 1965: 163). Her name means 'center Ñamca' and she may, like her spouse, have had a following across various valleys; in 1562, Mama was the place chosen for a summit meeting of native lords from the whole region (Murra 1980: xviii). The devotees of Paria Caca have conceptualized her component cults as a fivefold sisterhood, so as to match the form of Paria Caca (chap. 10, sec. 147; chap. 13, secs. 175-183). In the synthesis that dominates the manuscript, Paria Caca and Chaupi Ñamca are made into siblings (chap. 13, sec. 172; Avila [1645] 1918: 64). These claims apparently mirror increased Yauyo penetration of domains in which her marriage to Pacha Camac had once formed a dominant cultic axis. Indeed, in one version the newcomers' claim goes beyond fraternal symmetry to imply superiority: Chaupi Ñamca's five selves are styled "daughters" of Paria Caca, whereas Paria Caca's selves are never said to be her children (chap. 8, sec. 101). The "children of Paria Caca" have also expressed their conviction of political superiority by making their father-huaca wife-giver in relation to Pacha Camac.

In establishing a sibling relation between male mountain-huaca and female valley-huaca, the Paria Caca priests may have followed an earlier prototype. A partially obscured tradition (chap. 8, secs. 106-107; chap. 10, sec. 143) allows us to glimpse Mana Ñamca or Mama Ñamca, a synonym or comportent of Chaupi Ñamca who was once a lowland female counterpart to the male highland huaca Huallallo Caruincho (and who, like him, was a fiery power expelled by Paria Caca).

Within the dominant synthesis, ritual order richly embodied the idea of male-female symmetry at the apex of huaca genealogy. Paria Caca's priesthoods and his great festival the Auquisna (chap. 9, secs. 117-140) find their explicit counterpart in Chaupi Ñamca's festival the Chaycasna (chap. 9, sec. 122; chap. 10, secs. 149-151), so that the ritual and mythical order opens out in a grand sexual complementarity on the pattern Invader : Paria Caca : Male :: Aborigine : Chaupi Ñamca : Female. The fraternal treatment of the male-female apical deities tends to express ideals of harmony or equilibrium between groups at a totalizing, wholesociety level.

However, as one passes from the supreme powers to their huaca offspring and their human descendants, the dominant gender metaphor changes from brother-sister fraternity to marriage. A repetitive motif concerns invaders, sons of Paria Caca, who marry Yunca-descended women, implicitly daughters of Chaupi Ñamca (chap. 5; chap. 24, secs. 305314). Thus the sons of Paria Caca become indebted wife-takers to Yunca groups and their huacas.

This conviction of indebtedness parallels what was remembered as invader expropriation of the female and Yunca element in nature, namely, irrigable land. On the ecological plane the relationship between invader and aborigine is likened to the union of wild water from the heights (Yauyo-like, male) with the soil of the valleys (Yunca-like, female). The motif is discussed in the section below on the concept pacha. Because this union—irrigation—was in fact the greatest agrarian wealth of the western Andes, many myths can also be read as combined cosmic and political charters for local groups' rights to specific lands and canals. The abundance of landmark detail in the text, which we have tried in notes to key to modern cartography, reflects the function of myth as a memory bank of information about the tellers' generally invasionbased claims on resources.

The conjugal metaphor for invader-aboriginal relations carries a different symbolic load from the fraternal one. Whereas the siblinghood of the apical huacas is a static, merely classificatory relation, marriage myths are myths of social dynamics. They express not only an ideal of productive and reproductive union, but also an image of the many tensions involved in creating such union. For example, Collquiri, a water-huaca from the heights (chap. 31, secs. 408-436), must fight with his prospective inlaws and submit to their humiliating discipline before his desire for their beautiful land-huaca daughter can turn from destructive lust (flooding) into productive marriage (irrigation). Indeed, all the myths related to marriage treat it as an image of social stress and change latent in union. In chapter 5 Paria Coca's protégé Huatya Curi, a foreign male who wants to marry a human incarnation of Chaupi Ñamca, garners the hatred of his brother-in-law and establishes himself in his new household only by fighting within it (secs. 49-70). On the political level, accepting a foreign spouse is always a gesture of submission that one accepts at peril to one's autonomy, and for this reason Paria Coca's son Maca Uisa turns down the Inca's honorific gift (chap. 23, sec. 300). Even the Inca himself forfeited his empire by accepting a bride from a secret power that turned out to be the Spanish (chap. 14). While one may read these texts as comments on marriage as such, they also give voice in kinship idiom to fundamental doubts about the political stability of conquest-based agropastoral society.

In sum, the gender mythology of Huarochirí, though centered on an idealized complementarity, is at the same time emphatically a conflict model of society. It envisions every complementarity, whether marital, ritual, ecological, or political, as shadowed by submerged conflicts that had to be repressed in order to institute it. The inseparability of complementarity from conflict is implied to be a motor force in the mutability (what we would call the historicity) of west Andean society.

The Incas as seen from Huarochirí

Chapters 18, 19, 20, 22, and 23 vividly illustrate how the Inca conquest looked to provincial natives and give important clues as to how Incas manipulated the local pantheon. The Inca himself was said to have acknowledged and even subsidized local huacas—both those that local people thought of as ancient, like Paria Caca (chap. 18, sec. 220), and those newly emerging, like Llocllay Huancupa (chap. 20, sec. 243). This policy may be related to the fact that the Yauyos were reputed to be generally pro-Inca. Early evidence from the northernmost part of the Huarochirí orbit indicates that the Incas favored Yauyo efforts to penetrate the Yunca and gradually enlarged Yauyo enclaves based at Chaclla to the detriment of the anti-Inca Yunca of Collique and rival non-Yauyo highland groups (Rostworowski 1988: 148, 161, 178). Guaman Poma ([1615] 1980: 1 : 240-241) drew a picture of the Inca adoring an image of Pocha Camac in the shrine of Paria Caca.

The tellers of the manuscript likewise expressed their alliance with Incas in the idiom of ritual, adoring and helping finance some of the pan-Andean deities that Inca propaganda promoted. Among these figured Pocha Camac, the originally Yunca 'World Maker and World Shaker' whose shrinecitadel lay at the western edge of the mythic landscape (chap. 22, secs. 276-284; Patterson 1984[?]). So thoroughly had the Incas intercalated regional and local cults with royal religion that the Huarochirí people thought the Incas themselves owed some of their victories to help from Paria Coca's offspring (chap. 19, secs. 228-229; chap. 23).

The tendency to equate Cuni Raya, a coastal Trickster-demiurge embodying the transformation of landforms by water, with the invisible Vira Cocha fostered by Inca cult may also have been heightened by interaction with Incas. But for the Huarochirí narrators, this huaca stood over and against Inca power as against all other human power. Chapter 14 recounts an enigmatic Cuni Raya Vira Cocha myth alluding to the fall of the Incas. It apparently refers to the fact that the Incas had divided their domains in dynastic struggle just before the Spaniards arrived. The narrator tells how Cuni Raya Vira Cocha inveigled the Inca king with a beautiful bride and then, inducing him to "draw a line across the world" (chap. 14, sec. 196), made him fatally retreat from his sacred center at Cuzco. During the resultant outbreak of chaos, "people scrambled for political power, each saying to the others, 'Me first!"Me first!"' (chap. 14, sec. 197). While they wrangled uselessly in civil warfare the Spanish appeared at Cajamarca.

The Spanish invasion as seen from Huarochirí

Even more remarkable are the unique chapters that give us a glimpse of how natives, after a lifetime of colonial afflictions, looked back on the first confused moments of their contact with the Spanish. The first inkling of the epochal events arrives in the same way that the prophecy of Paria Coca's imminent victory had arrived long before (chap. 5), as a secret that Paria Caca vouchsafed to a despised ragamuffin from the heights. This time the prophet of the coming crisis is the "Mountain Man" Llacuas Quita Pariasca (chap. 18). The same chapter tells how one huaca priest survived the Spaniards' attempt to burn him alive and became a leader in preserving and sheltering the huacas amid Spanish attacks. Many references (e.g., chap. 9, secs. 122, 125, 136, 137; chap. 10, secs. 144, 148, 149; chap. 13, sec. 175; supp. I, secs. 449, 469, 473) show that those who rescued the huacas succeeded in hiding their cults (sometimes, a censorious voice warns, by camouflaging them in Catholic ritual) far into the colonial era. At least one of the contributors to the text regularly obliges Father Avila by warning that many Indians' conversion was a faÁade (chap. 9, secs. 133-134). Other chapters (20-21) reveal that even Father Avila's most obsequious ally, Don Cristóbal Choque Casa, in a certain sense still believed in—was even dominated by—the "evil ancient demons." These chapters tell with astonishing vividness and intimacy of Cristóbal's visionary combat with the huaca Llocllay Huancupa and of how Llocllay returned to battle Cristóbal once more in a dream. Cristóbal's dilemma—the need to validate his Christianity by conquering huacas, combined with inability to conquer them convincingly without invoking the same mythic paradigm he proposed to replace—adds up to a uniquely moving image of the Andean convert's stressful and compromised position.

Specialized chapters

The manuscript contains some specialized chapters perhaps given in reply to questions about arts that Tridentine Catholicism forbade as diabolically inspired, heretical, or superstitious. Chapter 29, short but important, sketches Andean astronomy or astrology. It suggests that the tellers thought of certain "black" constellations and certain star clusters as the celestial prototypes of the earthly beings they resembled. Visionaries saw these constellations "descend" to earth and shower their protégés with specific vital force.

Chapters 27 and 28 treat the cult of the dead. Chapter 27, a droll explanation of why the dead no longer come back, echoes a motif from chapter 1, namely, that humans must accept irremediable death as the alternative to Malthusian disaster. Chapter 28 recounts funeral and commemorative custom, somewhat evasively in the crucial matter of how ancestors' bodies were treated. Perhaps the tellers hoped to avoid setting persecutors on the trail of their beloved and (ideally) everlasting ancestral mummies.

Two "supplements" (which may be rough drafts) lay down the ritual duties of couples who give birth to magical children—twins or babies with small birth defects. These unusually difficult sections afford an idea of the complex ritual obligations embedded in in-law ties, with emphasis on the ritual exchange of immense amounts of coca and other wealth.

The Huarochirí region's people and their historic situation

These myths relate intimately to the real-life landscape and the historic conjuncture that generated them. For one thing, the deities themselves are land features and local climatic forces. For another, these priesthoods and celebrations themselves organized productive work, so that the religious regimen encodes practical as well as ideological information. Finally, every facet of religious organization—spatial, calendric, and hierarchical—reflects the pressure of colonial circumstance.

The westernmost range of the high Andes runs northwest-southeast, parallel to Peru's Pacific shore. The long slope from its icy crests down to the desert beaches forms a rugged watershed cut at intervals by many small rivers carrying meltoff from the heights to the ocean. The scene of the Huarochirí mythology is a segment of that slope. Most of the places mentioned in it are close to three valleys: the Rímac River valley, on whose lower banks the Spanish built Lima, but whose upper tributaries, the Chaclla and Mama rivers of the manuscript, were still mostly native territory in 1600; to the south, the Pachacámac River valley, today called the Lurín, on whose headwaters stood the colonial parish of San Damián; and still farther south the "River of Huarochirí," today the Mala, which gave its name to a colonial parish as well as to the province and to the manuscript. The people who made the manuscript also seem to have included in their range of meaningful geography one valley to the north of Lima, the Chillan, and two more to the south of Huarochirí, the valleys of the Omas and Lunaguaná (today Cañete) rivers.

Karen Spalding's Huarochirí: An Andean Society under Inca and Spanish Rule (1984) meticulously reconstructs the colonial society from which the manuscript emerged. Spalding (1984: 3) cautions us that colonial Huarochirí Province was no huddle of hamlets but a region the size of Massachusetts. Yet neither is the landscape whose meanings the myths unfold too huge for Andean travelers to have absorbed from personal experience. Taylor's map (1987b: 39) suggests that the widest span of the mythic landscape reaches about 8 5 air km (5 3. miles) from ocean to mountaintops, and about 120 air km (74.5 miles) from north to south. A separate reckoning by John Treaty (1984) estimates the main mythic scene as about 108 air km (67 miles) east to west and about 56 air km (35 miles) north to south. Of course, estimates using air kilometers hardly simulate the twisting paths pedestrians and llamas actually followed across the contorted mountain landscape. Even on foot, highland travel involves innumerable hairpin turns up and down the faces of river chasms. Yet, even allowing for steep climbs and zigzags, these distances would hardly have discouraged Andean travelers. Averaging Cotler's (1958: 113-114) and Spalding's (1984: 31) statements on walking speed over the Huarochirí terrain yields 2.4 km/hour as a usual rate for the area. If so, a person walking eight hours a day might have descended from the high shrines of Paria Caca to the Pacific shores in about four days. These are only order-of-magnitude estimates, but they suffice to remind us that the world pictured in the Huarochirí text is mostly a world of experienced rather than imagined landmarks and spatial relations.

Ecologically, the mythic scene spans a remarkable spectrum of "vertical tiers" (Murra [1972] 1975a; ONERN 1976). The uppermost points mentioned are icy peaks more than 5,o00 m above sea level. The myths clearly take as their backdrops the "pluvial tundra" of Mullucocha (5,000-4,300 m above sea level) where Paria Caca annihilated a mountain. The heroes' downward course traversed the "very humid páramo" (4,500-3,900 m above sea level; páramo means high grassland) and the "humid woodland" (3,800-2,300 m above sea level). This is all high, seasonally rainy and snowy country with numerous lakes and probably included the llama-herding lands of the ancient peoples up to 4,600 m above sea level. The tellers of many myths identified with the splendor of the heights and with their product, camelid wealth on the hoof. One of the major ceremonies was a llama race to the sacred heights, with the first llama acclaimed and given a sacred name (chap. 9, secs. 120-122, 128).

Agriculture belonged to lower tiers, four cultivation zones compressed within narrow valleys: the "montane steppe" at San Damián, and three increasingly lower, dryer grades of "desert scrub" (matorral desértico). The colonial resettlement villages in which the manuscript seems to have been made stood in the upper part of this range, 3,000 to 3,500 m above sea level. The agricultural terraces immediately around them were and are irrigated, but on the "montane steppe" above them dry farming of high-altitude crops is also possible. There people cultivated tubers, especially potatoes, and the Andean grain Chenopodium quinua, which one tradition mentions as a mythic plant that gave birth to people (chap. 24, sec. 302).

Irrigable crops grew best in places where canal water fertilized pockets of sun-warmed soil amid the rocky river canyons. Irrigated valley lands symbolized abundance, and access to warm land with water was seen as the limiting factor in winning wealth. The tellers of the myths especially associated maize with irrigation, for example, in the poignant figure of the young woman who stood in her field and cried as she watched her insufficiently watered maize shrivel up (chap. 6, sec. 83). Chapters 30 and 31 are vivid mythic renderings of the struggle between water-poor and water-rich groups.

Still further down-valley but short of the coastal plain, at altitudes around 1,000 m above sea level, irrigation takes place in small patches of warm riverbank land commonly called chaupi yunca ('mid-yunca' or 'semi-yunca'). These sheltered, gardenlike river margins were coca lands in antiquity, prized for the abundance of the sacred leaf that Huarochirí natives ceremonially lavished on each other and on their superhuman patrons. The manuscript is especially detailed in explaining how coca cultivators fit into Paria Caca's cult regimen (e.g., chap. 8, sec. 109).

Where the fertile valleys meet the sea, a rich fishing industry combined with delta and riverbank agriculture to yield wealth that awed even the Incas. Prehispanic seaboard dwellers developed pisciculture intensively (Espinoza Soriano 1974). Although the myth-tellers thought of themselves as highland-descended, they, too, imagined salt water as part of the humanized and culturalized landscape. Fish-farming was so familiar that the teller of chapter 2 thought of ocean fish as runaways turned wild:

At that time there wasn't a single fish in the ocean.

Only Urpay Huachac used to breed them, at her home, in a small pond.

It was these fish, all of them, that Cuni Raya angrily scattered into the ocean ... (chap. 2, sec. 26)

At the time when the manuscript was written, much of the prehispanic technology and some of the prehispanic land tenures that once governed use of these productive zones still functioned. But the Yunta, Yauyo, and Inca claims were no longer the only factors. The social organization of land—administrative divisions of territory, settlement patterns, and legal land rights—had become an even more complex, multilayered scheme composed of native and colonial elements. Diego Dávila Brizeño's ([1586] 1965) report on the "Province of the Yauyos" and Karen Spalding's monograph (1984) afford valuable clues to the de facto world the myth-tellers inhabited.

The most modern part of the settlement pattern was the one Dávila Brizeño himself had helped impose—concentration into forced resettlement villages (reducciones). Each village gathered together numerous non-nucleated settlements scattered up and down the valleys to make a conveniently taxable package on the model of a planned Spanish village (see the section on the concept llacta below). These parishes inorganically merged fragments of various units earlier defined by the Inca and aboriginal political orders. One of the things that makes the Huarochirí manuscript hard to read is the fact that the tellers are constantly comparing the (to them) arbitrary reducción map against the (to them) intelligible distribution of groups and deities that their hidden ritual still commemorated. Thus their toponymy refers to both ancient and current mental maps.

Nonetheless, the Spanish administrative plan did still make use of some older orderings. A map made by Dávila Brizeño representing Spanish administrative geography of the 1580s in the context of native toponymy (Rostworowski 1988: unpaginated insert) shows the Yauyos Province bisected by what appears to be an Inca-style moiety dividing line: the southern half is called Anan ('upper') Yauyos and the northern Lorin ('lower') Yauyos. This feature the Spanish apparently retained or at least recognized. Colonial Huarochirí Province corresponded to Lorin Yauyos. And when it came to sharing out the right to rule and tax Andean people, the Spanish availed themselves of other Inca demographic and organizational schemes. The Spanish repartimiento, or political apportionment of tributary subjects, proceeded by assigning each established native lord to a Spanish encomendero. The patterning of colonial lordship thus legalized appears to echo Inca rules albeit with innovations and distortions. So the initial colonial system, while it functionally overturned many native norms, retained the gross anatomy of Inca political divisions, and some of these are detectable in the Huarochirí manuscript even after subsequent reorganizations.

The Incas saw each half of the region as composed of large blocs of people (perhaps ethnically defined), each comprised of several huarangas or 'thousands' of tributary households. In what Dávila Brizeño called Lorin Yauyos, which housed the ancestors of the manuscript's authors, the Incas recognized and the Spanish retained three major groupings:

• Lurin Yauyos
  • Huarochirí People
    • Huaranga Colcaruna
    • Huaranga Quinti
    • Huaranga Langasica
    • Huaranga Chaucarima
    • Huaranga Checa
  • Chaclla People
    • Huaranga Chaclla
    • Huaranga Carampoma
    • Huaranga Casta
  • Mamaq People
    • Huaranga Matucana
    • Huaranga Huanchor
• Anan Yauyos

(Spalding 1984: 54)

Each "thousand" had a native lord, or curaca. An overlord or curaca principal was recognized by the Spanish, on purported Inca or aboriginal precedent, as lord of all huarangas in the bloc. This was the highest office held by a non-Inca lord. By such a title the Ninavilca lords of Huarochirí (chap. 7, sec. 93; chap. 19, sec. 230) were recognized as having precedence among, if not power over, the huarangas most prominent in the manuscript.

Finally, within these Inca-influenced political arrangements there remained elements of a still older order. The pre-Inca (or non-Inca) settlements that the informants called their llactas, some already reduced to ghost towns, were still the points of mythic reference. At the microscopic level, a system of partly localized extended kinship reckoning, discussed below, remained the matrix for relations within and among households. To what degree the large, purportedly hereditary categories that dominate the overall scheme of the myths corresponded to any territorial reality or to any practical pattern of residence by 1600 is uncertain. But it is likely that the large, purportedly genealogical formations whose founders are the greatest heroes of the text, and the ayllus or ancestor-focused kindreds, still counted, at least theoretically, as the corporate holders of important rights in productive assets (see below, especially on the concepts yumay and ayllu).

In the later sixteenth century the Huarochirí people were not numbered among the poor. They were heavily taxed, and Spaniards vied for chances to exploit them. Far from being an obscure hinterland, their region lay astride the best-known route from the Spanish capital, Lima, to the Inca one, Cuzco. But their relative wealth did not protect them, and in some respects the inhabitants of Huarochirí were an afflicted people. Epidemics of the 1550s, 1580s, and 1590s, within the living memory of myth-tellers, had mowed down a huge share of a population with weak immunological defenses. "If we accept Dávila Brizeño's estimate of approximately 6,000 household heads in Huarochirí in the 1540s, shortly after the entry of the Europeans, then the province lost. . . almost 30 percent of its adult male population in the quarter century between 1545 and 1571" (Spalding 1984: 173). The rest of the century was a period of continuing severe population decline in Huarochirí as in many Andean regions (Spalding 1984: 176). Around 1560 and again in the 1580s, survivors of the epidemics in some parts of the Andes (including southerly Yauyos) had asked the huacas to defend them, but suffered bitter disappointment in the failure of nativist movements (Taki Onqoy Moro Onqoy, and others; Curatola 1978; Stern 1982: 51-71). When Jesuit fathers conducted a conversion campaign in Huarochirí in 1577, they reported finding many people sick in their fields (Arguedas and Duviols 1966: 244). Missionaries routinely procured conversions by exploiting the desperation of the mortally ill.

Some of the myth-tellers were probably old enough to remember other calamities. When Spanish forces crushed the forty-year-old neo-Inca redoubt in Vilcabamba and then in 1572 executed the last independent sovereign of Inca blood, Túpac Amaru, Andean people everywhere went into mourning. The administrative overhaul that followed increasingly reduced native chiefs to the compromised status of tribute enforcers and cultural brokers. Some of the myth-tellers may have been among those routed from their homes and chased into strategic hamlets by Spanish functionaries in the 1580s. The "people called Indians" also had to comply with increasingly exhausting tribute regimens. Meanwhile, the authorities allowed Spaniards to appropriate landholdings that the natives lacked enough population to defend and sometimes let clerics exploit parishioners outrageously.

So by the probable date of the manuscript—1608—the inhabitants of the Huarochirí region had reason to wonder, like Quita Pariasca the Mountain Man (chap. 18, sec. 221), whether the world itself had turned bad. Several passages in the manuscript are colored by anxiety about sickness and depopulation (chap. 19, sec. 233; chap. 25, sec. 345; chap. 26, sec. 357). Observing that the natives of the rocky heights seemed to be holding their own even as the aborigines of the coast were perishing (and indeed became entirely extinct, through both depopulation and assimilation), the tellers of chapter 9 (sec. 140) wondered whether perhaps the rustic highlanders did not owe their vitality to greater faithfulness to the Andean religious tradition.

Into the world of the huacas

Some of the basic terms in Huarochirí myths embody unfamiliar Andean assumptions or categories. While glosses cannot capture their full senses, a reader at least needs some idea of what is lost in translation. The following paragraphs comment minimally on the most important ones.

Pacha: 'earth, world, time, place'

Huarochirí people called the world and time together pacha, an untranslatable word that simultaneously denotes a moment or interval in time and a locus or extension in space—and does so, moreover, at any scale. In chapter 18 (sec. 221) the Mountain Man, foreseeing the arrival of those destroyers who would turn out to be the Spanish, says, "Alas, brothers, the pacha is not good." He could have meant anything from 'this is not a good situation' (or 'moment' or 'conjuncture') to an idea as grand as 'the world is no longer good' or 'the epoch is no longer good'. The same word pacha is also the name of earth in general and in modern folk religion means Earth as a personified female being.

Cosmological ideas are not spelled out in the Huarochirí manuscript. Despite the use of solar gnomons for calendric astronomy (chap. 9, sec. 108), the Sun appears as a person in Checa lore only tangentially (chap. 4; chap. 13, secs. 172, 188). Sun worship figures as an Inca idea (chap. 22, secs. 276, 279) and as a motif in the genealogy of the huacas as the people of the lower Rimac valley imagined it (chap. 13, sec. 172). The moon is not personified anywhere. The Andean cosmos can be partly understood from these myths, but not because the myths explicitly describe it. We deduce what we can by noticing what the tellers take for granted. To some degree the picture can be augmented with data from other written sources and from recent ethnographic research that offers conceptual analogies. (Earls and Silverblatt 1978 propose a synthetic model.)

Two pervading generalities seem to order the data. One is that earth has predominantly female associations and water predominantly male. The second, partially overriding the first, is that altitude and motion connote maleness, while depth and stability are associated with femaleness.

Earth's living mass was imagined rising up from the waters of the surrounding ocean. (Whether the waters circled earth like Saturn's rings or whether earth's base was immersed like a boat's hull is not clear.) Especially in local instances, particularly when visualized as the green irrigable valley lands, earth is usually female. But the land's highest points, the great peaks whose ice-crusted crowns overtower the habitable earth, are usually male. One of them is Paria Caca. Roughly, the solid part of the world might be imagined as a single world mountain made of all the Andean ranges, rising from femalelike valleys to malelike snowcapped heights.

Water—rainstorms and mudslides, snow and glacial runoff, tiny irrigation canals and mighty rivers, even that astral river we call the Milky Way—is the kinetic part of the world. Water moves over pacha in a circular path. It rides up from the ocean into the sky along the Milky Way "river." Chapter 29 (sec. 375 ) tells us that water rises because a celestial llama constellation carries water up from below by drinking before ascending. Then water washes down onto the heights of the earth as storm and rain, bathing and fecundating earth as it descends to the ocean. People worshiped the great snowcapped peaks "because that is where their water comes from" (Avila [1645]1918: 83). Water is often male, especially storm water and downward-flowing water. Several myths liken the chasm-cutting power of rivers in spate, or the storm and flash flood that create disastrous mudslides, to warlike male violence:

As soon as they brought him up a hill, Maca Uisa, child of Paria Caca, began to rain upon them, gently at first.

The natives of that country said, "What could this mean?" and began to ready themselves.

When they did so, Maca Uisa reduced all those villages to eroded chasms by flashing lightning and pouring down more rain, and washing them away in a mudslide. (chap. 23, secs. 295-296)

Although the Pacific Ocean plays a key part in organizing mythic space, it is not personalized in the manuscript. In modern Andean myth, the ocean (Mama Cocha 'mother lake') is usually female. The huaca most closely associated with the sea is Pacha Camac (chap. 22, secs. 276-284), but this great force is not clearly defined as a maritime deity.

The hydraulic embrace of moving water and enduring earth was imagined as sex. Their embrace yielded a biotic system (Dumézil and Duviols 1974-1976) in which life forms emerge from mixed earth and water. Hydraulic sex was sometimes imagined as a turbulent affair. The myth-tellers, who seem to have faced the harrowing vicissitudes of a water-poor irrigation economy with a good deal of humor, likened its hazards to the comical chaos of undisciplined desire. In four myths (chap. 6, secs. 82-90; chap. 12, sec. 170; chap. 30; and chap. 3r, secs. 406-432) voluptuous earth-women offer their parched bodies to the virile water-huacas who rush down from the heights. The earth-women's self-serving tricks and the water-men's lecherous ineptitude can turn their meetings into comic disasters. In chapter 31, when the lake-huaca Collquiri rushes downhill to his earth-lover Capyama, the bursting pressure of his virility squirts out of every channel and sprays destructive floods all over Capyama's people:

... Capyama's elders shouted at Collquiri from their <crossed out:> [spring] village:

"Son-in-law, everybody's mad at us! Don't send us so much water!"

"Shut it off!"

"Hey, Collquiri! Hold back on the water!" they yelled.

With them shouting like that, Collquiri plugged the hole with a blanket and other stuff.

But the more he plugged it the more the barrier crumbled and the more the water kept bursting through over and over again.

Meanwhile the people from down below kept yelling at him nonstop:

"PLUG IT UP!!" (chap. 3r, secs. 425-426)

Pacha, the world as a given arrangement of time, space, and matter, is not supratemporal. It clearly admits change, even cataclysm. There have been times when pacha "wanted to come to an end," and the manuscript tells us how this can happen. Water might fail to ascend into the astral "river" and the world might drown as the ocean rose (chap. 3; chap. 29, sec. 375). Or Pacha Camac Pacha Cuyuchic, the 'World Maker and World Shaker' who sleeps under the ruined shrine that still bears his name, might turn over in his dreams and pulverize the world in an earthquake (chap. 22, sec. 284); the region is in fact subject to devastating earthquakes. Everything, including humanity, has been crushed and refashioned. The social order, too, is constructed and transformed in superhuman violence; the tellers imagine the original theophany that was their group birth as one of many cataclysms that simultaneously gave form to the land and to society.

In gross terms, then, the Huarochirí world opposes the qualities of still centricity—depth, solidity, dryness, stability, potential fecundity, womanliness—to those of a restlessly moving outer orbit—height, fluidity, wetness, movement, potential for insemination, virility. As the outer waters wash over the inner earth, these two fundamental lives mix in the circulation of water over soil. Lives that are both watery and earthy emerge: plants, animals, people. Born fat, wet, and juicy, all beings eventually—in the space of a dry season or a long life—separate out again into their original substances (Allen 1982). Moisture departs as vapor, leaving a weatherbeaten husk and the dry seeds of a future cycle. Nature and ritual combine to return the parts to their sources so the cycle can start anew. Humans like all others emerge fat and wet, but at the end of life their dried husk containing the potential for future life goes as a mummified ancestor (mallqui) back to earth. In chapters 27 and 28, a dry, seedlike being that emerges from the dead human husk—a fly—is the living residue of a dead generation. The function of ritual and sacrifice is to ensure a steady circulation of biological energy through pacha by conducting social exchange among its living parts.

Camay: a concept of specific essence and force, 'to charge with being, to infuse with species power'

The act by which huacas bring other entities into being is not expressed with the plain word meaning 'to make' (ruray), nor with the verb huallpay, which Gonçález Holguin took to mean a divine act of creation ([1608] 1952: 174), but with a different verb: camay. Camay escapes the seemingly handy glosses 'to create' (because 'create' connotes an ex nihilo act, while camay connotes the energizing of extant matter) and 'to fashion' (because 'fashion' suggests only an initial shaping of inert matter, whereas camay is a continuous act that works upon a being as long as it exists). But what does camay mean? The astronomical or astrological chapter 29 gives a crucial clue: it labels a llama-shaped constellation the camac (agentive form, 'camay-er') of llamas. On descending to earth this constellation infuses a powerful generative essence of llama vitality, which causes earthly llamas to flourish. All things have their vitalizing prototypes or camac, including human groups; the camac of a human group is usually its huaca of origin. Religious practice supplicates the camac ever to vitalize its camasca, that is, its tangible instance or manifestation. Taylor (1974-1976) has likened this idea to Platonic idealism, an insight that helps one understand the profoundly plural and ongoing nature of Andean creation but also minimizes its earthiness. Camac in the manuscript seems to suggest a being abounding in energy as physical as electricity or body warmth, not an abstraction or mental archetype.

Huacas could be camac to great or small categories of beings. The great coastal deity Pacha Camac bears in his very name—'Camac of space and time'—an all-embracing function as the vitalizer of worldwide realities, while local huacas animate smaller entities. Likewise, ordinary beings could be camasca (participial form; 'infused with camay') to different degrees. An ancha camasca person is a 'very powerful' one. But one can clearly see that camay means specific form and force, not general potency. In chapter 14 (sec. 191) three men boast of their speed, saying:

"I am a condor shaman!" some men answered.
"I am a falcon shaman!" said others.
"I am one who flies in the form of a swift!" replied still others.

What they said more literally is:

"I am the camasca of the condor!"
"I am the camasca of the falcon!"
"I am one who flies as a swift!"

The point appears to be that these men are three shamans whose patrons, the archetypes of birds who symbolize speed, have infused in them the species powers of speed and range of the condor, the falcon, and the swift.

Huaca: 'superhuman person, shrine, holy and powerful object; huaca priesthood

The Huarochirí manuscript is in large measure a reading-out of its space. The horizon, not the cosmos—geography, not metaphysics—poses the questions to which its most vibrant deities give answers. Andean numina lodge in places or placed objects: mountains, springs, lakes, rock outcrops, ancient ruins, caves, and any number of humanly made objects in shrines: effigies, mummies, oracles, and so forth.

Like all the other persons English forces us to call "deities," Paria Caca is a huaca. The half-Andean historian Garcilaso Inca de la Vega tried in 1609 to convey the sense of this all-important term by telling us that:

huaca ... means "a sacred thing," such as ... idols, rocks, great stones or trees which the enemy [i.e., Satan] entered to make the people believe he was a god. They also give the name huaca to things they have offered to the Sun, such as figures of men, birds, and animals.... Huaca is applied to any temple, large or small, to the sepulchers set up in the fields, and to the corners in their houses where the Devil spoke to their priests.... The same name is given to all those things which for their beauty or excellence stand above other things of the same kind, such as a rose, an apple, or a pippin, or any other fruit that is better or more beautiful than the rest.... On the other hand they give the name huaca to ugly and monstrous things ... the great serpents of the Antis ... everything that is out of the usual course of nature, as a woman who gives birth to twins ... double-yolked eggs are huaca ...

They use the word huaca of the great range of the Sierra Nevada.... The same name is given to very high hills that stand above the rest as high towers stand above ordinary houses, and to steep mountain slopes ... (Garcilaso Inca de la Vega [1609] 1966: 1:76-77; Livermore's translation)

A huaca was any material thing that manifested the superhuman: a mountain peak, a spring, a union of streams, a rock outcrop, an ancient ruin, a twinned cob of maize, a tree split by lightning. Even people could be huacas. Modern ethnography tells us that one extraordinary ethnic group, the Uru of Bolivia, was collectively called haqe huaca 'a human huaca' by Aymara-speaking neighbors because the Aymara thought the Uru had survived from primordial times (Manelis de Klein 1973: 143). In the manuscript, too, people can become huacas. Mummified ancestors of high rank (see below) could be huaca. The Inca rite of Capac Hucha (chap. 22, sec. 280) turned live humans—spotless children and youths-i-nto new huacas by burying them alive. The discovery of new huacas never ceased. Avila recalled that when one native brought home a black silk button with gold thread, which he had found in a garbage heap in Lima, a "master" of huaca worship revealed to him that it should be enshrined as a household huaca ([1645] 1918: 74). Chapter 20 details the discovery and career of another newfound huaca.

People owed huacas reverence and also plenty of goods: llama and guinea pig meat, brilliantly colored mineral powders, thorny oyster shell, clothing, coca leaf, maize dumplings, and maize beer. The Huarochirí manuscript lays down the priestly regimen for sacrifice and gives examples of the responses priests delivered on huacas' behalf:

They [huaca priests] gave people advice, telling them all sorts of things:

"You are to bathe in the confluence of two streams."

"You must sacrifice one of your llamas."

People were more than happy to obey their dicta. (chap. 13, sec. 186)

The Huarochirí manuscript tells a good deal about priesthoods, but not much about rules regulating lay worshipers' individual or group affiliation to huacas. People appear to have belonged to the cults of huacas considered as apical ancestors of their patrilineages, of their ayllus, and of the clanlike large groups identified with major founder-huacas. Apparently, heredity was a primary determinant of local religious duty, because when people went on their own initiative to consult the huaca oracle of the five Ñamca sisters, the Ñamcas would ask if the petitioners had properly consulted their hereditary huacas beforehand:

. . . these huacas would ask those who went to them, "Have you come on the advice of your own Con Churi [hereditary household deity], your father, or your elders?"

To those who answered "No," the huacas would reply, "Go back, return, consult your Con Churi first." (chap. 13, sec. 185)

Huaca shrines were powerful social and political corporations. When a huaca gained legitimacy, we learn from the "biography" of the huaca Llocllay Huancupa (chap. 20), worshipers built it a temple precinct and offered costly service. Politically sponsored huacas had endowments of both herds and fields that common people were required to serve:

The Incas worshiped these two huacas [the Sun and Pacha Camac] most, beyond all the others, exalting them supremely and adorning them with their silver and gold, putting many hundreds of retainers at their service, and placing llama herds for their endowments in all the villages.

The llamas of Pacha Camac sent from the Checa people stayed at Sucya Villca. (chap. 22, sec. 277)

Huacas also had human servitors and even spouses. Avila in 1645 reminisced about having met a good-looking eighteen-year-old girl, crippled and walking on two staffs, who had "been dedicated as a woman of an idol" (the huaca Maca Uisa). Maca Uisa lived with her in the form of a sacred blue stone, "no bigger than the palm of a hand," which she kept in a basket with tiny but luxurious garments to dress it ([1645] 1918: 69-70). He also mentions that a "saçerdotisa [priestess] very celebrated among the Indians" was known as a "woman of Pariacaca's" ([1645]19 18: 65). Both examples use the Spanish term muger, meaning 'woman' literally and 'concubine' or 'secondary wife' more freely, rather than the legal Spanish term for 'wife'. Because the Spanish terminology of marriage implies a sacrament, Father Avila might have scrupled to use it in connection with huaca marriage no matter how natives regarded the matter; so we cannot tell whether these unions counted as full marriages in the native system.

Some huaca priests clearly enjoyed class privileges. For example, common people did all the agricultural work on fields belonging to the yanca priests who had hereditary authority over a lake and its irrigation water (chap. 31, sec. 435 ) and they gave priests huge amounts of meat (chap. 24, sec. 335). Water priests are credited in the manuscript with almost dictatorial power over some of the processes by which society reproduced itself, including vital decisions about people's rights and duties in basic subsistence agriculture:

Because he was the yanca for this purpose, all the arrangements of the season were made in compliance with his commands.

When it came to irrigation he'd be the only one to give orders about it, saying, "It'll take place now" or "It'll be so many days." And all the Concha obeyed him to the last word. (chap. 31, sec. 434)

Early in his extirpation campaign, Father Avila imprisoned an important priest of Chaupi Ñamca in Mama village named Hernando Pautar (who may be a source of the distinctive subregional mythology contained in chapter 13) and wrung from him a confession about the privileged life of huaca priests. It should be read with caution, since it comes to us via a paraphrase composed four decades after the event:

... it is true that I've been a priest [saçerdote] of Chaupiñamocc since I was a youngster, and I inherited [the priesthood] from my father, and in all these villages of this parish and others, they've respected me a great deal, and I used to come visit them twice every year, and if I was late they used to send someone to call on me, and they would send me horses, and people to serve me on the road, and whenever I entered a village they would erect arches, and they would come out dancing, with the women beating their drums, and they would give me lodging, and fed me, and served me, and they gave me so much that I didn't know what to do with it all. At my rear they made something like a cabin of boughs, and they would cover it and close it off with cloaks, and the floor used to be covered with fresh straw, and I would enter into it alone, by day or by night as I preferred. There they would come to consult me, and I responded, and sacrificed guinea pigs, pouring out maize beer, and I used to perform other ceremonies in view of those who attended, and some used to say that they wanted to hear a response given by Chaupiñamocc, and I used to make her speak by placing there a little idol that represented her, and sometimes I would talk in a very high voice and other times very low. . . . And for this everyone respected me, as much as they do you [Father Avila], and much more. On the third or fourth day they would bring together maize, potatoes, and a lot of food for me, and they would dispatch it to my wife in my village. In this village I'd be lodged in one fellow's house the first time, a different fellow's the next ... (Avila [1645] 1918: 68-69)

Through oracles and through their privileges in ratifying rites of passage, the huacas' priestly representatives closely governed the ongoing business of society. Another sort of priesthood, the nonhereditary huacsa office, rotated among members of appropriate kin groups. Yancas seem to function as calendric and technical authorities, oracles and mediators, while the huacsas' main duty was to impersonate the great huatas in festivals and reenact their myths. Some of the narratives may be verbal "scripts" of huacsa performances.

In comparison to the priesthoods, the manuscript has little to say about the power of nonpriestly leaders such as the political "native lords" called curacas or the native magistrates (alcaldes) appointed by Spanish colonial officials. It mentions several postconquest curacas by name, often commenting on their attitudes toward huatas or as a device for setting a chronological context. Several passages imply that the curaca was expected to take a prominent part in huaca ceremonies (chap. 13, sets. 173-174, 176) and that he could exert leadership over a community's decision to adopt or neglect particular huacas (chap. 19, sec. 231; chap. 20, sets. 244-245).

Huacas had vibrantly individual personalities. Paria Caca, whether in his incarnation as the hailstorm of the icy heights or as the five-in-one hero who beat the Yunca down toward the sea, seems haughty, brilliant, and cold, a driving deity. Llocllay Huancupa (chaps. 20 and 21) lurks in the dark like an inarticulate beast, roaring dully over his immolated meal. Chuqui Suso, the sexy agricultural huaca (chap. 6, sets. 82, 88-90), and her sister who seduced the Paria Cacas into irrigating her (chap. 12, sec. 170), as well as the great maternal huaca Chaupi Ñamca, were associated with sensuality and playful ease:

"Chaupi Ñamca enjoys it no end when she sees our <crossed out:> [cocks] private parts!" they said as they danced naked.

After they danced this dance a very fertile season would follow. (chap. 10, sec. 151)

They used to stay there all night long, staying awake till dawn, drinking and getting drunk. They got real happy performing the dance called Aylliua, and danced that night drinking and getting drunk until dawn.

After that they went out to the fields and simply did nothing at all. They just got drunk, drinking and boozing away and saying, "It's our mother's festival!" (chap. 13, sec. 174)

Clearly huacas are living beings, persons in fact. Avila and other Christian seventeenth-century observers seem to think of the huacas as real beings, material in form but animated by demonic spirit. When Avila applied the terms "god" and "goddess" to huacas (for example, Cuni Raya's beloved Caui Llaca; Arguedas and Duviols 1966: 202-203), he seemed to be thinking of them as linked to natural forms, anthropomorphic and tangible, something like the deities of Greco-Roman mythology and the "demons" of medieval Europe whose lore he apparently knew in some detail (Arguedas and Duviols 1966: 207, note 3). In this partly Greco-Roman sense, the word "deity" might fit the huacas. But one must be careful; the dualism of substance enfolded in the Christian usage of the words "god" and "divine" seems superfluous for understanding Andean worship. The world imagined by the Checa does not seem to have been made of two kinds of stuff—matter and spirit—like that of Christians; huacas are made of energized matter, like everything else, and they act within nature, not over and outside it as Western supernaturals do.

The tellers give pacha shape by mapping onto it a society of huacas that mimics idealized human social structure, mainly genealogical and affinal. Earth forms, superhumanity, and society match each other in a structure of correspondences. Great snowcaps are great creators. Their myths are often shared among large populations (perhaps tens or hundreds of thousands before the Spanish conquest) united by language and other markers of public identity; often they are the groups called "nations" by some Spanish authors and "ethnic groups" by modern students of Inca polity. Paria Caca, whose body was a double-peaked snowcap high in the rugged western range, was known to Guaman Poma ([1615] 1980: 241) as one of the greatest deities. The pilgrimage to his shrine (Bonavia et al. 1984) united thousands of people in a trek over the last lap, a race up the "steps of Paria Caca" still visible today. The pilgrims visited the lake where Paria Caca quenched the Huanca deity Huallallo's fires, sacrificed llamas at Paria Caca's dwelling, and, according to Dávila Brizeño ([1586] 1965: 161), "climb[ed] the peak of the snowcap to offer their sacrifices."

The Huarochirí narrators explained their various ancestor huacas as heroic offspring of this great regional huaca. In the Huarochirí manuscript Paria Caca is said to consist of five beings who are, in various contexts, equal or ranked component selves of a fivefold being. In the overall textual architecture, which seeks to weld huacas of diverse origin into a single ideological kindred, the five heroes were felt to have established what were apparently large fraternal sections of society. The main teller of the Paria Caca mythology thinks of the landscape as divided into domains of influence corresponding to the various large highland-derived groups, each defined by putative descent from a persona or child of Paria Caca. These domains are not said to be contiguous territories. They emerge in the form

of mythic trajectories: the paths of heroes, along which the heroes achieved society-defining deeds and left areas of dominion over specific lands and people in the scattered or "archipelago" form known from secular and bureaucratic sources.

Yuriy/yumay: concepts of human birth and descent

Colonial Andeans imagined human descent groups as continuations of the genealogy of huacas. Intermediate-sized, named ascriptive groups like the Checa "thousand" (whose viewpoint often dominates the text) were fitted into a unified regional ideology by defining their founders as the progeny of Paria Caca's component heroes. Paria Caca's followers thus construed the "original" organization as sib- or clanlike. The spacial foci of superhuman and human genealogy seem to have been symbolized by shrines called pacarinas or 'dawning places' that represented founders' and heroes' appearances on earth. Pacarina myths are by their very nature peculiar to each group; since the Huarochirí mythology as a whole sometimes seeks to coordinate the mythic legacies of several "originally" separate groups, it is not surprising that pacarinas are the theme on which Paria Caca mythology is least consistent. One example is the pacarina where a people-producing quinua plant grew (chap. 24, sec. 302); the person who composed the chapters seems uncertain how to reconcile this myth (which Taylor [1987b: 163] takes as a component predating the subsumption of origins in Paria Caca) with other accounts of emergence.

Andeans traced the descent lines of actual humans from revered ancestors, whom they credited with living on as guardians of fertility and order as long as their mummified bodies endured. We know from "idolatry" trials and the testimony of chroniclers that the cult of mummies (mallquis) formed the link between the mythology of huacas and the purportedly known genealogy of named groups of the living. One witness on trial for "idolatry" explained: "This mallqui [the mummy Guaman Cama] was a nephew of the idols Caruatarqui Urau, and Ticlla Urau, and the progenitor of this ayllo Chacas, and he was a son of Libiac ('Lightning'], and he nurtured people, multiplied them, guarded their fields and gave them money and wealth" (Huertas 1981: 104-105). Mummified ancestors lived in caves or special houses (chap. 11, sec. 155). Their progeny dressed them richly, periodically "embraced" and feted them, served them with food and sacrifices, and petitioned their approval of major transactions (Guaman Poma [1615] 1980: 1:262-271; Polo de Ondegardo [1571] 1916: 116-119). Garcilaso ([1609] 1966: 1: 76-77) mentioned that mummies of special importance were huacas and in Huarochirí it seems to be mummified leaders of successful invasions who achieved this role. The two Huarochirí chapters about the fate of the dead (chaps. 27 and 28) are evasive about mummification but difficult to understand unless one assumes that people preserved human remains in some way. A few Huarochirí passages seem to refer to mummy huacas:

The one [huaca] called Ñan Sapa was a human being.

Later on, the Inca took away the huaca himself. But they made another one to be his proxy.

This is the one that we know Sehor Doctor Francisco de Avila carried away.

They say Ñan Sapa, when he was human, wore the quisay rinri in his ears and bore the canah yauri scepter in his hands. (chap. 24, secs. 319-320)

In connecting ancestors to living humans, unilineal principles play an explicit role. The myths employ a concept of patrilineage (yumay 'sperm') used in relation to a concept of sibling or birth group (yuriy 'birth'). The preface tells us that the peoples whose story is to be told here constitute a group by virtue of sharing a father: "I set forth here the lives of the ancestors of the Huaro Cheri people, who all descend from one forefather" (pref., sec. 2). The vocative yaya 'father' as a form of address is used to express deference to any male authority, human or superhuman.

The most prominent collective ego of the myths, the Checa, thought of their group as defined by mythic descent from one of the component Paria Caca persons; it is not stated to be patrilineal, but the Yauyo subgroup called Morales as a whole is treated as male relative to the Yunca, and the persons whose acts define it are males. Within such clanlike categories, there appear to have been more clearly genealogical patrilineages uniting people with specific rights and duties. Portions of society characterized as patrilineages (yumay 'sperm') are mentioned as the holders of hereditary water rights and priesthoods (chap. 7, sec. 91; chap. 31, sec. 445).

When comparisons among sectors of a collectivity are made, whether at a vague clanlike or a more genealogical lineagelike level, they are usually made in terms of birth order within an original sibling set (yuriy). Genealogical myths from all over the world "justify existing stratifications by denying them (we are all brothers) while at the same time providing detailed guidance to inequality by distinguishing between 'elder' and 'younger' branches. They also record alliances by tying allied groups into common genealogies" (Vansina 1985: 103 ). The concept yuriy 'sibling group' carries exactly this contradiction. In all cases, siblinghood seems to imply rank even more saliently than it implies solidarity. "The Quinti thoroughly despised the Checa because the Checa were born last" (chap. 11, sec. 153). The notion of unranked kin solidarity is treated as an abstraction or mystique at the highest level of mythic generality: Paria Caca, the union of all Huarochirí-area invader groups, came into being as five unnumbered eggs. The set of eggs could be considered a sibling set, but one without birth order because not yet born. No such transcendent paradox occurs among humans. The transition from power in ovo to power in action is also the transition to rank order.

As in many Andean contexts (for example, the political organization of colonial Andean communities), the firstborn or leading member of a set (e.g., noble heads of a village's component ayllus) functions within the set as first among equals, but outside the set as the totalizing representative of it. (The political authority of curacas over ayllus, for example, seems to have worked on this principle.) So, too, among huacas. Paria Caca can be considered the first among five brothers (chap. 8, sec. 99), yet at the same time as the overarching deity of which the five are only parts. In this latter function Paria Caca is often spoken of as the father rather than the brother of the heroes. Chaupi Ñamca has the same structure (chap. 10, sec. 147).

So thoroughly does the ideology of genealogical inequality govern ritual order that the birth of human siblings whose rank order closely approached equality—that is, twins—was seen as a major anomaly warranting the immense expiatory efforts that occupy supplement I.

Naturally there is some tension between rank ascribed by descent and rank achieved through warfare. The "children of Paria Caca" fought for the military supremacy of their ethnic group, but also for precedence among themselves, and this produces anomalies; chapter 17 expresses the Checa claim to have risen above their Quinti rivals by their achievements. Sometimes a group's junior standing seems to reflect its lesser achieved power. The Concha, a local group described as an ayllu (see below), were said to have originally been a yuriy of five brothers and a sister; the resulting five groups' unequal land resources are explained as the suit of the first three brothers' having conquered energetically while the last two lagged or got lost:

. . . these two [Hualla and Calla] fell somewhat behind.

So, lagging behind, they missed the trail and headed instead toward the Yauyo country, thinking, "Maybe our brothers went over there."

A long time afterward, only after the other three brothers had finished dividing up the fields and other goods among themselves, they did come back. (chap. 31, sec. 391)

Patrilineal grouping and birth-order ranking do not exhaust the kinship content of the ideology exessed in the manuscript. Some episodes suggest that descent through females also played a large part in the tellers' idealized vision of social organition. No named principle used in myths of group origin offers a matrilineal counterpart to yumay. Yet, in rituals, there is a strong tendency to treat the engendering of females as a separate type of parentage from the begetting of males. In the fertility-giving ritual game of spearing giant dummies called yomca and huasca,

Once they'd prepared everything, they named one of the effigy bundles Yomca and set it as a target symbolizing males.

The other, the one called Huasca, they set as a target symbolizing females.

After they set them up, the men would put on their best clothing and feather ruffs called tamta, and they'd begin to let fly at the targets. (chap. 24, sec. 328)

Then they threw spears at the Huasca effigy for females, saying,

"She'll give me daughters and all kinds of food!" and then at the Yomca effigy, saying,

"He'll give me sons, agave fiber goods, and all kinds of animals!" (chap. 24, sec. 334)

Separate female huacas were credited with being great producers of female offspring: "Chaupi Ñamca was a great maker of people, that is, of women; and Paria Caca of men" (chap. 13, sec. 172). Mama 'mother' was a term of honorific address strictly parallel to yaya 'father'. Certain priestesses held rank apparently equal to that of priests (chap. 13, sec. 178; chap. 31, sec. 417; see also Avila [1645] 1918: 65). Overall, the implicit suggestion seems to be that male and female fertility run through separate channels, which cross over each other in a braidlike pattern as each generation reproduces itself sexually. The female channel of fertility was fortified by a priestly religious structure, as was the male. The Huarochirí data seem compatible with—but do not declare the, existence of—inheritance of some religious identity or obligation through parallel descent (Lambert [1977] 1980: 37; Silverblatt 1987: 20-39).

Ayllu: corporate landholding collectivity self-defined as ancestor-focused kindred

The tellers of the myths habitually described their society as built of collectivities called ayllus. In many passages the ayllu figures as the basic unit of ritual action:

... in the old days, people used to go to consult Paria Caca at night, taking along llamas or other things.

They used to go taking turns, ayllu by ayllu. (chap. 24, sec. 309)

A person's immediate religious responsibility was to his or her ayllu's senior members. When Lanti Chumpi discovered what she guessed might be a new huaca (perhaps a buried figurine or unusual stone), her first thought was to take the find to them (chap. 20, sec. 237). We know that each ayllu made its own claim to religious authority because each told its own version of certain myths (chap. 13, sec. 187).

But what was the makeup of the ayllu? The classic sources look unhelpful at first glance. The great lexicographer Diego Gonçález Holguín ([1608] 1952: 39-40) gives a definition of ayllu so broad as to include virtually all kinds of descent, kinship, and even territorial solidarity. Avila thought an ayllu was something like the Spanish kin group defined by a shared surname (Arguedas and Duviols 1966: 257), which would suggest patrilineal bias.

Internal evidence from the manuscript, however, suggests something other than a corporate unilineal principle. Chapter 7 (sec. 91) says, "There is within this ayllu a patrilineage [yumay] which bears the name Chauincho." From this we infer, first, that ayllu is a separate concept from patrilineage, and second, that an ayllu could contain more than one patrilineage. ayllu is therefore not the minimal or the only unit of descent ideology. However, the fact that the passage just cited goes on to speak of the Chauincho patrilineage as an ayllu (with a crossing-out eloquent of hesitation on someone's part) suggests that the term ayllu could subsume the concept of lineage. This makes it difficult to distinguish the two in certain instances.

We also know, because chapter 13 (sec. 187) implies as much, that a given territorial settlement (llacta; see below), which we usually gloss 'village', could have multiple ayllus: "in each village, and even ayllu by ayllu, people give different versions . . ." People understood the internal dynamics of their local communities as a play of more or less rival ayllus.

The tellers saw rights to land and other immovable assets as lodged in the ayllus: "As soon as Tutay Quiri's children had expelled those Yunca, they began to distribute among themselves, according to their own ayllus, the fields, the houses, and the ayllu designations" (chap. 24, sec. 316). This passage then tells us that even the invaders felt bound to redefine their own organization on the pattern of preexisting local ayllus, which suggests they had a high degree of corporate definition and legitimacy as well as important functions. Since it goes on to say the Yasapa ayllu people were silversmiths, one may further speculate that some ayllus practiced, or at least were traditionally associated with, occupational specialties.

So it is relatively safe to think of the ayllu as a named, landholding collectivity, self-defined in kinship terms, including lineages but not globally defined as unilineal, and frequently forming part of a multi-ayllu settlement. But what exactly were the kinship criteria of inclusion? This question, an ancient mare's nest in Andean research, yields partly to Karen Spalding's exploration (1984: 28-30, 4852; see also Castelli, Koth, and Mould de Pease 1981). Gonçález Holguin shows us that the most general sense of ayllu and its derived words is "that of grouping elements or persons together on the basis of similarity or species, or dividing up a larger group on the basis of the same criteria" (Spalding 1984: 29). "Similarity or species" could mean tax such as animal or plant species, but, when applied to people, ayllu usually meant "descendants of a common ancestor." "The term was commonly defined as any group—family, lineage, or generation—whose members were related to one anothe through their descent from a common ancestor" (Spalding 1984: 28-29), that is, an ancestor-focused bilateral kindred. Zuidema (1973: 16-21) has developed the argument toward a detailed model of ancestor classification and cultic organization.

Spalding's definition has useful corollaries. First it reminds us that, like such spatial terms as pacha, ayllu is the name of a concept of relatedness and not of an entity with specific dimensions. It has no inherent limits of scale; in principle, it applies to all levels from sibling groups to huge kindreds, clanlike groups, or even whole ethnic groups defined by reference to common origin and territory. An ayllu can readily be understood as consisting of multiple patrilineages (or, in principle, matrilineages) insofar as any given member can trace descent from the "founder" or apex via a given child of the "founder" (and so forth, in potentially segmenting ramifications). Platt ([1978] 1986: 230231) has clearly demonstrated a varied-scale usage among modern Bolivian highlanders, who reckon upward from the "minimal ayllu"—small clusters of patrilocal rural neighborhoods—up through "minor," "major," and finally "maximal" ayllus that ascend to include the entire ethnic group. The various levels of ayllu organization may each have specific terminologies, typically referring to their political functions. For example, in Platt's area, the "minimal" ayllu was called cabildo (civic council) in its political functioning, and Spalding (1984: 51) adduces a Spanish witness who understood the Inca decimal term pachaca ('hundred') to mean an ayllu of a hundred households or over, suitable by its size for treatment as an administrative entity in its own right. In the Huarochirí manuscript, the usage of ayllu terminology becomes less confusing if one recognizes that an ayllu may be part of a larger ayllu. In this sense, the "children of Paria Caca," the large (perhaps ethnic) group that forms the mythology's collective subject, a group of people "who all descend from one forefather" (pref., sec. 2), is a "maximal ayllu."

Second, as Spalding also emphasizes, for practical purposes it was not precise genealogy that finally decided who belonged to an ayllu, but rather social conduct—including political alliance—befitting a genealogically connected person. As with many concepts in the domain of kinship, ayllu may be understood partly as an ideology built up to explain patterns of behavior rooted in the residence rules, which in turn often reflect the demands of a given geographical, technological, and demographic reality. Access to ayllu-held assets (and claims to collective ayllu ownership continued far into the colonial era) was given in return for exchanged labor and exchanged ritual participation on a kinship model. One can see in the myths of Concha ayllu (chap. 31, sec. 391) that genealogical connection alone was insufficient to bestow land on the two Concha lineages that had become politically disconnected. But adoption combined with political or marital alliance was seen as sufficient to create ayllu entitlements even when there was no genealogical tie (chap. 31, sec. 403). ayllu was a political fact, and cultic practice lodged in it regulated practical matters of economy and power.

Llacta: 'village' as cultic and territorial unit

By the time the Huarochirí manuscript was written, colonial coercion had overhauled the relationships between people and territory. If the prehispanic Huarochirí region resembled those known elsewhere in the Peruvian Andes, each of its major settlements probably controlled "archipelagos" of mostly non-nucleated residence spread over the various productive tiers of the mountain slope and the coast. But by 1608 the parishes where Avila was working—for example, San Damián de Checa, focal point of the manuscript—no longer entirely followed this model. Villages like San Damián had been carved out in a scheme of forced resettlement (reducción) that Viceroy Francisco de Toledo had begun almost forty years earlier. When administrators herded Andean peasants into resettlement villages (Gade and Escobar 1982; Spalding 1984: 214-216) they fused together multiple Andean settlements—the ones called llacta in prehispanic usage—into a larger, more accessible, more governable and exploitable "Indian town" on a Spanish plan with plazas, churches, and streets laid out on a grid. Most of the places with Spanish (saint) names in the manuscript belong to this Toledan reordering of territory.

Why, then, did people keep orienting their religion around a map that no longer strictly fit the productive space of politics and economy? After the initial coercions of resettlement, from the 1580s on, older relations to the landscape partly reasserted themselves both in practice and in ideology. In practice, productive efficiency and the chance to hide from tribute collectors enticed people back upward into the non-nucleated pastoral hamlets of the heights and downward into relatively hidden farming settlements in river canyons (Málaga 1974). Nor, in imagination, did the establishment of a Christian sacred geography centered on parish churches empty the landscape of non-Christian meaning. The Huarochirí storytellers c. 1600 saw all around them their parents' ruined "old settlements" (pueblos viejos, a common term in papers of the period) and their ancestors' stone "houses of the dead." The pre-resettlement scheme of territoriality, a mental map of social groups attached to place-deities and localized ancestors, still formed a complete and intelligible shadow-geography projected onto the landscape that colonial organizations had already reshaped de facto. Immense huaca-studded spaces of canyons and high tundra, fields and trails, embodied an Andean world view at least as cogently as the small, dense space inside the new churches figured Christianity. Since outdoor space was also the space of work and livelihood, the very cycles of herding and farming continuously retaught what Sunday sermons sought to erase.

The commonest term for the anciently defined settlement, llacta, is not the simple equivalent of 'town' or 'village', which denote a portion of territory or the legal corporation that governs it. A llacta in its old sense might be defined as a triple entity: the union of a localized huaca (often an ancestor-deity), with its territory and with the group of people whom the huaca favored. The word llacta could thus be used to mean the deity that was master of a settlement. In chapter 24 (sec. 325), where the original text says that the "llactas" divided up the llama herds among themselves, a marginal note clarifies that this means the "idols," that is, the huacas that defined the llactas, and our translation uses the latter sense.

The word llactayuc (glossed 'aborigine', 'native', or 'founder') therefore means something more complex than 'original resident'. It implies both being possessor of a local huaca's sanctum and being possessed by it. When the heroes of the myths conquered lands and peoples, they also acquired local ritual obligations and even, it seems, grafted themselves into the genealogical categories reckoned from the huaca: "As we said in another chapter, this land was once all full of Yunca. As soon as Tutay Quiri's children had expelled those Yunca, they began to distribute among themselves, according to their own ayllus, the fields, the houses, and even the ayllu designations" (chap. 24, sec. 316).

The chain of human movements and transformations by which the Checa people explained their social organization emphasizes at every stage a pattern of huacas among whose territories human groups move and fight. huacas might travel on the way to establishing their dwellings but once victorious they had—they were—their locales, and it was the deity-locale that gave wealth and identity to human groups. The Checa explained all changes, both prehispanic and recent, with reference to the geography and the relative fortunes of huacas. To the tellers, the explanation of society and its genesis was written out in the landscape—even where Spaniards had wrecked every visible monument or substituted crosses for huacas.

Even now llacta is essentially the name of a relationship, not of a type of settlement. Like ayllu, the term implies no particular scale. In modern Quechua one calls any unit from one's hamlet to one's country "my llacta." Since the word gives no suggestion of size, either demographic or spatial, the manuscript usually leaves us guessing as to whether a given llacta is a hamlet, village, town, city, region, or country. The answer will come, if at all, from archaeological work as yet undone. Most of the actions in the myths seem from context to concern agricultural villages, and in most cases we have glossed llacta as 'village'. But 'town', 'city', 'region', and 'country' are hardly out of the question. When one visits the immense ruined city of Cajamarquilla, within the space of the manuscript but predating it, one wonders if the mythic scene could not have been far more urban than generally imagined.

The original text

The Huarochirí manuscript is ff. 64r-114r of manuscript number 3169 of the Biblioteca Nacional, Madrid. The original lacks title, date, and author. It is the fourth item among six manuscripts about Andean religion, from Francisco de Avila's own collection, all bound together. Among these writings are whole texts and abstracts from some of the most important sources on the subject, such as the Relación de las fábulas y ritos de los Incas ([1575?] 1959) of Cristóbal de Molina "cuzqueño" and the original of Joan de Santa Cruz Pachacuti Yamqui Salcamaygua's partly Quechua Relación de antigüedades deste reyno del Perú ([1613] 1968) as well as works by Juan Polo de Ondegardo and Garcilaso Inca de la Vega. It also contains the manuscript of Francisco de Avila's Tratado, a translation, or more accurately a paraphrase with digressive commentary, of the same material contained in chapters 1-7 of the Huarochirí manuscript.

The Huarochirí manuscript is written in an ordinary competent scribal handwriting. The chapter divisions and titles used here are from the original. But the original does not have a consistent system of paragraph divisions or sentence divisions: some sentence boundaries are ambiguous. Various pen strokes resembling parentheses, commas, and other small symbols occur in the text, but they do not seem to add up to a consistent system of punctuation. (Scribal papers of the period normally lack consistent punctuation.) The text is in a variety of Quechua with the exception of the Spanish chapter headings of chapters 1-6, some borrowed Spanish words, and some marginal queries and annotations, probably by Father Avila. Irregular pagination and handwriting size indicate that the manuscript may have been compiled in noncontinuous bursts of effort. The text comes "from the hand and pen of Thomás," according to a marginal note over halfway through the text (chap. 23, sec. 291), but we do not know anything about "Thomás" save that he seems (from his Quechua-influenced errors in Spanish) not to have been a native-or a particularly accomplished-Spanish speaker. Whether he was only an amanuensis or was also the compiler of the testimonies is unknown. Thomás' practiced handwriting suggests he may have been an escribano de naturales (bilingual scribe), village council scribe, or other native functionary of the colonial regime.

The date of the Huarochirí manuscript is debatable. Avila's Tratado partially paraphrases the manuscript, or a draft of it, and bears the date 1608, so the manuscript seems prima facie to predate the end of 1608. But by how much? In chapter 9 (sec. 133) a narrator tells us it is cay pisi huatallarac, which at first glance seems to mean 'scarcely a year', since Father Avila came onto the San Damián scene. Were that reading unambiguous, we could agree that the date must be 1598 because, as Duviols ascertained, Avila arrived in Huarochirí in 1597 (Arguedas and Duviols 1966: 235). But cay pisi huatallarac could also mean 'just a few years', admitting a later date. And as Antonio Acosta's important researches (1987a, 1987b), summarized in the next section, suggest, there are strong reasons to consider dates considerably later than 1598.

The possible genesis of the text in the local conjuncture

The common attribution of the manuscript to Father Francisco de Avila is erroneous, but he certainly was a key figure in the local conjuncture that forced the manuscript into being. The following paragraphs summarize recent research by the Spanish historian Antonio Acosta, whose archival inquiry (1987b: 551-616) brings the local crisis into clearer focus and suggests a more plausible genesis for the text than the self-serving version available from Avila's own testimony.

The pioneer of seventeenth-century anti-huaca persecution was born in Cuzco probably in 1573, the abandoned, perhaps illegitimate baby of an unknown couple. The Spanish couple who found him at their doorstep gave him the Spanish name Francisco de Avila Cabrera. Avila later, but dubiously, claimed he knew his ancestry to be noble; others claimed he was half-Indian, and to his misfortune this became the common opinion. Acosta (1987b: 557-558) finds reason to question it. If Avila thought himself mestizo, it would have been against his interest to say so given the racial impediment to ordination during his youth, so his silence on the point cannot be decisive. One witness in 1610 called him mestizo, without proof, and a 1641 witness thought he looked Indian (Acosta 1987b: 556-557 note; Spalding 1984: 253-254). But these accusations of mixed birth came after Avila had risen to prominence and could well have been falsehoods intended to sabotage his career. Such doubts leave room to question the common supposition that his vengeful attitude to Indian religion grew from the pain of mixed (therefore impure and stigmatized) birth or from resultant abandonment.

Avila showed talent in a Cuzco Jesuit school and went to Lima to study at San Marcos University in 1592. In 1596 he was ordained and in 1597 posted as curate to San Damián de Checa, one of the reducción parishes near Huarochirí. It was a plum appointment. Avila's Jesuit connections probably had much to do with his career there, for the Jesuits over twenty years earlier had pioneered what they recognized as an unusually rich mission field—one where Andean religion had already shown strong resistance to some forty years of intermittent Catholic intervention. One passage (chap. 20, sets. 244-247) sketches the fluctuating, but never extinguished, fortunes of the huaca priesthood through three generations of colonial leadership.

Even within normal limits of law, a curate like Avila had access to ample native labor. But, in addition, by the date of Avila's arrival it had become usual though illegal for priests serving Indian parishes to parlay their ecclesiastical holiday levies and salaries, combined with legal leverage over native nobles, into business enterprises large enough to rival the incomes of rural Peru's opulent semi-feudal civil elite (LavaIIé 1982). In time, it seems, this sort of practice led Avila into controversy. When he was a new curate, in 1598, an inspection praised his pastoral work. But a sign of conflict appeared in a 1600 "secret inspection and hearing" that turned up accusations of commercial abuses. Although he was officially cleared, the inspection reveals that like many rural clerics he had begun to make local enemies.

Four later inspections left Avila's record superficially restored. In 1607 Avila gathered testimonies on his "life and morals" to buttress his continuing appeal for a higher post. At the last minute, how ever, he met a fateful opposition: a faction of his native parishioners mounted an ecclesiastical lawsuit against him. Both colonial chieftains and commoners accused him, and their accusations were more numerous than in ordinary cases of this sort (frequent, to be sure, in the period). They accused him of absences from his parish (perhaps, Acosta suggests, he went to Lima in connection with his studies toward the doctorate; 1987b: 574), of charging excessive fees, and of exacting illegal labor levies. Like many other curates, he was said to collect huge amounts of native crops and sell them for private profit. Maybe natives tolerated this because priests of the Andean huacas, too, had enjoyed huge gifts of produce (Acosta 1987b: 576). But Avila may have gone too far in helping himself to his parishioners' labor, which he used to support his partly illegal business enterprises in gunpowder, charcoal, and textile manufacture and to build himself a house in Lima using beams he made his parishioners remove from the roofs of their pre-resettlement village. In a period of declining native population, curates' rising and increasingly arrogant expectations of cheap or unpaid labor acutely angered some natives. It was apparently Avila's project to open a new textile factory at native expense that finally provoked chiefs of the "thousand" of Chaucarima to litigate against him.

The Indians went beyond economic grievances to introduce a doubt about Avila's religious and moral regularity, alleging sexual abuses of various native women (some married) and the fathering of an illegitimate son. He even forced Indian women, "under physical threats, to suckle at their breasts the same puppies which ... when grown, chased and killed the Indians' chickens" (Acosta r 987b: 574). He beat up villagers, including his sacristan, for not serving his desires zealously enough. Once at a christening he hurled the oil at his sacristan's chest. Accusers said he encouraged Indians to give their venerated ancestors silver (see chap. 21, sec. 262) and then himself collected some of it (Acosta 1987b: 572). If true, this suggests that in the early days of his curacy Avila had no urgent scruples about a modus vivendi with "idolatrous" religion. Apparently coexistence entailed the curate's acquiring traditional privileges of huaca priests—access to women, labor donations, crop gifts—and, in return, tolerating (or even profiting from) worship of huacas on Catholic holidays (see chap. 7; chap. 9, sec. 125).

Avila spent some time in a church prison while charges were pending. But his emissaries pressed the natives to recant their accusations, and some did. The record of the trial contains a neat sheaf of the recantations, witness by witness, each with a cover sheet bearing Avila's elegant handwriting. One of Avila's allies in collecting these was the same Cristóbal Choque Casa whose visionary combats chapters 20 and 21 glorify, at that time an obscure son of native nobles. He prepared one of the recantations in Quechua (Taylor 1985: 180). In Acosta's judgment (1987b: 596), "It is believable that the narrative [i.e., the Huarochirí manuscript] could have been compiled at Avila's order in 1608, when his lawsuit with his Indians was in full swing, maybe on his emergence from prison and as a part of his reaction against the Indians who had accused him." Avila probably secured the testimony through one or more native cat's-paws who either interviewed anti-huaca natives or else secured the testimony of huaca believers by lulling them into unawareness of their testimonies' future utility. Cristóbal Choque Casa probably played a key role. It is hard to guess what parts fear, self-interest, and sincere conversion (Cristóbal's or others') played in the extraction of testimonies. Certainly intimidation played a part; for example, Avila made a point of interrogating Indians made vulnerable by frightening illnesses (Acosta 1987b: 600-601, 603). Although in later years Avila never mentioned the manuscript's existence, he probably did use it in hunting down huacas.

When the court responsible for the Indians' suit took depositions in San Damián, a decision only minimally damaging to Avila appeared likely. But instead of letting procedure take its course, he seized the offensive by asking the Ecclesiastical Chapter of Lima to authorize an inquiry into "idolatries" under canon law. Avila, according to his own much later testimony written in 1645 (1918), then led the ecclesiastical judges to the huacas Llacsay Huancupa (perhaps equivalent to Llocllay Huancupa, chaps. 20 and 21), Qqellccas Ccassu (probably quillcas caxo or 'Engraved Rod', chap. 24, sec. 320), and Maca Uisa (chaps. 18, 19, 23). In later testimonies Avila claimed his campaign had arisen only from disinterested zeal, but Acosta's research makes it believable that the campaign was undertaken in revenge against the natives who had accused him of venality and immorality.

According to the 1645 testimony (Arguedas and Duviols 1966: 220), it was Cristóbal Choque Casa's revelation that the feast of the Assumption in Huarochirí for 1608 would be used to cover, a rite of Paria Caca that provoked Avila to undertake massive anti-huaca campaigns. It is likely, of course, that in reality nothing new was being revealed. The only new element was Avila's urgent need for favorable publicity. Avila intensified his sleuthing in 1608 and procured public confessions of "idolatry" in a parish meeting that stirred and mobilized his pro-Christian allies. He asked the Jesuits for assistance in confessing a deluge of penitents and was sent two helpers. They pioneered the routine of breaking images, burning mummies, extracting public confessions, and punishing believers that was to become the periodic scourge of Lima archdiocese Indians until at least the 1660s. Avila was able to collect a great deal of stolen religious gear and forced testimony. By September 1609 he was amply ready to dictate his answer to the native accusations, laying the groundwork for a false autobiography that painted the accusations as a reaction to his anti-huaca zeal.

It was also during this crisis that Father Avila prepared the Tratado, that unfinished pamphlet retelling and commenting on the same myths that make up chapters 1 through 7 of the Huarochirí manuscript. The relation between the redaction of the Huarochirí manuscript and Avila's work on the 1608 Tratado has been debated between Hartmann (1981) and Taylor (1982, 1987b: 17-18). The former thinks it probable that Avila based his treatise on the manuscript, while the latter thinks that the two are separate workings of a prior source probably consisting of interview notes. In either case, it is likely that Avila intended the Tratado as a readable exposé designed to win Spanish support for his career. At the time of the Tratado's composition Avila needed public support because the archbishop of Lima, Toribio Mogrovejo, had set policy against the aggressive persecutions that Avila favored.

But, luckily for him, an opponent of gradualist policy toward "backsliding" Indians and a bitter foe of huaca religion, Bartolomé Lobo Guerrero, succeeded to the Archbishopric of Lima in October 1609. According to Taylor (1987b: 18), Avila left continued compilation of anti-huaca intelligence to an Indian associate. Although he did read and annotate part of the product, the Huarochirí manuscript, he never fully edited it.

A bare ten days after Lobo Guerrero's accession, Avila seized the moment to make his case behind the archbishop's closed Lima doors. Shortly afterward he staged a public propaganda spectacular. Avila had carted a load of huacas and mummies "800 years old" into Lima, perhaps the ones he tracked down thanks to the testimonies recorded in the manuscript. Amid immense pomp and publicity he directed a giant auto-da-fé in Lima's great cathedral square on December 20, 1609. Thousands of natives were forced to attend. A spectacle of music, Quechua sermons, and whipping of believers culminated in public sentencing. Among the victims was Hernando Paucar, the same ex-priest of Chaupi Ñamca and early denouncer of huacas whose confessions had opportunely helped Avila. Paucar was tied to a post, lashed two hundred strokes, and condemned to confinement among Jesuits in faraway Chile. Avila then burned the huacas and the ancestral mummies, to the inconsolable grief of those who felt themselves orphaned.

Four days later an ecclesiastical judge absolved Avila of all the accusations pending. Moreover, he was granted the title of "visitador [traveling judge] of idolatries."

These events marked the opening salvo of the "extirpation of idolatry" campaigns, in which Avila and his Jesuit allies developed standardized methods to destroy the partly clandestine forms of Andean religion that had grown out of the long confrontation with Christianity (Duviols 1972). Within a year of becoming visitador, he executed repressive campaigns in the main parishes that the Huarochirí manuscript mentions and collected, as he claimed, some 5,000 "idols." In 1611 he went with a Jesuit crew to climb the heights via Yampilla just outside Huarochirí (chap. 31, sec. 409). The climbers trekked for several days up what may have been the ancient pilgrimage route. After the ascent they followed a stairway hewn into the rock (Bonavia et al. 1984), up to what Avila believed to be the shrine of Paria Caca himself. They demolished everything they could. Then the extirpators "put in its place a cross and in the afternoon they returned to San Lorenzo de Quinti, where [local people] received him with illuminations, and the Indians said in their own language, 'Paria Caca has died"' (Duviols 1966: 224).

Avila continued to campaign widely and became an influential figure in the Jesuit-centered antiindigenous movement that intermittently lashed the archbishopric for most of the seventeenth century. While legally separate from the Inquisition, and in some respects different from it, the "extirpations" were to inflict on Andean society some of the same sufferings, and the same clandestinity, that the Inquisition visited on Iberian heterodoxy. Avila's track of destruction crossed the wanderings of the great native chronicler Felipe Guaman Poma de Ayala ([1615] 1980: 3:1017, 1022-1023), who met some of his victims and, notwithstanding his own dislike of huaca religion, cursed Avila for his greed and mercilessness: "Oh what a fine doctor, where is your soul? What serpent is eating you?"

But if many of "the people called Indians" had reason to hate Avila, many of the Spanish, and not just his Jesuit allies, had cause to rejoice at his success. One reason was that huaca hunting appealed to clerics' and laymen's greed for treasure. Spanish law allowed confiscation of pagan deities' wealth, and the persecutory movement that Avila promoted opened a rich vein. Guaman Poma was among those who suspected that Avila's motives included the theft of precious objects belonging to huaca cults ([1615] 1980: 3:1022-1023). The anti-Andean campaigns also offered career opportunities to Peru's numerous churchmen at a time when the church was hugely overstaffed relative to its parish infrastructure.

Avila built a long and prosperous career on these campaigns and was to win renown for his antihuaca scholarship as well as for his militancy in missionizing. In 1611, at Archbishop Lobo Guerrero's request, he wrote a report on "idolatrous" domestic rituals and local deities. After a series of attacks on deities mentioned in the Huarochirí manuscript and in other places, in 1615 he wrote at the request of the new viceroy, Principe de Esquilache, a project on the means to achieve "real conversion." The next year he conducted a gigantic tour of repression affecting some 35,000 persons. He wrote extensively on ways to influence and intimidate huaca worshipers, emphasizing both intimate persuasion and institutional ways to confine those sorts of Indians (notably the sons of native nobles) thought prone to lead or protect clandestine worship.

In 1618 he won a mid-ranking post in the Archbishopric of La Plata (seated in Chuquisaca, now Sucre, Bolivia) and held the job for fourteen years. Why the church hierarchs decided to employ him in a place so far from his political base is unknown; nor is much known about his career there. In 1632 he returned to the Cathedral of Lima to serve the newly appointed archbishop, Hernando Arias de Ugarte, an old acquaintance, as a canon of the Lima Cathedral. He came to be known as a durable eminence of the church, popular among Spaniards for his charitable donations. Almost at the end of his life, he petitioned to enter the Jesuit order, but his alleged mestizo background was used to thwart him. When in 1641 Archbishop Pedro Villagómez saw fit to remobilize the old persecutor as instructor for what was to become the second great wave of "extirpation," Avila seized the occasion to build his own monument as the "discoverer" of crypto-Andean heterodoxy. The self-portrait he painted in the 1645 preface to his Tratado de los Evangelios (Treatise on the Gospels) enduringly and, as Acosta shows, misleadingly influenced his historic image. The treatise reached the press in 1648, the year after his death. Its two volumes of Quechua and Spanish sermons, a little-known monument of literary Quechua's Baroque florescence, memorialize Avila's vision of a Counter-Reformation culture in Quechua.

Some twenty-three years after the making of the manuscript, Don Cristóbal still figured in a Concha lawsuit as a minor native official (Taylor 1983: 266, note; the date of Choque Casa's signature is 1631, not c. 1660 as Taylor holds). Certain of the curacas mentioned in the manuscript appear in tribute records and lawsuits both before and after the manuscript. In general the careers of Cristóbal Choque Casa and the other Indian makers of the Huarochirí manuscript remain obscure.

Among provincial religious sources the Huarochirí text has no peer, but it has many companions. Detailed testimonies about peasant and provincial belief appear, for example, in the "extirpation of idolatry" trials of which the Huarochirí crisis (chaps. 20-21) was a forerunner (Acosta 1987a; Duviols 1972, 1986; Millones 1967; Silverblatt 1987). Some of the Catholic priests who organized "extirpation" themselves wrote monographs on the Andean provincial religions and how to persecute them (Albonoz [1583?] 1984; Arriaga [1621] 1968; Avila [1608] 1966). Many missionary treatises and reports from the field reveal local cults and the memories of older practice in vivid detail (for example, Calancha [1638] 1974-1982; Hernández Principe [1613]1919, [1622]1923). After the waning of the "extirpation" campaigns, the record becomes thinner but still workable through, for example, trials of shamans implicated in political assassinations via magic (Dammert Bellido 1974, 1984; Millones 1984; Salomon 1983).

There may be room for doubt about some extirpators' personal sincerity, but little doubt that they played on widely believed ideological propositions. Local social conflicts like those of Huarochirí c. 1598 conspired with endemic conflicts arising from challenges to the Spanish state's commercial and religious hegemony to infuse in Jesuits and other anti-Andean militants a grim seriousness. Fear of "heresy" (meaning Protestantism and the northern European powers beginning to rival Spain) and enduring hatred against Iberia's Muslim and Jewish cultures, long since driven into clandestinity, helped clerics persuade the state that the fight against Andean religion contributed to a decisive world-historical struggle. The papers born of this effort offer ethnographic evidence, but because of their heavy ideological freight must be read with caution.

Given these facts, one might expect to find the particulars of Andean religion, and eventually the very fact of Andean worship, distorted in the direction of familiar European fantasies of anti-Christianity: satanism, the black mass, and so forth. Such distortions did happen, especially at later dates and more urban locations than those of the Huarochirí manuscript (Silverblatt 1987: 159-196). Likewise, one might expect that missionaries would picture Peruvian huacas in the image of more familiar "gentile" deities or "idols"—for example, the dei of Greco-Roman antiquity. This also did occur (MacCormack i985).

But to a surprising degree the testimonies of the victims retain freshness and unfamiliarity that give prima facie evidence of an origin other than Iberian demonology or the classical legacy as enshrined in seminary curricula. Perhaps because many of them were provincials lacking the know-how to package and process their culture in terms familiar to Spanish speakers, the myth-tellers in the Huarochirí manuscript created an image still largely framed by conceptual categories proper to local thought. The Huarochirí stories retain for us a certain irreducible strangeness, resistant to translation because, unlike the preprocessed Inca lore available in chronicles, they were seized by Spain but not made for it.

Previous editions of the Huarochirí manuscript

The following are the extant complete editions of the text. No attempt is made to cover excerpts, retranslations, or popularizations. (For critical discussion, see Hartmann 1975, 1981; Taylor 1982.)

Adelaar, Willem F. H. 1988. Het boek van Huarochirí: Mythen en riten van het oude Peru zoals opgetekend in de zestiende eeuw voor Francisco de Avila, bestrijder van afgoderij. Amsterdam: Meulenhoff.

Dutch translation (no Quechua) with introduction and glossary.

Arguedas, Josh María (trans.) and Pierre Duviols (ed.). 1966. Dioses y hombres de Huarochirí: Narración quechua recogida por Francisco de Avila [¿1598?]. Lima: Instituto Francés de Estudios Andinos/Instituto de Estudios Peruanos. [Republished 1975] México: Siglo XXI.

First full and published Spanish translation. Includes Quechua. Although less accurate than Taylor's translation of 1987, Arguedas' translation is esteemed for literary merit. Duviols' biobibliographical essay on Avila, questioned by Acosta, remains important for its pioneering historic inquiry and for its primary source appendices. These include Avila's Tratado and other papers relevant to Avila, such as extracts of important Huarochirí reports by mostly Jesuit observers. Mexican republication is incomplete.

Galante, Hipólito (ed. and trans.). 1942. Francisco de Avila de priscorum Huaruchiriensium origine et institutis ... Madrid: Instituto Gonzalo Fernández de Oviedo.

Contains the thirty-one chapters but not the appendices. Introduction in Latin, facsimile, transcription, critical notes, glossary of hispanisms, Latin translation, Spanish retranslation by Ricardo Espinosa M.

Mejia Xesspe, Toribio [unpublished translation]

Toribio Mejía Xesspe, a bilingual scholar who worked closely with Julio C. Tello in the pioneering days of Peruvian prehistoric archaeology, left in his posthumous estate an as yet unpublished version of the thirty-one chapters of the Huarochirí manuscript. Prepared in 1941-1943 it includes a rephonologization, a Spanish translation attempting morpheme-by-morpheme correspondence, an incomplete "literal translation," and an incomplete "free translation" with pictorial sketches (Szeminski 1989).

Szeminski, Jan (ed. and trans.). 1985. Bogowie i ludzie z Huarochirí. Cracow/Warsaw: Wydawnictwo Literackie.

Polish translation with brief introduction and general glossary. The title means 'Gods and Men from Huarochirí'.

Taylor, Gerald (ed. and trans.). 1980. Rites et traditions de Huarochirí: Manuscrit quechua du début du 17e siècle. Série Ethnolinguistique Amérindienne. Paris: Editions PHarmattan.

Bilingual edition containing introduction, transcription with variorum notes, and French translation with interpretative notes providing original solutions to some dialectological and lexical problems. Interpretative glossary, bibliography, and supplementary notes follow.

Taylor, Gerald (ed. and trans.), with Antonio Acosta. 1987. Ritos y tradiciones de Huarochirí de] siglo XVII. Historia Andina, no. 12. Lima: Instituto de Estudios Peruanos/Instituto Francés de Estudios Andinos.

Bilingual edition containing interpretative introduction, transcription, "phonological reconstitution," Spanish translation, bibliography, several indices (to Quechua words in the translation, hispanisms in original, names of places and groups, names of huacas and heroes, names of rites, and names of historical personages). Copious critical, variorum, and interpretative notes closely address dialectological, geographical, and lexical problems. The preferred study edition.

Trimbom, Hermann (ed. and trans.). 1939 Francisco de Avila: Dämonen and Zauber im Inkareich. Quellen and Forschungen zur Geschichte der Geographie and Völkerkunde, vol. 4. Leipzig: K. F. Koehler Verlag. [Republished with additional introduction and notes as] Hermann Trimbom and Antje Kelm (eds. and trans.). 1967. Francisco de Avila. Quellenwerke zur Alten Geschichte Amerikas Aufgezeichnet in den Sprachen der Eingeborenen, vol. 8. Berlin: Ibero-Amerikanisches Institut/Mann Verlag.

The earlier (1939) edition was the first publication of the original Quechua. Research for the 1939 edition was interrupted by the Spanish civil war; later bombing destroyed most copies. Contains preface, bibliography, introduction, transcription, German translation with notes, glossary, exegesis (1967), index of proper nouns. Hartmann (1975, 1981) judges that superior accuracy justifies the republication.

Urioste, George (ed. and trans.). 1983. Hijos de Pariya Qaqa: La tradición oral de Waru Chiri (mitología, ritual, y costumbres). 2 vols. Foreign and Comparative Studies, Latin American Series, no. 6. Syracuse: Maxwell School of Citizenship and Public Affairs, Syracuse University.

Bilingual Quechua-Spanish edition: introduction, transcription with variorum notes and marginal material, translation with variorum and interpretative notes, indices (to proper nouns, Quechua words in the translation), bibliography.

An unreliable English version of Avila's Tratado by Clements R. Markham (Avila [1608] 1875) was the first modern edition of the Huarochirí mythology.

The character of the present translation

This book is a fresh working not based on any previous translation (including Urioste's 1983 Quechua-Spanish edition). Our study started from a newly made transcription from microfilm (reproduced here), collaboratively rendered into English by the co-authors. It is meant as a reader's edition rather than a study text. Scholars in need of a study edition should turn to Gerald Taylor's Quechua-Spanish version (1987b), which offers abundant apparatus as well as fuller contextual references. The 1966 Arguedas and Duviols edition also contains indispensable primary sources and bibliographic detail. The present version refers to other translations by way of mentioning some significant discrepancies, difficulties, or contextual findings but does not provide full coverage of alternative renderings.

We hope to address nonspecialists and have striven not only for accuracy but also for immediacy. We have intentionally left enough unresolved strangeness on the surface to keep a reader aware that this text is untranslatable in all the usual senses, and perhaps a little more untranslatable than most.

Aside from the hazards intrinsic to all translations, Huarochirí readers face some specific ones.

Language substrates and non-Quechua languages

Oddly enough, it is not known exactly what relationship obtained between the Quechua of the manuscript and the language in which Huarochirí religious life was conducted. Many dialectological and linguistic questions about the manuscript remain pending.

Even the origin of the Quechua dialect used in the text itself is less than obvious. Its general affiliation is clear: it is one of the many and far-flung kinds of Quechua grouped as Quechua A or Quechua II. At least one member of this group spread widely through Andean America long before the Incas, and the Incas promoted a Quechua II dialect (probably not identical to Cuzco Quechua) as their administrative tongue. Perhaps working from this precedent, the Spanish invaders styled a widespread Quechua II for perhaps several overlapping Quechua II dialects) the "general language" (Cerrón-Palomino 1985: 552-553; Taylor 1985: 158-160). In ecclesiastical councils especially, Spaniards promoted it for colonial use through efforts such as standardizing orthography and providing norms for translation and nontranslation of religious concepts.

qqq

"General" Quechua functioned c. 1600 as a lingua franca shared by many linguistically diverse peoples including Spaniards, among them clergymen who used it to simplify missionary work in a dialect landscape already reverting to pre-Inca diversity. Colonial native chiefs and others who traded on relationships with the Spanish also relied on it. On the whole, this church-influenced "general" Quechua is the language of the manuscript. The person who actually did the writing appears to have learned the art of Quechua writing for ecclesiastical purposes. But precisely because it was a partly artificial lingua franca—many speakers' second or third language and few if any speakers' first language—specific examples of "general" Quechua would normally be affected by underlying patterns of local speech. Linguists disagree on whether the text's peculiarities indicate an attempt to render speech similar to Cuzco Quechua (Urioste 1973: 4) or similar to the Quechua of modern Ayacucho (Hartmann 198x: 189). Mannheim (1991: 195) notes phonological reasons for questioning whether the manuscript records any member of the group of south highland dialects to which both Cuzco and Ayacucho belong.

The question actually goes far beyond Quechua dialectology. There is room for doubt about whether any Quechua was the language of religious practice or of the original testimony. In Huarochirí the particular version of Quechua that functioned as a "general language" for such churchmen as Father Avila still, at the probable date of the manuscript, thinly overlay at least one non-Quechua Andean tongue. Twenty-two years before the manuscript, Diego Dávila Brizeño ([1586] 1965: 155) noted that the common folk did not all know Quechua. A 1577 Jesuit report tells us that missionaries needed to have Quechua sermons repeated in a local language "because the women there don't know the general language" (Arguedas and Duviols 1966: 245). Almost two centuries after the Huarochirí manuscript, another visitor observed that local use of the "general language" sounded peculiar and "mixed" (Taylor 1983: 270, 273). It is therefore highly likely that the religious life of the generation Avila persecuted had been conducted at least partly in a language other than Quechua.

So the text probably stands at some distance from habitual and traditional usages in religious speech. We do not have any specific proof that the texts were translated from a local tongue into Quechua, but the possibility cannot be discarded. At a minimum it seems likely that processing of local discourse by a person fluent in "general" Quechua played a role in creating the text.

The following are languages that appear to influence and in some cases to underlie the Quechua text.

Quechua other than the "general" dialect

"Thomás" the scribe (or persons dictating to him) probably knew at least one dialect of Quechua other than the one the church promoted as "general." It may have been influenced by an Aymara-like tongue (see below). It is also a possibility, though a difficult one to demonstrate, that local Quechua was influenced by a coastal speech proper to the Yunca people whom Huarochirí folk considered aboriginal; their language, too, may have been a Quechua different from the "general." Finally, some usages in the Quechua of "Thomás" (e.g., chacuas 'old lady'; chap. 21, sec. 261) are attested in Quechuas of the Quechua I or Quechua B group native to the central Peruvian highlands. Quechua I languages differ widely from the Quechua of the text, but other colonial examples of mutual influence are known. In some cases the writer has hesitated between words of the Quechua II or A group, as, for example, in the repeated crossings-out of punchao 'sun' and substitutions of another word for 'sun': ynti. The discrepancy may have to do with Inca versus non-Inca religious vocabulary.

Language(s) of the Jaqi (Aymara) family

The terms "Jaqi" and "Aru" denote a group of languages today represented most notably by Aymara, an important language spoken from the Lake Titicaca basin southward into Bolivia. Jaqi tongues share some lexicon with Quechua but are entirely separate languages. Although Quechuas have displaced Jaqi tongues from a large part of their formerly enormous range, certain Jaqi languages (Kawki and Jaqaru) were and to some degree still are spoken in two areas close to the Huarochirí terrain (Briggs 1985: 546). Both areas are in Yauyos Province, from which Huarochirí people were thought to have immigrated (chap. 23, sec. 297; chap. 24, secs. 305-309; chap. 31, secs. 391, 408, 443). Taylor has argued that some of the testimony may have been given in a Jaqi tongue, because Aymarisms are common in the ritual terminology (e.g., the names of the two greatest celebrations described, Auquisna and Chaycasna). He also notes that certain "Aymara-type" phonetic alternations (ñamca/ñamoc, etc.; Taylor 1985: 162) commonly occur in words connected with Huarochirí religion, suggesting Jaqi interference. These Jaqi-like phenomena offer the strongest clue for identifying the ethnic language of the mythtellers. It is probable that at a minimum the text has been modified from a more Jaqi-influenced speech toward "general" Quechua.

Non-Quechua, non-Jaqi native lexicon?

A few common nouns and many names of persons and places do not seem to derive by any evident route from Quechua, Jaqi, or Spanish. There may be an additional influence from an unknown ethnic tongue, perhaps predating the supraethnic Jaqi and Quechua diffusions. Huallallo Caruincho's name, the hugi monster he created, and the untranslated common nouns callcallo (chap. 31, secs. 413, 415) and llaullaya (chap. 21, sec. 271) may be examples. If so, the persistence of such words in both Quechua and Jaqi cultic vocabulary may eventually yield a clue to the antiquity of Huarochirí religious categories.

Spanish

Father Avila reported that most Huarochirí people knew at least some Spanish (Arguedas and Duviols '1966: 255). However, as Urioste notes, not all of the many Spanish words that found their way into the manuscript are there for obvious reasons. Some words do lack Quechua equivalents and are therefore hardly surprising: cauallo 'horse', yglecia 'church'. Others overlap Quechua terms but "cover a different semantic space" (Urioste 1982: 106). For example, animalcona ('animals'; Spanish with a Quechua plural marker) combines categories of wild and domestic beasts that, in Quechua, are named with separate words. But a few do have close Quechua counterparts and are not used out of any obvious necessity: doze año 'twelve years' (in Quechua a plural numeral obviates the need for a pluralizing suffix on the head of the phrase, hence the singular form año), gato montés 'wildcat', and so forth, which have close and obvious Quechua equivalents. It is not known why the writer chose them.

Although the text was probably made as a tool for excising Andean belief from the religion of converts, it sometimes uses Christendom's lexicon to name Andean religious categories. All the seven occurrences of saçerdote and saçerdotisa (Spanish for 'priest'/'priestess'; chap. 13, secs. 178, 183; chap. 18, sec. 224; chap. 20, sec. 252; chap. 21, sec. 273; supp. I, secs. 462, 466) refer to priests of huacas. Saçerdotisa suggests an analogy from the Renaissance lore of Greco-Roman antiquity, but some usages are more markedly Christian: the preface (sec. 2) categorizes the regional belief system as a fe 'faith', implying that as a whole it is an entity comparable in scale and kind with the True Faith. Given the tellers' and writers' free use of specialized Andean religious terminology, they probably did not include these words for the convenience of Spanish readers. It is more likely that there already existed some habitual code of correspondences perhaps arising unconsciously from the habit of expressing Andean ideas in a fashion responsive to Christian hegemony.

Sometimes the influence of Spanish norms operates in a hidden fashion by affecting choice between Quechua words. This is especially notable in sexual lexicon, where the pro-Christian speaker substitutes shame-oriented and therefore Spanish-like phrases for plainspoken Quechua ones. For example, in chapter 10 (sec. 151), the word ollonchicta 'our cocks' has been crossed out in favor of pincayninchicta 'our private parts' or, more literally, 'our shame'.

The problem of redaction

An editing process has given the manuscript a veneer of organizational unity. But no one has yet subjected it to text criticism detailed enough to uncover the "seams" where testimonies have been stitched together or to determine how many voices enter in. For the time being, this can only be recognized as an unknown requiring future study. It appears likely that the myth-tellers and retellers are multiple, probably from a minimum of three different places. Whether a separate translator beside the scribe and/or editor intervened is unknown. We neither know whether the interviewer used a written questionnaire (a common Spanish practice) nor whether the edited text as we find it was organized by a native researcher, by the scribe Thomás himself, or by someone employing native informants (Avila?).

It is highly likely that chapters 1-31 are rewritten and edited text, because they are fair copy (standardized at thirty-six lines per page) already augmented with contextual material, especially at beginnings and ends of chapters (Urioste 1973: 7-10). These chapters, however, appear to be an intermediate draft because editorial work continues in the form of interlinear and marginal comments and corrections (some seemingly in Avila's hand). Supplements I and II differ in grammar from the body of the manuscript, are written in a less polished handwriting, and come after a word meaning 'the end'. They may be surviving fragments of less-processed testimony. Or they may be a set of rough notes prepared by the editor/redactor on the basis of personal knowledge, apart from the compiled testimonies.

The problem of validation

One property of the text that may yield clues about the editing process if closely studied is the problem of validation. Quechua, like many American languages, requires the speaker to attach suffixes that clarify his or her relationship to the data conveyed. When conveying data learned from personal experience, the speaker uses the witness validator -mi (alternatively -m, -n), which implies that the content of the sentence or (sometimes) larger speech unit is something learned through direct sense experience. When passing on data learned secondhand—for example, an account heard from somebody else—the speaker will switch to the reportive (sometimes called "hearsay") validator -si (alternatively -s). We suspect that predominant reportive validation may reflect the intervention of a re-teller, perhaps a translator; but when speaking of legendary events, original informants might well have used it, too. When speculating without evidence, or when uncertain, the speaker employs the conjectural validator -cha.

There are two types of passage in the manuscript in which witness validation predominates. One is the contextual remarks and interpolated comments that appear to be supplied by someone other than the myth-telling informants, probably the editor/redactor. Sentences with -mi often introduce or end a chapter, stating with witness validation the relation of a huaca or ritual to a narrative or geographic context. Witness validation is normal in chapter titles. The preface has predominant -mi. Also, at least one contributor often uses -mi witness validation to report apparently contemporary local circumstances, as opposed to mythic or ancient events. For example, a mythic place name will be identified with a modern landmark using -mi. Accusations that local people are secretly carrying on some of the rites that the myths explain carry -mi: for example, ynatacmi musiasca tucoy ynantin llactacunapipas rurancu 'those who are privy to these customs do the same in all the villages' (chap. 9, sec. 133). The other passages in which -mi witness validation predominates are the two supplements, which appear to be rough notes. Perhaps their witness validation is related to their focus on contemporary ceremonies (and not ancient rites or myths of the past), which the writer or informant might have seen. Or perhaps, if they are transcriptions of unprocessed testimony, the validation is the original informant's own.

In the remainder of the text (the great majority of it), which consists of mythic narratives and descriptions of rites and rules of the past, reportive -si validation predominates. It is the usual validation of narrative passages: cay Chaupi hamca sutiocsi huc runa anchi cochapi apo Tmta ñamca sutiocpac churin cancan 'the one called Chaupi Ñamca was the daughter of a man in Anchi Cocha, a lord named Tamta Ñamca' (chap. 10, sec. 142). Because it contains -si, this sentence could be rephrased 'Chaupi Ñamca was, they say, the daughter of...' Such locutions might be expected of any person telling a myth with a slight authorial distance. But -si could also come from a translator or reporter paraphrasing or repeating, without necessarily endorsing, an informant's words. A rendition using that assumption would be 'Chaupi Ñamca was, (s)he says, the daughter of...' These rephrasings are overtranslations insofar as the original does not contain the verb 'say' or imply anything one way or the other about who does the saying.

The exceptions to predominant -si reportive validation in narrative passages occur mostly in dialogue, where the characters speak what is putatively their own experience, and also in utterances that are not dialogue but performative speech (i.e., speech that actually changes the world rather than telling something about it): camca vinaymi causanque tucoy hinantin sallcacunamanta huañuptinca huanacuctapas viconactapas yma ayca cactapas camllam micunque chaymanta camta pillapas huañochi sonque chayca paipas huañuncatacmi ñispas hircan 'You'll live a long life. You alone will eat any dead animal from the wild mountain slopes, both guanacos and vicuñas, of any kind and in any number. And if anybody should kill you, he'll die himself, too' (chap. 2, sec. 19).

There are passages and even whole chapters (e.g., chap. 9) in which it appears that the editor/redactor validates substantive content (as opposed to expository asides) with -mi, as if telling what he had seen with his own eyes. These passages tend to be descriptions of ritual, typically set in explicitly colonial context. A good example is the explanation of how Chaupi Ñamca's ritual dances were scheduled into a purportedly Catholic calendar (chap. 10, secs. 148-151 ). The authorial voice claims witness knowledge of incriminating facts, but coyly slips into reportive where naked dancing is concerned.

There are also passages in which significant mythic material is witness-validated. The crucial passage in chapter 9 (secs. 112-116) where Paria Caca lays down the fundamental rule of his cult is one. This validation may relate to the convictions expressed in the preface, which credits the narrative tradition with evidential worth comparable to that of Spanish writing. It is difficult, however, finally to decide whether such validation indicates belief in the factuality of the Paria Caca myth (on a witness's part or on the editor/redactor's part?) or only signals that some passages, the parts in the editor/redactor's own words, contain a true rendering of a story that may or may not finally be true.

Validation presents frustrating problems for translation, because nothing in English has rhetorical force similar to it. Added phrases like 'It's a fact that' (-mi), 'It's said that' (-si), or 'It could be that' (-cha) do convey validators' informational function, but at such expense to narrative quality as to falsify the text (by making it sound hesitant and colorless) more than they clarify it. We have availed ourselves of the regularities already noted in order to establish norms that minimize encumbrances on the text, while also alerting readers to each significant validation shift. (Some particulars of validation completely escape translation or are difficult to interpret, such as the many instances of a witness-validated discourse marker introducing a reportively validated sentence.) The redactor/editor's stage-managing contributions routinely have witness validation, so the reader may assume its presence in sentences that are obviously editorial, even if they lack explicitly translated validation, unless an overtranslated phrase such as 'it's said that, 'they say', or 'reportedly' suggests otherwise. Such passages occur at the beginnings and ends of chapters. Mid-chapter editorial asides such as 'we told this story in the fifth chapter' (chap. 10, sec. 142) also routinely carry witness validation. In the chapter titles, too, readers should assume -mi witness validation. At mid-chapter validation shifts, notes signal how far the new validation extends. The two supplements are exceptions to these rules. In them the reader can assume witness validation throughout unless reportive validation is explicitly indicated.

Following such curtain-raising editorial formulas as 'the story goes like this' validation regularly shifts to reportive. 'In ancient times' and similar formulas also herald reportive passages. Generally, in the narrative passages of chapters 1-31, except in direct quotations, -si reportive validation may be assumed unless a phrase such as 'it's a fact that', 'in fact', or 'as we know' suggests otherwise. In direct quotations and performative utterances within the narratives, -mi can normally be assumed and is not specially marked. Readers interested in explicitly translated validation should look at Taylor's 1987b Spanish version, which translates validation more fully than we do (though still less than universally).

Translation of style

Validation is not the only respect in which the manuscript varies somewhat in tone and "feel" from passage to passage. Despite the likelihood that coerced elicitation, editing, and translation (or at least "correction" away from a local dialect of Quechua) put us at considerable distance from the myths' oral embodiment, and despite the probability that some of the testimony's oral qualities have been blurred as a result, nonetheless there do seem to be identifiable variations of oral-derived styles within the manuscript. We have tried to suggest them in the translation, giving the more bookish and editorial sentences a slightly different sound from utterances that probably replicate storytelling, ritual speech, and maybe even verse or song.

Framing sentences

An editorial voice intervenes particularly at the beginnings and ends of chapters as a stage manager of testimonies: 'we have already told', 'now we shall tell', and so forth. In the translation we have sought to make this "framing" verbiage easily recognizable by its semiformal and faintly pompous diction. Perhaps the person who composed the framing sentences is the same one who occasionally obtrudes with politically opportune comments:

... no matter how people behave, neither the alcalde nor anybody else would ever try to stop them by asking, "Why do you do these things?"

On the contrary, they dance and drink right along with them until they get drunk. And as for the Catholic priest, they fool him, saying, "Padre, I'm back from cleaning the canal, so I'm going to dance, I'm going to drink."

As far as that goes, all the people do the same thing.

True, some don't do it anymore because they have a good padre.

But others go on living like this in secret up to the present. (chap. 7, sec. 95)

One possible source of such language is Cristóbal Choque Casa, Avila's political ally, who glorifies his own struggle against the huacas in chapters 20 and 21. The same voice or pen may well be the source of the many passages that tell about non-Christian rites surviving clandestinely at the time of writing.

Narrative passages

The bulk of the chapters consists of myths and descriptions of past ritual practice. Often the "framer" alerts us to the shift to mythic discourse and the recalling of past practices with a formula such as chay simire caymi 'that story goes like this'. On the whole what follows is vivid and powerful narration, with -si reportive or "hearsay" validation predominating.

It is hardly accidental that translators differ on how "oral" or how "bookish" the manuscript is. Adelaar called his Dutch translation (1988) Het boek van Huarochirí. Our introduction also emphasizes the likelihood that its redaction was influenced by a European concept of the book and that the redactor thought of myth as material for the making of history books. On the other hand, Urioste gave his 1983 Spanish version a subtitle meaning 'The oral tradition of Huarochirí', and Taylor says that the manuscript "preserves the spontaneous composition of its oral informants" (1987b: 9). Arguedas, sensitive to both qualities, described it as "the voice of antiquity transmitted ... through the mouths of common people," yet also called it "a little regional bible" (Arguedas and Duviols 1966: 9-10).

Taken separately, both the oral and the bookish emphases can easily turn into distortions that obscure the text's peculiar qualities. We see the intersection of orality and literacy as essential. Rather than filtering out bookish traits the better to approximate oral sources, we seek to capture a form of literacy in the act of engulfing one or more oral genres. This approach has the advantage of faithfulness to the manuscript's historic function. But it entails technical problems.

Something would be lost by simply leaving the translated text in solid blocks of page-filling prose like those of the scribal page. It would be a sacrifice of authenticity not to make at least some use of Dennis Tedlock's ([1971] 1983) and Dell Hymes' ([1977] 1981) demonstrations of how internal evidence helps the translator restore qualities of spoken performance to transcriptions that bury performance rules in prose conventions, because some qualities of oral genre do shine through the editorial veils.

It might even be feasible to produce a fully ethnopoetic rendering of at least some Huarochirí texts (that is, one whose goal is to restore oral organization, with versification, by detecting written correlates of pauses, turns, etc. ). We have not followed this method entirely, because, for reasons already mentioned, we doubt our ability to create a hypothetical likeness of the performances from which the manuscript was compiled (or, more likely, recompiled from earlier notes). Also, to make explicit the features that warrant line and stanza breaks (etc.) requires putting into the English text many particles that impede fluent reading of narrative content. Finally, it may be that some of the pauses marked with chaymanta 'and next' and the like do not reflect local norms of oral performance but rather are effects of making the speaker (translator?) wait for the scribe to catch up.

Though inhibited by these caveats, we have sought to indicate, partially and tentatively, some performance qualities of the Huarochirí material detectable within the bookish frame that is also part of its substance. Like most Native American narrative texts, the Huarochirí manuscript is spotted with innumerable "empty" words or discourse markers, especially at the beginnings of sentences, that add no literal content to the flow of action. Common ones include, for example, chaymanta 'and next' or 'after that', chaysi 'so' (with reportive validation), and chaymi 'so' (with witness validation). Translating these words directly gives an effect the very opposite of oral. It makes the text sound hesitant and finicky, disrupted. But they are, of course, not "empty" in spoken performance; they serve to separate utterances in a patterned way, imparting a quality of measure and helping the listener keep pace. We believe they served this function in the oral performances that the manuscript partly mimics and have tried to convey as much. Some words that we read this way are sentence-initial cay 'this'/chay 'that'; ychaca 'however'; yna, ynaspa, and other derivatives of yna 'thus'; ha and hatac 'already', 'then', 'so', 'again'; huc ... huc and huaquin ... huaquin 'some ... others'. Certain other touches, such as emphasized changes of topic or words indicating change of time or setting, also are taken as signaling a new "turn." Many sentences begin with clauses whose informational contribution is minimal (chay yna captinsi 'while it was like that') but which similarly seem to serve a function of measure or pace, introducing a new burst of the storyteller's speech.

Our objective is a compromise, an attempt to make a likeness of the manuscript that suggests its "feel" as a manuscript: that is, a prose presentation following norms of book organization, which for long stretches subsumes oral performances without strictly transcribing them. Stopping short of an attempt to reconstruct the oral performances that the writers heard, but also stopping short of an undivided prose that would silence their clearly audible echo, we have chosen to render the abovementioned discourse markers and other devices of measure into prose divisions that somewhat suggest the quality of turn or strophe, namely, short indented paragraphs. (The section numbers alongside them are only meant as aids to correlating with the transcription.)

Another "oral" characteristic of the Huarochirí manuscript is heavy reliance on direct quotation and dialogue. Quechua ñispa ... ñin 'saying ... (s)he said' delimits quotations, but instead of translating the formula we have simply used 'said' and quotation marks (unlike Taylor 1987b, who often favors indirect quotation). Because the Quechua verb ñiy can mean 'to think' or 'to intend' as well as 'to say', some direct quotations are rendered as quotations of thought: ñocaracpas ñaupac umanman chayaiman ñispa 'thinking, "I mean to get to the summit first!"' (chap. 9, sec. 120).

In order to clarify the dialogue structure of many passages, we have placed each turn of quoted speech in a separate indented paragraph, even when the original quotes more than one speaker within a single sentence. Most Huarochirí dialogue (and narration) sounds lifelike and colloquial, and to convey this quality we freely use English contractions ('isn't', etc. ). In some cases, a quotation contains several sentences that sound like simultaneous comments on the same point, and in these cases we have interpreted the effect as a chorus of comments:

Then they derided Quita Pariasca with spiteful words:

"That smelly mountain man, what could he know?"

"Our father Paria Caca has subjects as far away as the limits of the land called Chinchay Suyo. Could such a power ever fall desolate?"

"What does a guy like that know?" (chap. 18, sec. 223)

Question marks in the translation securely reflect Quechua grammar, but exclamation points are supplied. Parentheses are also supplied.

Versified speech in semantic couplets

A few passages clearly employ the Andean oral device called "semantic coupling" or "thought rhyme" (Mannheim 1985), and we have rendered them as couplets. They appear inset in the translation. The semantic couplet is a pair of sentences that express closely related ideas phrased in related (sometimes syntactically parallel) fashion. Semantic coupling is common in modern Andean folk poetry and song, as in many of the world's oral traditions. Huarochirí examples include:

We go in Tutay Quiri's steps,
We go in the path of his power. (chap. 11, sec. 158)

Tell me, I beg: what have I done to make them ill?
For what fault of mine do I live in suffering? (chap. 13, sec. 179)

Semantic coupling typically occurs in invocations to, or sayings of, sacred beings—huacas themselves, legendary ancestors, priests and oracles delivering responses, or Christian deities. Cristóbal Choque Casa's moving prayer to the Virgin Mary (chap. 20, secs. 254-255) consists largely of couplets. It might be an example of those "devout and elegant" religious songs that Jesuits in Huarochirí heard "idolatry" defendants sing in their cells at night c. 1620 (Arguedas and Duviols 1966: 259-260). Much earlier witnesses, the Jesuits who conducted the pathbreaking mission of 1571, found out that such songs were reworkings of prehispanic hymns: "The same [hymns of praise] which in former times they used to give to the sun and to their king, they have converted into praise of Jesus Christ by taking material from what they heard preached" (Arguedas and Duviols 1966: 243). Some couplets are brief ritual formulas rather than hymns and a few might be humbler bits of folk song or familiar sayings. Margot Beyersdorff (1986: 219) comments that in the Huarochirí manuscript "intercalated fragments of sacred verse ... follow the same norms of versification as the 'prayers' of [Cristóbal de] Molina ["cuzqueño"] and [Felipe Guaman] Puma de Ayala, and can be included in the genre wayllina." She does not identify the fragments. Beyersdorff infers from Molina that wayllina denominated a specific verse form unique to sacred speech. Husson (1985: 330331), on the contrary, emphasizes Guaman Poma's couplets' affinities to popular and profane song.

Other translation conventions

In addition to translation problems arising from idiosyncrasies of the Huarochirí text, the translation reflects some decisions that face anyone working from Quechua. Readers who want to avoid spurious accuracy in using the translation should take note of them.

This book does not aspire to morpheme-by-morpheme translation, for several reasons. Rendering an agglutinative language morpheme-by-morpheme (to the degree that it is possible at all) yields verbose, obscure formulations. The following example of a translation process uses the slash (/) to signal morpheme boundaries and suggests the sort of problems that occur within a single random and typical clause. The original is ña chori/n/cuna/pas collo/pti/n/rac/si ... Analyzed using Urioste's (1973) terminology, this clause consists of three roots (italicized), each with additional morphemes: already child / (3d person) / (plural) / (additive) perish / (contrastive adverbial) / (third person) / (continuative) / (reportive validator) ... This could be rendered, forcing a morpheme-by-morpheme translation, as: already child / his / (plural) / even perish / (verb with different subject to follow) / they / first / it's said ... To make intelligible English one must drop a morpheme (-pti) that has and needs no English analogue and make choices whether to translate -pas, -raq, and -si explicitly. We chose not to translate most -si reportive validators, and the remaining content can be made clear by implication in a simple English form: 'When his children had already perished. . . ' (chap. 19, sec. 233; the referent of 'his' is supplied in the body of the translation).

Our translation generally preserves sentence boundaries, but in some cases the original sentences contain chains of subordinate clauses so long that a single-sentence translation (without Quechua's devices for clarifying switches of subject) comes out congested. In these cases sentences have been divided. Sometimes the order of clauses has been altered. The order of sentences has not been altered. Where sentence boundaries are significantly uncertain, notes indicate the alternate readings. Where the grammatical subject of a sentence is unstated but unambiguous in the original, an explicit subject has been supplied in English. In cases of ambiguity, we tried for English translations with a similar ambiguity. Where this was not achieved, notes indicate the nature of the ambiguity.

Aside from unique cases, there are generally common sources of ambiguity (as the English reader sees it) in regard to gender, number, and time. Quechua does not have grammatical male-female gender and in some cases the sex of a named person is unclear (for example, the master weaver mentioned in chap. 1, sec. 8). Also, because singular forms may be used to mean plural, Quechua nouns are sometimes indefinite as to singular and plural. Finally, the tense in Quechua that most resembles English present sometimes refers to a (usually recent) past, making certain sentences ambiguous as to whether they describe current reality. Such cases are noted where they cause significant doubt about the meaning of a passage.

A given root, especially a verb root, is not always translated by the same English word because the root with its varied suffixes often gives a net sense that varies from the root sense by a distance that only a different English word can render. In the following example the common verb root yalli- ('to exceed, surpass') yields English 'to compete' when coupled with the suffix -naco- (denoting reciprocal action) but English 'to win' when not modified: pomacta aparispa yallinacoson 'Let's compete in putting on puma skins' (chap. 5, sec. 64); pomancunacta aparispa yallita munarcan '[he] wanted to win by wearing the puma skins he had' (chap. 5, sec. 64). However, in passages where repetition appears to be an element of rhetoric or measure, as, for example, with the four-times repeated canancama 'until now' in the preface, the translation mimics the repetition.

The numbers given in the margin are added by the translators. Numbered sections are there only to help readers compare the original and the transcription and to facilitate index references and cross-references. They do not define a unit intrinsic to the method of composition.

The translation does not show the small pen strokes and marks (symbols?) scattered in the original manuscript. The punctuation is added. In addition to sentence-, quotation-, and clause-delimiting punctuation, we have added:

( ) Parentheses to signal apparently parenthetical remarks in the original text.
(This spring flowed from a large mountain that rises above San Lorenzo village.
This mountain is called Suna Caca today.)
At that time, they say, it was just a big lake. (chap. 6, sec. 81)

[ ] Square brackets to signal marginal or crossed-out material on manuscript pages.

< > Angle brackets to identify translators' words referring to marginal or crossed-out material.
<margin, in Quechua:> [Parid Caca

We have translated crossed-out material (words or phrases) that adds substantively to the text, but not crossings-out that only change a spelling or crossed-out single morphemes or letters. All these details are reproduced in Taylor's (1987b) study edition.

Note conventions

Unless otherwise noted, English versions of quotes from non-English sources in notes and in this introduction are by Salomon. Sources cited in notes appear in the general bibliography. Notes provide context sufficient only for novice readers and do not contain full comparisons with other translations or exhaustive references to literature.

Transcription conventions

The transcription strives for the nearest simulation of the original page's qualities compatible with easy comparison to the translation.

The page breaks, signaled with foja or folio notations R for recto and V for verso, are those of the original.

In the transcription the following conventions are used:

Uppercase letters stand for large letters in the original. The original does not use uppercase (rare in scribal papers) but does use large lowercase letters in headings.

[ ] Square brackets contain the section numbers for keying passages to corresponding parts of the translation.

< > Angle brackets identify crossed-out material in the original.

( ) Parentheses identify the expansion of abbreviations. For example, the original's capi^o is rendered capi(tul)o.

Toponymic and onomastic spelling conventions

It is important to note that in the translation (not the transcription) we have standardized the names of persons, places, and huacas so that each has (as nearly as possible) only one name in English context, regardless of the often considerable variations in the original. (Exceptions are cross-referenced in the index.) When the original spelling varies, it is standardized in the translation according to the following rules:

• First: If a proper noun appears only once in the manuscript, or if all occurrences are spelled the same way, we have re tained the original spelling.
• Second: If a proper noun has multiple spellings in the manuscript, we have chosen one of these and maintained it throughout the translation and index. The criteria for choosing the form are:
  1. For words ending alternately in o or u, we selected the u ending.
  2. Wherever i, e, and y vary, i was selected.
  3. Wherever s, ss, x, and ç vary, s was preferred. The manuscript's apparent sound system does not preclude reading x as a palatalized fricative different from s, nor does it conclusively prove this to be correct.
  4. Wherever n and m alternate in final position, n was selected.
  5. Wherever al and ay alternate, ay was selected.
  6. Wherever hua, gua, and ua alternate, hua was selected.
  7. Wherever quia and quio alternate with cya and cyo, the latter spellings were selected.

Some place names as standardized are different from their nearest modern equivalents (identified in notes at the toponym's first occurrence). For example, the modern place is written Huarochirí but the manuscript uses guano cheri and huaro cheri. We use the latter in the translation. Such standardizations from the text, not modern toponyms, appear on the map.

In Spanish names (Thomás, Luzia, etc.) we have conserved the manuscript's spelling and added the acute accents corresponding to their modern forms.

A caution: standardizing is intended to minimize the novice reader's problem in keeping track of the dramatis personae and to make the index simpler. But it sacrifices data important to the analysis of the original text's sound system, to its dialectology, and to source criticism. Students who need such data are urged to use the transcription.

In the transcription and translation, we have followed a common Quechua structure of proper names by treating them as two-word phrases (Paria Caca, Huatya Curi, etc.). Large onomastic corpuses such as the 1588 revisita of Sisicaya clearly show that Huarochirí names are binomial, whether or not a space separates the elements (Archivo General de la Nación, Buenos Aires, ms. 13-17-5-1) . Applying this norm involves some interpretation because word boundaries in the original are sometimes inconsistent or unclear. By normal Quechua syntax rules, the first word would express an attribute applying to the second; some names can be interpreted using this rule (e.g., Sullca Yllapa 'Lastborn Lightning'; chap. 8, sec. 108; chap. 16, sec. 202). Indeed, Huarochirí onomastics remains problematic in general; uncertainty about whether all the names are binomial is only one of many obstacles preventing general translation of proper nouns. A few name meanings are nonetheless clear enough and important enough to warrant bringing them into the translation (e.g., Pacha Camac Pacha Cuyuchic 'World Maker and World Shaker').

Index and glossary

The index gives numbered section (not page) references to both proper nouns and topics and themes, classifying many specific items under rubrics common to anthropology (e.g., for material on sibling relationships, look under "kinship"; for the symbolic value of red, see "color"). The references are to the translation, not to the Quechua original. An instance of a proper name supplied in the translation will be indexed even if only a pronoun or an unambiguous implied reference is present in the original. Personal names are alphabetized according to the first term of the native name. For example, Diego Chauca Huaman is alphabetized under Chauca. The glossary of untranslated words lists all the terms that appear italicized in the translation, giving simple glosses and numbered section references to all their occurrences. Material in notes is not indexed.