Aztec scribes respond to Europe, furthering an ancient literary tradition

Gordon Brotherston

It would be hard to exaggerate the pain caused by Europeans when they invaded Mexico in 16th century, or their destructiveness. We may surmise this thanks not so much to European as to native accounts of what went on. From that period, we have a host of alphabetic narratives written by the Aztecs and other groups which report on the unkinder deeds of Cortes and his conquistador companions (León-Portilla 1959).

These alphabetic narratives continue a long-standing practice, among the Aztec and in Mesoamerica generally, of making paginated books of paper and skin in screenfold format that are written in the local script known as tlacuilolli in Nahuatl (Nowotny 1961; Brotherston 1995). Only a few examples of these old books or codices now survive, almost all having fallen victim to Christian bonfires. In the specifically Aztec tradition we have two examples, and they belong respectively to the two literary genres characteristic of the classic screenfold texts: annals, and ritual (divinatory or dream) books. The Codex Boturini annals tell of the migration from Aztlan to Tenochtitlan (Boone 2000: 209-17); the Codex Borbonicus dream book reflects on the longer Mesoamerican story in three thematic chapters (Reyes et al. 1991).

On the same subject of the European invasion, we also have a host of texts which directly continue the tlacuilolli tradition, adapting to Mexico's changing circumstances. In part these were produced as a result of the strategy used by the Spanish Crown to gain a stronger footing in the newly invaded American territories. The idea was to hold at bay, on the one hand, the feudal conquistador-encomenderos, and, on the other, the Roman church which had begun its efforts at conversion with the arrival of the Twelve Franciscans in 1524. To this end, in 1535 Charles V sent Antonio de Mendoza to be his representative or viceroy in New Spain. Through Mendoza, he set up a state apparatus epitomized by coin currency and law courts. Spanish metal coinage, with its hard definition of exchange value, was introduced systemically only then, fourteen years after Cortés's military takeover. The Real Audiencia, the royal court, turned to the Indians precisely to limit or even undermine the rival powers of church and conquistador-encomendero, encouraging them to submit complaints using as evidence documents written in their own native script.

Responses to invasion in Aztec texts

Against this background, Aztec texts emerged in native script that had wider scope, a famous example being the codex that came to be named after the viceroy himself (Berdan & Anawalt 1992). In this text - the Mendoza Codex -, on sheets of European paper, survivors of Aztec ruling class present the new viceroy with an account of the place he had arrived in, revealing their deep understanding of the economic and social system of times before Cortés. Alluding to the suns or world ages of genesis, they relate the city's foundation history, the setting up of its tribute system and of an economic empire which reached many hundreds of miles eastwards to a frontier which is that of Mexico with Guatemala today, the commodity tribute that continued to pour in from the four quarters long after 1521, and norms of education, behaviour, work and social control. The argument and logic of this text stem directly from the older books, specifically chapters on cosmogony and political expansion, the tribute year, and social organization that can be found in dream books like Féjérváry and Borgia. In all this, the Mendoza Codex bears comparison with the long letter sent to the Spanish crown by Guaman Poma, several decades later, from the viceroyalty of Peru, - the Nueva coronica which claims to have been transcribed from the quipu "books" of the Inca.

Overall, the nature and consequences of the changes enforced on Mexico by the invaders are the subject of many 16th-century texts. They are most finely chronicled in a book of annals or xiuhtlapoualli identified sometimes with the year 1576 - in fact the story goes on past that date - and sometimes with the French scholar Aubin (Vollmer 1981). Today, this Aubin Codex is housed in the British Museum, just as Mendoza is housed in the Bodleian Library, Oxford. Unlike Mendoza, it appears not to have been written in order to be presented to the invaders. Beginning in the 12th century with the migration from Aztlan and noting the foundation of the Aztec dynasty in Tenochtitlan, it is concerned rather with history from the New World perspective, affirming the continuity of that history through and despite the upheavals caused by Spanish colonization. It makes meticulous note of the impact of missionaries, markets scheduled according to the Christian calendar, coinage, judges, and above all the viceregal system begun by Mendoza, which succeeded in gradually diminishing the political power of the line of Aztec rulers that after Moctezuma and Cuauhtemoc ran directly on through the 16th century.

An indication of how Aubin interweaves the stories of the invaders and the invaded can be found in its account of the year 1559, 2 Reed in the Aztec calendar. For this was the year of major moment in both. In November, New Spain was ordered to commemorate Charles V, who had died the year before, in a grand funeral, and this was duly done in an elaborate night-time ceremony which involved the construction of a large wooden edifice, the Túmulo imperial, described in great detail at the time (Cervantes de Salazar, 1972); this 'tomb' is depicted in Aubin. Above it however there is another image, that of the year bundle (xiuhmolpilli) that represents the beginning of a new 52-year cycle in native chronology, of "our years" (toxiuh) as the Nahuatl gloss puts it, in the kindling of New Fire (the start of the previous cycle in 2 Reed 1507, under Moctezuma, is grandly featured in Borbonicus p.34). As accounts of the year 1559 in Aubin and other annals suggest, the chance was seized to heighten the link between the two events - funeral and New Fire ceremony -, since by then the Spanish colonial authorities would not tolerate overt celebration of the latter. Native prompting led to the funeral being held at night, in darkness dramatically broken by the lighting of a candle in the four-sided edifice, whose flame and fire was then spread outwards. As a result, in many key respects, emperor Charles's funeral became a covert kindling of New Fire, and it was done on a night in November well placed in the Aztec calendar for this purpose.

A further, yet more intricate Aztec response along the same lines can found in the Mexicanus Codex, now in Paris (Mengin 1952). This text shares the historical perspective of Aubin in the annals which, beginning in Aztlan, make up the main body of the text, and indeed they invoke the New Fire year 2 Reed 1559 in the same terms. In the more philosophical outer frame into which these annals are set, Mexicanus is formally closer to Borbonicus and Mendoza. Overall it may be understood as a wry reflection on the course that Aztec history had taken before and after 1519, and in particular as a highly intelligent response to the papal initiative of 1582. That is, the Gregorian Reform which attempted to iron out at least the worst wrinkles of the Julian calendar, and in so doing changed for ever Christian articulation and understandings of origins and time.

On the evidence of its own annals, the writing of Mexicanus was completed in the decade immediately after the Gregorian Reform took effect in New Spain in 1583. Its authors note an interest in the College of San Pablo, founded in 1575 in Teipan by the Augustinian friar Alonso de la Veracruz, in a property in the southeast quarter of Tenochtitlan that had been endowed by Moctezuma's grandson Inés de Tapia (Toussaint et al 1990:139). A prominent Nahuatl gloss on the European calendar wheel in the text (p.9) reports that it was then that the Augustinians arrived in San Pablo. The resources and high intellectual standing of this college, which eclipsed anything the Franciscans or other orders had previously accomplished, became proverbial in native annals of the time. Aubin contrives to posit the construction of the College and Friar Alonso's name in terms that are significant in both native and imported chronology. For their part, chroniclers of the Augustinian Order note how the library of books freshly imported from Europe was adorned with maps, celestial and terrestrial globes, astrolabes, planispheres, chronometers and other state-of-the-art instruments (Grijalva 1985:327. I am extremely grateful to Elly Wake for this reference). Used by the Vatican in its attempt to redefine the time of western Christendom, instruments such as these were observed by native eyes in Mexico that had nothing of Europe's innocent other, but rather were informed by many centuries of precise local knowledge, as Mexicanus makes plain.

In a Nahuatl gloss in the annals (p.36), the last of Mexicanus's several authorial hands (a certain "Juan") states that the book was given him by a descendant of the royal Aztec house of Acampachtli (Inés de Tapia?) and that, loyal to the calendrical and philosophical tradition of that house he will do his best to expose the intellectual limitations and oppressive ideology of his country's new rulers.

Among the 400 or so extant tlacuilolli documents, Mexicanus is notable for the emphasis it puts on the notion of political and intellectual continuity among the Aztecs and it stands unrivalled as a legible and thorough examination and critique of the calendar and liturgy imported into Mexico in the 16th century. For this reason, it may well serve as a first guide into complexity of the book and script tradition which it effectively continues in making the critique, decades after the military defeat of the Aztecs in 1521. The matters and phenomena it deals with are generally well documented, well identified and understood, which reduces the need for speculation or guess work at least at the first level of reading. Nurtured by their own millennial tradition, its authors contemplate with keen professionalism the logic, origins, deeper history, and significance of the calendar imposed by Christians.

Mexicanus: hybridity, and the question of literary genre

In several senses, the text of Mexicanus is deliberately hybrid, it occupies a space between (nepantla) Aztec and European, being an example of what León-Portilla has called "nepantilism" (1974, the concept later taken up by Gloria Anzaldúa and Walter Mignolo, among others).

- It is made materially of native paper yet is paginated in European style rather than as a screenfold book, being bound at the spine, which affords the page opening the significance it may have in a European book.

- Though in principle its rectangular pages are subjected by the spine binding to a horizontal layout, it plays with the vertical layout typical of European (and Maya) books.

- It is written in tlacuilolli yet also makes use of European alphabetic script (for words in Nahuatl, Latin, Spanish and other languages) and numerical notation; moreover, it treats these latter with an ingenuity characteristic of the former (Galarza 1966) - according to need, numerals may be arabic or upper or lower roman; letters may be upper or lower case, more roman or more gothic, even uncial (a 4th-century amalgam of Latin and Greek).

- Thoroughly imbued with Mesoamerican calendrics, it works easily with the rules and structure of the Christian liturgy and chronology.

- It construes the time depths and Eras of both its own and the Christian calendars, implicitly asking who entered whose history in 1519.

Mexicanus was written at a time when, as the Gregorian Reform indicates, Europe was endeavouring to bring its chronology and astronomy up to American levels. Unlikely as it may seem, before Charles V's protegée Mercator (Gerhard Kremer) published his Chronologia, hoc est temporum demonstratio...ab initio mundi usque ad annum domini 1568, ex eclipsibus et observationibus astronomicis, Europe had no formal means of projecting dates accurately far into the past (years "BC" were simply not conceived of) and therefore was ill-equipped to match the Mesoamerican Era, let alone the larger cycles that Era is set in. And as we shall see, the corrections eventually made by Gregory had unfortunate side effects: the liturgically (and socially) disruptive loss of ten days, and the elimination of the night sky as a chronological reference.

Within its own tradition, Mexicanus furthers the achievements of millennia, synonymous with the unbroken history of Mesoamerican calendrics and script, which begins with the Olmec inscriptions of 1st millennium BC and projects back to an Era date of - 3113. As a book, the text represents each of the two literary genres known in the surviving corpus of the classic screenfolds of paper and skin. For part belongs both to the ritual or dream book genre (temicamatl), which consists of thematic chapters (pp.1-17, 89-end), and part belongs to the annals genre (xiuhtlapoualli), which narrates events through time (pp.18-88). In other words, each component of Mexicanus, outer and inner, has its own genre and hence mode of exposition, means and principles of reading.

As the inner component of Mexicanus, the annals are a sequential narrative that moves ever forward through time, year by year. Each year is marked as such by a turquoise box (xihuitl is both year and turquoise), and is named, like the New Fire year 2 Reed, by a combination of number (from 1 to 13) and Sign (from a set or Series of four year-bearers) that makes up the 52-year xiuhmolpilli. This reading principle characterizes all known examples of surviving annals, above and beyond many regional differences, with respect to style of markers, year-bearer names and Series, intervals between years and subsidiary day dates, reading line (straight or boustrophedon) and direction. This is a literary principle, that determines the genre 'annals', xiuhtlapoualli or year count in Nahuatl, to which belong more than half of the narratives in surviving screenfold books that adhere to traditional norms.

By contrast, the outer component is subdivided into eight sections or chapters. Each of these has its own theme and own guide to reading, one determined not by the simple forward passage and count of years but by the signs and numbers of the calendar itself, specifically those of its two main constituent cycles, the year (xihuitl), and the human gestation period (tonalpoualli). The year consists of 18 Feasts of 20 days; the tonalpoualli consists of nine moons (expressed as 29 x 9 nights), and 20 trecenas of 13 days. This is no less a literary principle, that determines the genre 'ritual' books (in Nahuatl teoamoxtli, or temicamatl, 'dream'; the core texts in the genre are often loosely referred to as the 'Borgia group'). Without a single exception, every last chapter in the traditional screenfold corpus that cannot be assigned to the annals genre belongs to this other genre. In practice, compared with the annals, the texts in this genre tend to the larger view and are more conceptual, theoretical, meta- i.e reflexive on their own practice, philosophical, poetic, and may appeal to simultaneously levels of time, from the immediate present to the depths of creation and the world ages (Brotherston 1997). Hence they are often more demanding at the first level of comprehension, and may indeed elicit multiple readings, especially on pages that have the format of tables or complex time-maps. This is the case with the such pages and chapters in Mexicanus which explicitly demand successive readings or 'counts'.

Hence, the difference between the two component parts of Mexicanus, inner and outer, resonates deeply within whole corpus of screenfold books. It brings home the need to recognize characteristics of native texts in the terms they propose, which by no means always match those we have chosen to impose, like those western binary distinctions between dia- and synchronic, linear and cyclical, historical and calendrical, chronological and divinatory, and so on. Only by respecting the authority and indications of the texts themselves can we adequately gauge what genres they belong to, what functions they had, and how they may echo and correspond to each other, as we become sensitive to ritual tendencies in annals, and narrative tendencies in ritual books.

To examine the whole text of Mexicanus in the detail it deserves is clearly impossible here. It is remarkable for the subtlety with which it conveys and organizes its arguments and for the range and depth of knowledge these presuppose. It also has been heavily consulted and re-used over time, being much thumbed, and effaced in places; it may also have lost some initial and end pages. What follows is a preliminary account that builds on existing work by Mengin, Galarza and Prem and attempts to present the broader arguments made in the ritual chapters and the annals and their interconnection, ever with respect to literary precedent in the native tradition.

The ritual chapters

The 'ritual' outer frame of Mexicanus is made up of eight chapters, which engage with the Christian calendar in great detail, analyzing the philosophy implicit in it and revealing its technical limitations with respect to local and Mesoamerican norms. The analysis dwells on the relationship between the two calendars as this is exemplified historically in the Annals, and was foregrounded by the Gregorian Reform. The chapters are defined according to categories proper to the genre of dream books, and therefore necessarily have as their constant premise one or other of the two main cycles of the Mesoamerican calendar: the year of 18 Feasts of 20 days (F1.1- F1.9; F2.1-F2.9), and the gestation period of nine moons of 29 nights and 13x20 days (tonalpoualli).

In this respect, the eight chapters fall into two groups, based successively on the year and the tonalpoualli. The first group (chapters 1 to 3; pp.1-11) is grounded on the 18 Feasts of the year, as well as the 52-year xiuhmolpilli cycle, and the annual night sky; the second group (chapters 4-8; pp.12-17, 89-end) is grounded on the Twenty Signs, Nine Nights and trecenas of the tonalpoualli. In both chapter groups, the night is carefully distinguished from the day, and prominence is given to the night-sky cipher eleven. Both align native models against imported ones, using page-formats and cycles that are partly Christian and partly native in origin.

In opening up the concept of the year, the first chapter focuses on the cycles of Aztec and Christian liturgy. Running from May to December (in the text as it now stands), a narrow marginal column to the left on each page furnishes such details as zodiac sign, month name in alphabetic script and tlacuilolli, and length of month in days. On each page, the lower register is dedicated to glyphs of the eleven Aztec Feasts that run from Toxcatl (F1.3) to Panquetzaliztli (F2.4), plus the fractions of Feasts before and after, mementos of the great metamorphoses and catastrophes deep in the time of creation. These Feasts in the year are correlated with successive sets of tonalpoualli days. At the half-way point of one Feast, Hueypachtli (F2.2), the glyph is given of another of its names, Tepeilhuitl (p.7; Fig.1) (Ill. 1) . In images and phonetic glyphs, the upper register offers a selection of Christian Saints' and Holy Days: which are chosen and how they are depicted make a complex argument in its own right. All these days are fixed in the year, with the single exception of the Ember days which look back from Christmas to the end of Eastertide (and are therefore moveable).

Page by page, the lower and upper registers are separated by a double band of letters, consisting of the 7-letter count a-g that begins on New Year's day in any given year, and the 27-letter count of the sidereal moon, that is, its passage through the zodiac (the "lunar letters"). As it were rooted in popular custom below, four chosen Feast days grow upwards, stem-like, through this double band of letters, indicating that they may continue to flourish within the new colonial order. The lordly Feast Tecuilhuitl (F1.5) rises to become the solstitial St John's day in June, and a similar though less confident link is made between the hunting Feast Quecholli (F2.3) and St Martin's day in November, and the "falling water" Feast Atemoztli (F2.5) and St Thomas's day in December. The case of Toxcatl (F1.3) in mid May is the most striking since the emblem of that Feast, Tezcatlipoca's triple-ring sceptre, thrusts right up into the upper Christian register, higher on the page than the Holy Cross that just before and beside it denotes 2nd May. Glossed in Latin as the living tree (arbor), this pagan sceptre would recall for every native reader Tezcatlipoca's role in creation no less than in the Aztec expulsion of Cortes and his army from Tenochtitlan in that Feast in May 1520. The principle of such stem Feasts remains alive in the timing of ritual today (Broda & Baez 2001).

Suspecting this possibility, the Christians put great energy to suppressing the pagan Feast cycle. A poignant case concerns their determination to extirpate memory of the ancestors by enforcing observance of their imported feast of All Souls and 'Month's Mind' in November. In the Feast cycle, the proper time for such observance is Micailhuitl (F1.7) in August, days in the year invoked in the very beginnings of the Mesoamerican calendrical Era in the year -3113. Its perennial significance determined Cuauhtemoc's delay in surrendering to Cortés in August 1521, until the point in Micailhuitl that coincided with St Hippolytus's day (Thomas). At this time of year, the dead journey in their after life, passionately accompanied in thought by those who survive and remember them. Mexicanus shows the souls' journey as footprints (the only ones seen in the sequence of Feasts). Driven from their original home in Micailhuitl, four footprints go precisely as far as the day of St Hippolytus, midway through the Feast. Eighty days later, on All Souls Day in November, just two footprints, half their number, show them seeking shelter as refugees in the Tlaloc mountain glyph of Tepeilhuitl, midway through Hueypachtli (F2.2), Mexico's Day of the Dead today. This half Feast of 10 days in November also ingeniously intimates the 10 days eliminated from that month in the Gregorian Reform.

The days of two Saints, Francis and Martin, are noted as those of the arrival in Mexico-Tenochtitlan of Cortes and viceroy Suarez de Mendoza, in 1519 and 1580 respectively. The fashionable black hats and bearded faces of these Spaniards can also be seen in the annals at these dates, and as conquistador and viceroy they complement the church as the three arms of power extended by Europe into 16th-century Mexico. Cortes's intrusion on St Martin's day 11th November links him with the Feast Quecholli (F2.3), when migrant birds arrived from the north and the hunting season started on the lakes around the capital, a coincidence relished in many an Aztec narrative. On that day Cortes is reported to have kidnapped Moctezuma, having failed to convince him that the Bible was any kind of match for American accounts of genesis (Thomas 1993:334). Noting Suarez and Cortes in this way here means formally integrating historically specific events into the year cycles of both Christian and Aztec ritual.

In its clarity and comprehensiveness, this opening chapter grounds itself in native calendrics in order to affirm the possibility in principle of combining meaningful liturgy with precise time-measurement (this after all had been the aim of the Gregorian Reform), and going beyond that, in coordinating ritual with history.

The following chapter (2; Fig.2) (Ill. 2) deals with the main question addressed by Gregory's Reform, i.e. the need the measure the solar year more accurately than the Julian calendar had done when simply adding a leap day every fourth year. Julian practice is represented by the wheel of 4x7 Dominical Letters, the cycle of 28 years derived from the seven-letter count and the four-year leap-day peroid. Thanks to a reference to the final year of Mendoza's reign as first viceroy (1540), the cycle is located here in the Christian Era as the years Anno Domini 1551 - 1578. Matching it to the right is the comparable Aztec cycle, the 52-year xiuhmolpilli, of 4x13 rather than 4x7 years. The two wheels touch and mesh at the xiuhmolpilli ending in 1 Rabbit 1558 and the subsequent New Fire year 2 Reed 1559. Technically, the wheels are so engineered as to reproduce the dar superior leap-day formula of 29x52 (their lowest common denominator), flexibly used in the many regional variants of the Mesoamerican calendar and found at early dates in its far longer Era (Edmonson 1988). Visually, the Christian wheel defines its year units by means of rigid lines radiating from a centrally authoritative St Peter, papal master of locks and clocks and the printing press. The turquoise years of the Mesoamerican wheel are more like the segments or vertebrae of a coiled snake, that carry the attentive reader back to the start of the Mesoamerican Era. Announcing that it requires no less than four levels of reading (nauipoualli), this very dense page chapter also delves into year-name correlations within the Mesoamerican Era.

The opening set of the three year chapters culminates in the night sky, in the form of the Old World zodiac, the twelve stations the sun passes through annually, into which are inset the 28 stations of the moon. The zodiac appears twice, starting first in Aries (traditionally the spring equinox in March; p.10) and then in Aquarius (traditionally January; p.11), reflecting Christian indecision about when the year should begin.

The Aries zodiac concentrates on the further concomitant problem tackled by Christianity in the centuries-long Paschal Controversies, and again in the Gregorian Reform: how best to determine the date of Easter, as the first Sunday after the first full moon after the spring equinox (Parisot 1996). To that end, page 10 relates the zodiac in tabular form to the sequence of 19 Golden Numbers, which reconciles the synodic cycles of sun and moon over 19 years, as well as the letter sequences of the sidereal moon (extended now from 27 to 28). For its own purposes, the Reform solved this problem no less satisfactorily than it did that of the leap day. Yet in doing so it abandoned the night sky, that is the actual positions in it that the zodiac constellations had come to assume over the year, as the result of a slippage greater than that caused by the excess of Julian leap days. Known as the precession of the equinoxes, this slippage is caused by the difference between the solar year of the seasons and the slightly longer sidereal year of the zodiac. Precession amounts to one night about every 70 years and therefore by that date in the Christian Era accounted for more than an Aztec 20-day Feast. As far its liturgy went, in 1582 Rome as it were gave up if not on the phases of the moon then on the lunar letters of the zodiac and the night sky in general, a fact of some historical and philosophical consequence. To begin with, it obscured the rationale by which the very advent of Christianity was thought to have coincided with the equinoctial sun's precession from Aries towards the constellation Pisces.

That the night sky could not be so readily neglected in Mexico may be deduced from subsequent corrections of their own (p.15) which the authors of Mexicanus make to the Golden Numbers of the years 1579-82. These gauge the effects of precession with amazing accuracy, as Hanns Prem well observed (1978:278). In this respect, we may also usefully compare this zodiac that begins in Aries (p.10) with its companion that begins in Aquarius (p.11; Fig.3) (Ill. 3), and both with that in the initial chapter (pp.1-8).

The zodiac that begins in Aquarius recalls the fact that the 1582 Reform established the start of the liturgical year in the month of January, when the sun was supposed to be "in" this constellation. But by then the sun was in fact in Capricorn in January and in Aquarius in February. In other words Aquarius in January here announces the reified zodiac of Gregorian Christianity (and astrology) rather than one that shows the constellation the sun is actually "in" during any given month. The obsolescence of this arrangement is suggested in various ways in Mexicanus. As life forms, the zodiac signs beginning with Aquarius are subjected to rigorous critique in the name of native American genesis, and the four Aristotelean "elements" they are tied to are wittily exposed as nothing of the sort, and to be rather the product of effort and unending change. Yet more specifically, glosses on the two "fire" constellations that succeed Aries in the year, that is Leo and Sagittarius, reinforce and amplify the concept of precessional shift established already in chapter 1 with respect to this last. There (p.7), Sagittarius is shown firing his arrow backwards over his head as if to indicate the monthly positions in the year that he has occupied in the past, and is moving forward from, as the leading edge of his bow traverses the edge of the left-hand marginal column.

A further, highly eloquent diagnosis of the Old World zodiac is to be found in the annals in the eleven-page Nahuatl commentary on the twelve signs (pp.24-34), which goes on (p.36) to demand the "annihilation" of the sign Libra or "peso", that epitomizes unwelcome foreign weight and coinage (niman nimitz micquitiz peso. Libra is in fact the only one of the twelve that is not a life form - zoodion- , having been later inserted in the interests of Old World duodecimal standards). This elimination was to be done, as we saw, out of loyalty to Acamapichtli and in Mesoamerican terms the effect, clearly, is again to defend local understandings of the night sky in which eleven is privileged as a cipher.

In contemplating the year, the authors of Mexicanus were able to find a certain common ground with Christendom: they were less lucky when it came to the other main cycle of their calendar. This is the tonalpoualli of 260 nights and days, by which midwives measured human gestation, from conception (the first missed menses) to birth. The five chapters in Mexicanus relative to this cycle (rather than the year) seize on what parallels they can from Christian doctrines of Christ's birth and body, and the nature of his mother Mary.

This second tonalpoualli group of chapters starts by focusing on a white male body that because of its wounds could be that of Christ (ch.4; Fig.4) (Ill. 4). On these wounds we can just make out palimpsest words in Nahuatl for two of the Twenty Signs, the movement (ollin, XVII) in the heart, and the lizard (cuetzpalin, IV) in the stomach, among others reduced to marginal ghosts. This is a drastically reduced example of the convention in the classic screenfolds whereby the splendour and power of such major gods as Tlaloc and Tezcatlipoca are constructed through the placing of the Twenty Signs on and around their bodies (e.g. Laud p.45; Borgia p.17).

From pale naked male we move on (ch. 5) to a tonalpoualli table that, neatly aligned in eleven columns, carries us through half a gestation period, from late in Hueytecuilhuitl (F1.6; 30th July) to day 8 of Panquetzalitztli (F2.4; 8th December). This end date gives the clue to what is at stake. For as the opening chapter shows amid marine shells, 8th December is when Mary Mother of God was herself "immaculately" conceived, a doctrine resisted by Aquinas since it raised her status vis-à-vis her son and offered to modify a little the resolute maleness of the Trinity, so alien to Mesoamerican belief. In native eyes, Mother Mary's familiarity with the midwives' tonalpoualli would be obvious; the ranging of its days in elevens (and cumulative counts of eleven, i.e. 66+66) by implication confirms her kinship with both the angels and the Star-Skirt (Citlalicue) of Aztec genesis.

From mother we return to son, and the passion that wounded his body. Another table (p.15; ch.6) fixes the beginnings of the Easter story in Septuagesima Sunday and Ash Wednesday, over four years which continue the Dominical Letter sequence on p.9, 1579-1582, and lead directly up to the year of the Reform. Once a major moment in Christian ritual and now widely forgotten, Septuagesima demanded that Christ's Passion be matched with the story of creation itself in readings from the biblical Genesis. At first, Mexicanus follows this thinking in its own terms, infusing native wit into the Christian story. Like the stem Feasts in chapter 1, the decimal numbers of Anno Domini years grow into vigesimal trees of the land that sprout leaf-hair (tzontli, 400) and fruit flags (pantli, 20). Sunday (domingo) is shown as the face of the lord while that of Wednesday (miercoles) is blackened with ash to indicate the planet of its night (Wotan or Mercury). With its stigmata, Christ's four-digit hand (ma-itl) spells 4th March (marzo), while comparable dates are to be read through the roman numeral versions of the Twenty Signs (dog snake as xv, or 15th February), or reckon with their nature (like the lunar rabbit as its 31-day gestation period, in January). Then, the sap ceases to flow, the trees wither and play with the Signs of fellow species yields to games with month names in alphabetic script (fenelo as either enero or febrero or both) and with numbers (158ii). This points no doubt to the confusion caused by the Gregorian use of recently imported arabic arithmetic in the roman interest, its break with centuries of liturgy and custom, and the year's delay in its official impact on Mexico. It also hints at an endemic decline in Christian imagination that it might be beyond the power of any Reform to mend.

As for the night count, in rhythms traceable back to Mithra or even earlier, this begins at Septuagesima and runs over 9 weeks of 7 nights to Easter, over two weeks of 9 nights to Ash Wednesday, and over an exact Mercury cycle (116 nights) to Ember day, the final Wednesday of Eastertide prominently noted in ch.1 (p.2). Such arithmetic is readily found in classic tlacuilolli texts: the opening chapters of Borgia and Cospi, for example, express the nine moons of gestation as 2(92 +72) = 260, a formula that still informs ritual today (Tedlock 1982; see also Fonds Mexicains 20, Laud pp.31-38). The Mexicanus attempt to find ancient resonance was again however doomed to fail, for Rome's waning interest in the zodiac extended to the night sky in general and eventually led to the abolition of Septuagesima.

Subsequently (chapter 7), in preparation for the annals that follow straight after, the panorama narrows to specifically Aztec beginnings at Tenochtitlan, the birth of Acamapichtli's dynasty and its male and female lines. Here, the genealogies of emperors and their wives are laid out between east and west according to the twenty of the Signs in grid-form (p.16) and the seven and the nine of the moon arrayed as crescents (p.17). After the close of the Annals (p.88), the final tonalpoualli chapter has the page format of the 20 trecenas (chapter 8), which is exactly that found in the classic dream books, and it leads forward into a future where the old calendar still lives. Neatly inverting the procedure in chapter 1, the tonalpoualliv days in the trecenas serve as the yardstick for the 20-day Feasts of the year. In each case, the count or reading is multiple, fourfold.

In all, these eight frame chapters in Mexicanus well exemplify the reach of the ritual genre, at once technical and philosophical, and use it to engage with the imported calendar and dogmas of Christendom, with especial reference to the 1582 Reform. In the process they equip us for a fuller and closer reading of the Annals.

The annals

For its part, the annals component of Mexicanus belongs fully to the xiuhtlapoualli genre, in logic and mode of exposition. The year count begins with the Aztecs' departure from Aztlan in 1 Flint 1168 and ends, after Gregory's Reform, in 7 Rabbit 1590, covering 423 years in all. The narrative records a particular version of Aztec history, which as chronology and ideology stands somewhere between the earlier plebeianism of Boturini and the direct imperialism of Mendoza. With respect to the migration route, the role of Huitzilopochtli and the succession of emperors, there is a particular affinity with the Azcatitlan Annals, as Barlow showed long ago (1949).

Along the route they travel in Mexicanus and Azcatitlan, the Aztecs soon come to two "snake" places which otherwise tend to be mutually exclusive in their histories: Coatlicamac ("snake maw") recalled only in early accounts; and Coatepec ("snake mountain"), celebrated only in later, more imperialist accounts. In the texts where it appears, Coatepec is identified as the site of Huitzilopochtli's celestial birth as the sun, which prefigures his dominance atop the great pyramid of Tenochtitlan, later built and named in honour of this mountain event along the migratory route (Garibay 1958: 29-39, 77-83). In Mexicanus, Azcatitlan (and subsequent versions), this appearance of Huitzilopochtli at Coatepec directly complements his earlier appearance at the tribal birthplace Chicomoztoc ("seven caves") that lay to the northwest like the island Aztlan though was less specific to the Aztecs (like Coatepec, this toponym is absent in Boturini). Azcatitlan especially reveals a certain numerical logic inherent in this conjunction of terrestrial and celestial appearances, figuring seven and eleven caves or wombs in each respectively.

Huitzilopochtli's birth from the sky at Coatepec, allegedly in the year 2 Reed 1195, was made to coincide with the Aztec kindling of New Fire, the first of the seven such ceremonies shown in Mexicanus that were held every 52-years in the year with that name. As the new-born sun, Huitzilopochtli eliminates the stars of darkness at dawn, suspecting their treachery - an obvious enough trope. Yet these are the very stars which serve to determine the exact moment - the Pleiades directly overhead at midnight- when New Fire is kindled and a new 52-year cycle begins. The antagonism also confirms, then, how the New Fire ritual necessarily brings into conflict the year of the sun and slightly longer year of the stars, revealing the precessional slippage between the two, of about one night in seventy years. This meant that after four kindlings or 208 years, the New Fire ceremony would have to be scheduled three nights later in the solar calendar.

The seven kindlings featured in the Mexicanus annals run from 2 Reed 1195 at Coatepec to 2 Reed 1507 at Tenochtitlan, in the reign of Moctezuma II (see Borbonicus p.34). Both Mexicanus and Azcatitlan make a clear difference between the first four and the last three of these kindlings, the former being those of a people still migrant or "on the road" (otlica), the latter occurring after 2 Reed 1403 in the new home Tenochtitlan, under the aegis of the dynastic line of emperors begun under Acamapichtli (as shown in Mexicanus itself, p.17). In Azcatitlan, the before-and-after difference prompts a complete switch in format. With reference to the initial group of "migrant" kindlings, performed over the 208 years that run from 1195 to 1402, both texts correctly detail as "three lost nights" the lapse between solar and sidereal time that builds up during this period (which indeed offers the first ready correlation between the 52-year and the 70-year cycles. Cf. López Austin 1985). Thereafter, such adjustments may be presumed to be taken care of in the "Aztec arrangement" of time and space (pace the less optimistic Hassig 2001: in a calendar functioning continuously over millennia as it did, Mesoamerican fire-kindlings could not have failed to expose and respond to the effects of precession). In the Mendoza account of Tenochtitlan, for example, the periodicities of the sidereal and the solar year, as well as the sidereal and synodic moon, are embedded in the totals of garrisons and tribute towns allotted to the metropolitan area and its four surrounding quarters.

Absorbing enough as they may be in themselves, these details acquire yet more edge in Mexicanus when due attention is paid to the manner in which they are formally set out and articulated on the page.

In the annals genre, the rhythms and groupings of years on the page are generally significant and imply arguments in their own right. This is notably the case with annals recorded in the screenfold books like Boturini, which averages the years of its Aztlan migration story at ten per page, to achieve its total of 210, or three lifetimes of 70 years (denoted as such on p.2, and as "three score years and ten" in Mendoza, f.71). The far longer migration narrated in the 52-page Vienna screenfold observes a "trecena" maximum of 13 per page, and this in the Xicotepec screenfold becomes that of the page opening. Of the pair of screenfolds from Tlapa, Tlapa II has 8 years per page (and therefore always has same initial year-bearer Sign per page), while Tlapa I has 7, the complex year guardian sequence of the midwives spelt out in Borbonicus (pp.21-22). The semantic significance of the Tlapa page totals corresponds to the theme chapters found respectively on the verso side of these screenfolds: commodity tribute payments, per quarter and per four years in Tlapa II; genealogy in Tlapa I.

For over more than four centuries or eight 52-year cycles, along the road that runs from Aztlan in 1168 to 1590 (423 years inclusive), the annals in Mexicanus observe a duodecimal rule of 6+6 per page opening. Patently a hallowed norm in the imposed Christian system, this frequency in Mesoamerica terms invokes rather the bone-hard members of the Twenty Signs: VI and XII, skull and jaw. Moreover, it is ingeniously recast so as wholly to recover a harmony with the norms of native arithmetic, calendar and astronomy. This is done thanks to the breaking of the duodecimal rule on three highly significant occasions, at the beginning, the end, and once in between.

At the start, there is room for only 3 rather than 6 years on the first page (p.18; Fig.5) (Ill.5) since the travellers have first to leave the island Aztlan, and step up over water to the road level of the year- count, striding into their own rhythm and rule. At the end, the final page-opening squeezes in 10 years (1572-1581; p.86), and then 9, removing tlacuilolli colour and introducing arabic numerals (1582-90; p.87) as if to highlight the Gregorian Reform (which those numerals made possible), and its corresponding detachment of the 19 Golden Numbers from the night sky. Then in a single-page coda (p. 88), the path runs on after 1590 into a future of unindividuated years: this crosses a path of stars trodden by Christian converts which however veers dangerously back towards the pagan forest. Accordingly, of the twelve stars shown, one falls aside Judas-like, leaving a most meaningful eleven.

The only other occasion when the duodecimal rule is broken in the Mexicanus annals occurs on a page opening of 11 rather than 12 years (pp.80-81), which responds to the challenge openly posed to the native management of time, calendar and economics after the installing by the first viceroy Mendoza (1535-40), in particular for Moctezuma's direct successor Teuetzquititzin who ascended to the Aztec throne (the last throne shown in the annals) in 1540-41 (cf. Aubin f.46v). The steady imposition of European standards and law came to imply that of the Christian Era itself, first recognized and measured as such in Mexicanus in chapter 2, with respect to precisely these dates. Instead of 6+6 years, the page opening here in the annals records 5 + 6 years 1537-1547, again the locally resonant eleven, which in turn leads to impressive overall reformulations.

The three exceptions to the duodecimal rule may so far be summarized in years as less 3, less 1, and plus 7 (or plus 1 if page 88 is included). In other words, at six years per page 423 equals 35 page openings plus 3 years, 70 pages plus 3 years, or 71 pages less 3 years. In this formulation, the Aztec annals in Mexicanus recover the 70-year span of Boturini, doubling the total from 210 to 420; and in doing so they make explicit its significance not just as an earthly lifetime but as an astronomical frequency, since the precessional slippage of one night occurs somewhere between every 70 and 71 years (western estimates were still being revised in the 20th century).

Indeed, the recurrence to 11 years at the time of Mendoza and Teuetzquititzin can be fairly construed as the chronological crux of the whole annals sequence in Mexicanus. For the page fold between the 5 years that end in 10 House 1541 (p.80) and the 6 years that begin in 11 Rabbit 1542 (p.81) divides the 423-year narrative into a first main period of 374 years (1168 - 1541), and a subsequent period of 49 years (1542 - 1590). Hence, invoking longer time rhythms, it reveals a sublimely graceful formula that subjects the composite duodecimal of Christianity to the series of tough primes that exist within it, privileging precisely the higher primes that have most native resonance in sky and earth, 11 before and 7 after:

11 (32 + 52) + 72 = 423

That a reading of this order is appropriate is suggested by the fact that it fully accounts for irregularities in the text which can hardly have been random. It is also borne out by Juan's Nahuatl commentary on the zodiac and the need to annihilate Libra, referred to above, which refers to the middle one of the "elevens" sequences in question (1201-1266), and to the cumulative count of 11 (i.e. 1+2+3+4...+11= 66):

1168-1200 = 3x11 = 33

1201-1266 = 6x11 = 66

1267-1541 = 25x11 = 275

These annals in Mexicanus tell a story like all examples of the genre they belong to. True to classic models of the genre, they likewise go much further than simple report, establishing perspectives, meaningful rhythms in time, and a continuous interplay between earth and sky. Through their own serial mode of exposition, they respond to the ritual chapters that frame them, so that the message of the whole text is far more than the sum of these two components.

The textual message

Prompted by the Gregorian Reform enforced in Mexico in 1583, Mexicanus responds with immense finesse to the European invasion, its practical and intellectual demands. It does this precisely by remaining loyal to a philosophy of time and space which avoids the easy binaries of the west, and deploys a much better mathematics and astronomy. Above all, it relies on and furthers the genre expectations of the tlacuilolli literary tradition, tracing subtle interplays between theme and narrative, cyclic pattern and historical moment.

Mexicanus was written at a time when Rome felt threatened intellectually by America on these and several other counts. Try as it undoubtedly did, it could no longer hold on to the central authority that it had proclaimed for centuries in such matters. Caught by Ptolemy, its own dogmas and philosophical priorities, it lost the night-sky to the far more accurate yet heartless, socially abstract and unreflexive mechanisms of the "science" that Europe was beginning to embrace precisely when the Augustinians arrived in San Pablo with their astrolabes. And it soon found itself burning the treatises of those scientists with the same enthusiasm that it burned tlacuilolli screenfolds in Mexico and quipus in Peru.

For these reasons, within its own tradition Mexicanus may serve as a fine and much-needed guide to tlacuilolli literature, to wondrous articulations of time and re-livings of genesis, no more than hinted at here, that Europe suppressed, ignored or simply failed to understand.

© Gordon Brotherston, 2003.

Bibliography

Barlow, Robert. 1949 'El Códice Azcatitlan', Journal de la Société des Américanistes 38: 101-35

Berdan, Frances & Patricia Anawalt. 1992. Codex Mendoza. Berkeley: University of California Press

Bierhorst, John. 1992. History and Mythology of the Aztecs. Tucson: University of Arizona Press

Boone, Elizabeth. 2000. Stories in Red and Black. Pictorial Histories of the Aztecs and Mixtecs. Austin: University of Texas Press.

Broda, Johanna & Félix Baez Jorge. 2001. Cosmovisión, ritual e identidad de los pueblos indígnas de México. Mexico: FCE

Brotherston, Gordon. 1995. Painted Books from Mexico. London: British Museum Press.

Brotherston, Gordon. 1997 La América indígena en su literatura. Los libros del Cuarto Mundo. Mexico: Fondo de Cultura Económica.

Cervantes de Salazar, Francisco. 1972. México en 1554 y Túmulo imperial. Mexico: Porrúa

Edmonson, Munro. 1988. The Book of the Year. Salt Lake City: University of Utah Press

Galarza, Joaquín. 1966. "Glyphes et attributs chrétiens dans les manuscrits pictographiques mexicains du xvie siècle: le

Codex Mexicanus 23-24", Journal de la Société des Américanistes, 55:7-42

Garibay, Angel María. 1958. Veinte himnos sacros de los nahuas. Mexico: UNAM

Grijalva, Juan de. 1985. Crónica de la Orden de N.P.S.Agustín en las provincias de la Nueva España [1624], Mexico: Porrúa

Hassig, Ross 2001. Time, History and Belief in Aztec and Colonial Mexico. Austin: University of Texas Press

León Portilla, Miguel. 1959 Visión de los vencidos. Mexico: UNAM

León Portilla, Miguel. 1974. "Testimonios nahuas sobre la conquista espiritual", Estudios de cultura nahuatl 11: 11-36

López-Austin, Alfredo. 1985 'El texto sahaguntino sobre los mexicas'. Anales de Antropología (UNAM) 22: 287-336.

Mengin, Ernst. 1952. 'Commentaire du Codex Mexicanus', Journal de la Société des Americanistes 41:387-498

Mignolo, Walter. 2000. Local Histories/ Global designs: subaltern knowledges and border thinking. Princeton: Princeton University Press

Nowotny, Karl Anton. 1961. Tlacuilolli. Die mexikanischen Bilderhandschriften. Berlin: Mann

Parisot, Jaen-Paul & Françoise Suagher. 1996. Calendriers et chronologie. Paris: Masson

Prem, Hanns. 1978. 'Comentario a las partes calendáricas del Codex Mexicanus', Estudios de cultura nahuatl 13:267-88

Reyes García, Luis, Ferdinand Anders & Maarten Jansen. 1991. El libro del ciuacoatl (Códice borbónico). Mexico: Fondo de Cultura Económica

Tedlock, Barbara. 1982. Time and the Highland Maya. Albuquerque: Univrsity of New Mexico Press

Thomas, Hugh. 1993. The Conquest of Mexico. London: Hutchinson

Toussaint, Manuel, et al. 1990. Planos de la Ciudad de Mexico Siglos XVI y XVII. Mexico: UNAM

Vollmer, Günter (ed.). 1981. Geschichte der Azteken. Der Codex Aubin und verwandte Dokumente. Berlin: Mann