The Romances de los señores de la Nueva España, or Romances, as the codex has been called for short, is a hastily penned Nahuatl manuscript of forty-two folios, undated, unsigned, and with a few stray comments in Spanish, jotted even more hastily, by an anonymous glossator. Evidently the work is a transcript of an original now lost, with numerous scribal lapses that betray imperfect copying.
Preserved with the Relación of Juan Bautista de Pomar, composed March 9, 1582, as its author states, the Romances may have been intended as an adjunct to the Relación and therefore compiled about the same time. The extant copy of both Relación and Romances, however, has been assigned to the 1600s.1 As for the presumed Romances original of ca. 1582, behind it one glimpses no fewer than four urtexts, perhaps workbooks, from which the compiler drew the four partes of his compilation. These hypothetical workbooks, containing the texts of cuicatl 'songs', subsequently dubbed romances 'ballads', were evidently taken from the lips of native singers.
To whom, then, might we owe the workbooks or worksheets? Angel M. Garibay, the first modern editor of the Romances, suggested that during the second half of the 1500s, when the initial dictation seems to have taken place, there could have been no one but the tireless and resourceful Bernardino de Sahagún, orchestrator of the encyclopedic Florentine Codex and (very possibly) the unsigned Cantares Mexicanos, who might have recorded such material; or at least the method was his, if we assume that Pomar prepared the urtexts; and perhaps Pomar had been a student or follower of Sahagún's.2 Of Pomar little is actually known aside from his authorship of the Relación, which summarily treats the history, customs, and physical environment of Texcoco, the old seat of Acolhuan kings (thirty kilometers northeast of the center of Mexico City).3 As noted evocatively by the early-seventeenth-century historian Fernando de Alva Ixtlilxóchitl, Pomar was one of several infantes de la ciudad de Tetzcuco, hijos y nietos del rey Nezahualpiltzintli de Tetzcuco 'princes of the city of Texcoco, sons and grandsons of King Nezahualpiltzintli of Texcoco'.4
As for the thirty-six songs of the Romances itself, these are not only in Nahuatl but in oral Nahuatl, as distinguished from the written Nahuatl of missionaries and acculturated native speakers. The difference is between an older, economical (or paratactical) diction and a newer manner influenced by Spanish syntax. A barrier between the two styles is detectable, and in the case of song diction quite pronounced, since the singers distorted their utterances with untranslatable vocables. The point is worth making because Aztec songs were already beginning to be imitated, sometimes deceptively, by writers using Nahuatl. The songs in the Romances are oral productions.5
Like the old ballads of England and Scotland, which express a broad range of human concerns, the romances, or ballads, of Spain cannot, or should not, be characterized in a word. Yet the twentieth-century scholar and collector of Spanish and Latin American folklore, Aurelio Espinosa, undoubtedly expressed a sentiment of long standing in stating repeatedly that the Spanish romances were to be identified first and foremost as "the old historical ballads that sing the praises of the Castilian heroes in their struggles to reconquer Spain."6 These Spanish ballads, which began to be collected in the 1500s, derive from the celebrated cantares, or epics, of the eleventh and twelfth centuries, notably the Cantar de mío Cid, an epic of 3,735 lines, which itself is divided into three parts, or cantares.
Comparison, therefore, came easily to the sixteenth-century collectors of the Aztec cuicatl 'songs', who found the Nahuatl texts peppered with the vocabulary of war and the names of old kings and dubbed the two principal collections Cantares mexicanos and Romances de los señores de la Nueva España.
Señores, in sixteenth-century Mexican Spanish usage, are 'lords', meaning rulers of pre-Conquest city states 'in all that is now called New Spain' (en todo lo que agora se llama nueva españa), with the term rey 'king' reserved for the most important among them. As for Tenochtitlan, its señores were 'emperors' (los señores della fuerõ emperadores).7
The title of our manuscript, then, with its emphasis on "rulers," may be compared with the Nahuatl heading given to songs XX -XLIII of the Cantares Mexicanos by its glossator:
Melahuac cuicatl 'plain song' is the term used by missionaries for the heterometric canto llano of the church. If the native cuicatl of the kind found in the Cantares and the Romances were heterometric (without regular meter), as seems likely, this could have been a designation applicable to all these songs.8 The three locations refer to pre-Conquest Mexico and its two Triple Alliance partners, Acolhuacan (or Texcoco) and Tepanecapan (also called tlalhuacpan 'dry lands').9 Thus romances de los señores de la Nueva España may be taken as an approximate paraphrase of the heading used for Cantares songs XX-XLIII-four of which in fact recur in the Romances as songs II (Cantares XXV), XIX (Cantares XXVIII), XXIII (Cantares XLIII), and XXVIII (Cantares XL).
In addition to Romances songs II, XIX, XXIII, and XXVIII, eight other songs, or parts of songs, have close variants in the Cantares. These are Romances VI, XI, XIV, XVIII, XX, XXI, XXIV, and XXVII. In all, about thirty-six percent of the Romances reappears in the Cantares.
Minor differences in wording reveal that the songs were not simply transferred from one compilation to the other, and in some cases there is wholesale rearranging of stanzas. Evidently the texts derive from independent recitations.
The editorial style of the two manuscripts is similar, however. In both compilations the stanza form of the songs is clearly indicated, with each stanza presented as a discrete paragraph; and, in both, occasional drum cadences are given, using the syllables to, co, ti, and qui.
Like the Romances, the Cantares is unsigned. Various clues, such as vocabulary items from the Cantares that Sahagún utilized in his Psalmodia christiana and the stylistic resemblance between Cantares folios 60, 60v, and 73 and Sahagún's Memoriales con escolios, indicate that the Cantares may have been prepared under Sahagún's direction,10 as suggested above. A link between Sahagún and the Romances, also mentioned above, is less clear, though not inconceivable.
Like the Cantares manuscript, the Romances seems to have grown by accretion, adding songs in recognizable groups. Yet its songs taken together have a unity lacking in the Cantares. Nearly all have eight stanzas. Only the older rulers are mentioned, none who held office later than the 1520s, and none, even if baptized, who are called in the songs by their baptismal names. Historical events, whether pre-Conquest (i.e., pre-1521) or post-Conquest, are never directly talked about, so that the songs, at least superficially, have a timeless quality even if datable rulers are named.
A natural question is whether the arrangement of songs in the Romances reflects the compiler's taste or the order of native performance on (four?) particular occasions.
It may be noted that each of the four parts (except part three) begins with a kind of incipit or invitational phrase: "Friends, let us sing" (part one), "Now let us begin" (part two), "Begin in beauty" (part four). Not lacking are songs that challenge, or at least doubt, the native doctrine of the warrior's passage to and from the other world, the staple precept of the entire repertory. Yet the final song in each of the four parts of the Romances, or at least parts one, three, and four, reaffirms this religio-militarist dogma.
By contrast, the Cantares is more of a disorganized source book, with short songs and lengthy, rambling songs intermixed; songs with narrative and satiric elements; songs that have Bible stories sandwiched in with native material; and songs that reenact actual battles, including the Conquest of 1521. Far removed, no doubt, from the pre-Conquest mnemonic aids called cuicaamatl 'song sheet(s)' or amoxxotl 'book(s) more or less',11 the Romances manuscript with its neat division into four parts and its presumed date (1582) calls to mind the Spanish verse anthologies, or cancioneros, of the fifteenth and sixteenth centuries, especially the well-known Flores de baria poesía, compiled in Mexico City just five years earlier, in 1577, and neatly divided into five parts.12 With some justification we may therefore refer to the Romances as a Nahuatl cancionero.
In the highly schematized world of the songs all kings and their warriors are teteuctin, pipiltin 'lords, princes', interchangeable terms denoting members of the princely class. Warriors are not called by name unless they are the warrior kings themselves, with exceptions made for a few heroic military leaders and high officials of the city-state of Mexico.
In the Romances the named rulers may be divided into two major categories, each with three subcategories:
When named in combination within a single stanza, the rulers are always from either category 1 or category 2, never mixed, though representatives of both categories may appear within a single song. As is well known, Huexotzinco, Tlaxcala, and Chalco were the traditional enemies of Mexico and its Triple Alliance partners - and allies of Hernán Cortés in the Conquest of 1521. In a few songs Tepanecapan (category 1C) is represented by the city of Azcapotzalco, the seat of Tepanecapan prior to 1430, when its government was moved to Tlacopan. And while Acolhuacan (category 1B) is generally treated as Mexico's ally, in some songs it emerges as an enemy (as in Cantares song 15), since it defected - or a faction defected - to Cortés in the final struggle of 1521.
The singer or singers responsible for the Romances seem to have a particular fondness for songs that name Texcocan rulers. Or else the compiler, if he picked and chose, favored such songs. Number XXIII raises the possibility not merely of selection but of tampering.
Romances XXIII appears also as Cantares XLIII (which is the third in a suite of three songs) and again in Cantares LXII (as the third part in a three-part song). All versions treat "Cuacuauhtzin," presumably the fourteenth-century Cuacuauhtzin who was king of Tlatelolco. But "Cuacuauhtzin" is also the name of a fifteenth-century ruler of Tepechpan, a small city controlled by Texcoco - according to several sources,13 including the anonymous Romances glossator (who labels song XXIII "of Cuacuauhtzin, lord of Tepechpan") and the Texcocan historian Alva Ixtlilxóchitl. Ixtlilxóchitl relates, colorfully, that King Nezahualcoyotl of Texcoco sent the Tepechpan ruler to his death in battle in order to obtain his beautiful wife for himself; and before setting off, this "Cuacuauhtzin" gathered his "friends" (amigos) and sang for them "some sad songs" (unos cantos lastimosos).14 Did Ixtlilxóchitl use the Romances as his source?
The Cantares versions do not mention Nezahualcoyotl, while the Romances version, differently, has the third-stanza verbs in the second person (rather than the first person), interpolating "you, Yoyontzin [an alternate name for Nezahualcoyotl]." As for the "friends" and the "sad songs," the Cantares and Romances texts both have tocnihuan 'friends', cuicatl 'song(s)', nentlamati 'grieve', and choca 'weep', but these terms are formulaic throughout the two manuscripts. It is conceivable that the Romances singer has doctored the song to give it a Texcocan flavor and that Alva Ixtlilxóchitl cooked up the story about the gathering of "friends" and the "sad songs," using the Romances text and his own imagination.
The similar case of song XXVI, in which the singer appears to have adapted Mexica material by inserting the name of the Texcocan ruler Cacamatl, is discussed in the Commentary.
Both the Cantares and the Romances manuscripts have their incautious glossators, who evidently understand Nahuatl but, of course, are versed in the new written tradition. Their training in the oral tradition, if any, cannot be verified.
Both commentators are interested in labeling or detecting pre-Cortésian material. The Cantares glossator, for instance, has labeled Cantares XX-XLIII as songs "that used to be performed in the palaces," as noted above. But since the twenty-four songs in question invoke God, Santa María, and Jesucristo, the description is in need of qualification at the very least. In the same vein the Romances commentator constantly glosses the textual dios 'God' with the expression totecoyo, as he writes it (i.e., toteucyo 'our lord'), a useless observation if he means 'our lord' in the Christian sense. However, toteucyo is pre-Cortésian usage for addressing or referring to the god Tezcatlipoca.15 Further, song XVIII as a whole, which addresses the supreme power as father, God, and ipalnemohua 'life giver', is glossed a lo divino gentilico 'to the pagan divinity'. The name Ipalnemohua apparently did refer to Tezcatlipoca in pre-Conquest usage, yet in colonial texts it much more frequently meant dios.16 As he looks for the idol behind the altar, brushing aside the contemporary reality, the glossator stands at the head of a persistent and very long tradition.
Other songs are seen as eyewitness accounts of specific historical events, a kind of newspaper archive in which we may read the declarations and responses of the key participants. Thus song V, the glossator believes, reports the agony of Cacamatl (and in his own words) as he faced death at the hands of the Spaniards in 1520. Songs XIX and XXVII are also imagined as reportage (on the events leading up to the Tepanec War, ca. 1430, and on a war between Mexico and Huexotzinco, ca. 1498). The similar case of song XXIII and its connection with the historian Alva Ixtlilxóchitl has been reviewed above. Treated elsewhere is the difficult question as to whether this glossator could have been Ixtlilxóchitl himself.17
And it may be asked whether the gloss de Nezahualcoyotzin for songs XVII, XIX, XXVIII, and XXXIII means that the glossator considered Nezahualcoyotl the author of at least these four pieces. His preposition de 'of' need only mean 'pertaining to', not 'authored by'; but it is well known that Alva Ixtlilxóchitl regarded Nezahualcoyotl as a composer (probably on account of the phrase "I, Nezahualcoyotl" in some of the song texts). The internal evidence of the songs themselves, in context with the formulaic diction of the Cantares and Romances as a whole, fails to support Ixtlilxóchitl's conjecture. Nor is there native testimony that any pre-Conquest ruler had (or had not) been a composer or even legendarily a composer. Documentary evidence, frequently claimed to exist, has never been produced. The matter is discussed at length elsewhere.18
Filled with references to dios, the songs in their present form could not have been put together earlier than the Conquest period, 1519-21, and more likely after the beginning of Franciscan evangelization in 1523-24.19 A still later - perhaps much later - date may be required if the subtlety of the Christian allusions in the Romances is fully appreciated.
Time may be needed to convert dios into dios ipalnemohua and to add espiritu santo 'Holy Spirit' (1v:15) and S maria 'Saint Mary' (10:2). Subtler still are the huelica tzihuatl (read huelic cihuatl) 'fragrant woman' at 7:12-13 and the tocuiyc tonatihu (read to[te]cuiyo tonatiuh) 'Our Lord the Sun' at 10:1. These are evidently not syncretic terms for Mary and Jesus, as might seem, but adaptations into Nahuatl of established Old World concepts, in which the mother of Christ is identified with a proverbial garden and Christ himself with the sun.20
In song V there is a reference to "God's loved ones," meaning the dead warriors, or "loved ones," of pre-Conquest lore who reside in the house of the sun. But "God's loved ones" are the saints of Christianity residing as angels in heaven (see the Commentary for song V).
It might be supposed that Christian touches are irrelevant, since the terms could have been tacked onto old songs at the last minute. Such a possibility is hard to entertain, however, since the same phrases in the same positions (mentioning dios) are found in some of the Romances and Cantares duplicates, or near duplicates, that otherwise vary, indicating separate performances if not separate singers.
Further, stylistic unity throughout the Romances and across the Cantares argues for a single school of composition, not various schools scattered over centuries or even decades.
Two of the songs in the Romances, I and X, make use of historical figures from different time frames. Song I has them separated by a generation, song X by several generations. In both cases the more recent figures, mostly from category 2 (see above), belong to the Conquest of 1521. The older figures, proverbial heroes from category 1, rise out of a more distant past; and, in both cases, category-1 figures are given the last word. It is difficult, therefore, to imagine a date of composition very close to the most recent of these time frames (i.e., 1521), since the named rulers and heroes - with chronological distance - seem to have become mere symbols.
Four songs in the Cantares, LV, LVI, LVIII, and LIX, carry the dates 1553, 1550, 1564, and 1536, respectively. The aberrant 1536 is most likely a copyist's error for 1563,21 and the first and third of the four dates should probably be corrected to 1555 and 1562.22 Although none of these four songs has a variant in the Romances, the dates show that songs using the formulaic diction found in both manuscripts were being composed in the 1550s and 1560s.
A fifth date, 1581, can be proposed as an approximate terminal date. On March 27 of that year don Alonso Axayacatl, gobernador, or puppet ruler, of Itztapalapan and nephew of Montezuma II, made his will; presumably he died not long thereafter. Since the songs that carry dates name post-Conquest figures who died shortly before those dates, Cantares song XC, which names don Alonso, was probably composed in the early 1580s, if not 1581.23 No song in the repertory mentions any figure who died later than this.
None of the foregoing is meant to suggest that pre-Conquest songs were not being retrofitted during the years 1550-85, only that these are the dates we have to work with. And in view of the singers' sheer inventiveness, brilliantly displayed in the Cantares, it is unnecessary to believe that native people were mindlessly parroting old songs they did not understand. It should be kept in mind that the Cantares and the Romances are not salvage ethnography, as is the Florentine Codex, but evidence of an active sixteenth-century genre, marked by borrowing back and forth between singers, the reshaping of old material, new composition, and public performances.24
The songs repeatedly deal with the problem of reasserting the old native war ethic in the face of Christian values. In Cantares song VII a singer openly contrasts the lure of battle "in the old days" with the lure of a Christian heaven, "that world of flowers in the sky." In Cantares song LVIII we learn that "through Santa María he came to take his precious incarnation; through his precious death he came to save us, and he gave us everlasting life." The Romances does not include texts as explicit as these. But the ferment is there, as the singers struggle to forge a nativistic doctrine combining old and new elements.25
Linguistic hypertrophism may also serve as an age indicator, deceptively perhaps. One jumps to the conclusion that overloaded verbs and compound nouns prove ancientness. But the opposite may be closer to the truth. Florid locutions such as teoaxochicuauhcocoltica 'with flood-flower eagle sadness' (Romances 1v:7-8) or tiquetzalzacuanxiuhquecholhuihuicomacan 'let's make troupial-and-turquoise-swan plumes twirl' (Cantares 47:24), so unlike the leaner diction in the Florentine Codex, even in the no doubt pre-Cortésian "devil" songs, suggest a combative verbal stance more Aztec than the Aztecs. A sign of nativism in response to stress?26 Similarly choked compounds, like mochipahualizichpochaçucenaxochicelticayotzin 'your pure and maidenly lily-flower freshness' (from a prayer addressed to Saint Mary),27 rear up in the ambitious Nahuatl of missionaries; and even native writers were sometimes forced into weighty neologisms in order to approximate a European concept - such as 'myth' (tlamachiliztlatolzazanilli), literally, 'wisdom-word fable' (CC 75:1). Setting aside the question of who is imitating whom, or whether exhibitionism is at issue, the evidence is for novelty, albeit mingled with tradition.
Songs of the kind found in the Romances were regarded as either benign or potentially dangerous, depending on whether the vantage point was before or after the 1590s. By that decade or a little earlier the songs as we know them from the Romances and Cantares manuscripts seem to have been dying out.28
The two manuscripts, however, were in circulation during the 1600s, along with similar song compilations that have not survived. Charmed by these relics of a presumed distant past, the new writers treated the song texts as (1) fossilized accounts of pre-Conquest doings safely entombed in the amber of history, (2) evidence for the existence of pre-Conquest poet-kings, (3) echoes of glorious old failures on the battlefield, or (4) luscious imagery waiting to adorn a new polite diction in the service of Christianity. Examples of the four approaches, since these have a bearing on the devolution of Aztec song scholarship, may be summarized here:
(1) Chimalpahin (1579-1631?). The historian Chimalpahin, like the Romances glossator, used song texts as reportage, sprinkling his - nevertheless invaluable - chronicles with excerpts meant to add immediacy and color. A passage that has come to light recently in an edition by Arthur Anderson and Susan Schroeder includes the phrase hecamecatlon onmalintoc yn chalco, in Tenuchtitlan a ohuaya ohuaya, which the editors translate, faultlessly: 'Long discourses lie entwined in Chalco and Tenochtitlan'.29 Faultlessly, because this is no doubt what Chimalpahin had in mind. In view of the whole corpus of surviving song texts, however, the more probable reading is he ca mecatlon onmalintoc yn chalco in tenochtitlan a ohuaya ohuaya 'Hey! for the garlands are whirling in Chalco, in Tenochtitlan! Ah! Ohuaya, ohuaya'. 'Garlands', in the diction of the Cantares and the Romances, are incoming warriors from the other world; and while the entire song, had it been preserved, would probably recollect a pre-Cortésian struggle between Tenochtitlan and Chalco, it would also echo the Spanish conquest of Tenochtitlan in 1521, in which Chalco played a role.
(2) Cantares de Nezahualcóyotl (1618). Purportedly dictated to Alva Ixtlilxóchitl but more likely gleaned by him from earlier manuscripts, these cantares include a "lamentation" that begins: Tlacxoconcaquican hani Nezahualcoyotzin etcétera (read Tlaoc xoconcaquican ha niNezahualcoyotzin etcétera) 'Listen indeed to what I, Nezahualcoyotl, etc.'. No further Nahuatl text is given, but Ixtlilxóchitl offers his Spanish translation of the entire seven stanzas, beginning: 'Listen carefully to the lamentations that I, King Nezahualcoyotl, make regarding the empire [ . . . ]. O king [meaning Tezozomoc, chief ruler of the Triple Alliance during Nezahualcoyotl's younger years], [ . . . ] when that time comes after your death [ . . . ] then the mandate and governance of the empire will not be in your hands but in the hands of Dios the Creator and All Powerful'.30 Thus Nezahualcoyotl, whom Alva Ixtlilxóchitl calls a composer of songs "of much morality" and songs "as though they were prophecies," predicts the Spanish Conquest nearly a century before the event.
(3) Discursos en mexicano (1640s). A thinly disguised etiquette manual with an agenda of language improvement,31 the Discursos re-creates an antique milieu by drawing upon various Nahuatl manuscripts from the previous century.32 One passage, showing how a gobernador might elegantly reply to a group of singers who have just entertained him, incorporates phraseology from an actual song: Auh huel ōnonnotlamachtì, huel ic ōnompāc ĭnīc ōnichuālcactoca ōanconmēhuilìquè yāōcuīcatl ic pŏyōmicquè Ācōlhuàquè tlaxcallan yāōtěpānco, ye achi quēxquich cāhuitl in niquelēhuia noconcaquiz33 'Well now, I quite enjoyed myself - I was quite glad to be hearing your good selves sing that war song in which the Acolhuans died gloriously on the battlefield in Tlaxcala; for quite some time now I've wished to hear it!' The translation follows Garibay, who sees pŏyōmicquè as 'died a perfumed death' (poyon-, the combining form of poyomatli, as in Romances 1:13), freely 'died gloriously', which is no doubt what the author was thinking. But an idiomatic reading of the song diction requires pŏyōmicquè to be translated '[are] drunkards' (i.e., "drunk" on blood lust), turning the gobernador's 'I've wished to hear it!' into an unintended joke.34 Here again - as in example 1, above - an echo of the Spanish Conquest (in which Cortés was aided by Tlaxcalans, imagined in the song text as trounced by savagely victorious - not "gloriously" defeated - Acolhuans, who would normally be in league with Mexico).
(4) Laso de la Vega's Guadalupan Legend (1649). Like the Discursos, the famous "legend" is an exercise in antiquarianism, borrowing freely, in this case, from Cantares song I.35 The poor Indian Juan Diego "as he approached the hill called Tepeyaca, just at dawn, heard music . . . ," and, as the story unfolds, the Virgin (of Guadalupe) appears to him, enables him to gather flowers even though it is the winter season, and leaves an imprint of her image on his tilma. Laso's distinctive vocabulary may be compared with Cantares folio 1: tzinitzcan (1:7, LASSO 26:10/LASO 62:1); manoce (1:7, LASSO 26:23/LASO 62:13); nocuexanco nictemaz (1:10, LASSO 42:21/LASO 78:27 has quicuexanten); iuhquin tepetl quinnahnanquilia (1:13, LASSO 26:8/LASO 60:28 has iuhquin quinànanquilia Tepetl); coyoltototl (1:17, LASSO 26:10/LASO 62:1); tonacatlalpan xochitlalpan (1:28, LASSO 26:16/LASO 62:6 has inxochitlalpan intonacatlalpan); xixochitetequi (1v:3, LASSO 42:13/LASO 78:20 has xochitl xictètequi); teyol quima (1v:8, LASSO 26:9/LASO 60:28 has teyolquimà); etc. Cantares song I itself is a deritualized literary composition in which the 'flowers' xochitl have become tokens of divine grace. Standing behind these acculturated xochitl are the xochitl of the oral Cantares (all ninety-one songs in the manuscript except for I-IV, VI-IX, and XI-XIII), representing a somewhat different worldview. Laso further appropriates the flowers, turning them into Caxtillan tlaçòxochitl 'Spanish roses' (LASSO 42:18). With Laso, for whom the old songs were undoubtedly a source, we encounter the first milestone in the history of costumbrismo - the literary manufacture of palatable, nonthreatening folklore.
Observers who knew the songs before the 1590s, however, took a more cautious view.
During most of the 1500s Spanish authorities either forbade native song performances or carefully monitored them, making an attempt to distinguish genres.
Conquerors and early colonists had promptly recognized two kinds of songs, those in honor of "devils" and those in praise of "conquests" or, more fully, in praise of "past kings, recounting wars, victories, deeds, and such things." Both were performed with dancing, but of different styles: the former called macehualiztli in native usage; the latter, netotiliztli.36 Either could be referred to as song or dance, since singing and dancing were integral to all the performances. Evidently the Romances and the Cantares are netotiliztli, a term that will be adopted here - even though in native-language texts it was occasionally used synonymously with macehualiztli and the two terms were sometimes confused or even disregarded by writers of Spanish.37
The macehualiztli were the more worrisome. It was macehualiztli that had precipitated the soon-to-become-legendary Toxcatl massacre of 1520, when 80 (some say 130) of Cortés' men were stationed, uneasily, in Tenochtitlan.38 Made nervous by the exotic, large-scale performance, the Spaniards had ambushed the dancers, slaughtering several hundred. Retelling the event from the vantage point of the 1550s and 1560s, the academician Francisco Cervantes de Salazar, adding his own emphasis, could write that, although the cantares performed on that hair-raising occasion were sacred, en este trataron la conspiración contra los nuestros 'in this business they plotted conspiracy against our own'.39
The chronicler Francisco López de Gómara, the usual source for the Toxcatl massacre, whose description Cervantes borrowed, states explicitly that the Toxcatl performance was macehualiztli, not netotiliztli. Yet Sahagún's detailed Nahuatl-language account of activities for the feast of Toxcatl mentions netotiliztli only and repeatedly; and he indicates the same style of dancing (netotilo, mitotia) during the massacre itself.40
By Cervantes de Salazar's time the macehualiztli had been proscribed at least twice, once by ecclesiastical writ in 1539 and again in the penal code issued in 1546. The latter mandate also outlawed certain other "cantares," apparently netotiliztli, unless they were of the kind taught by the friars, on pain of a hundred lashes for each infraction.41
Nevertheless, netotiliztli of the kind not taught by the friars are known to have been performed in public after 1550 and could well have been witnessed by Cervantes de Salazar. Evidently the obscure figurative language made the texts impervious to censorship. According to Sahagún, speaking of the netotiliztli in particular, "no one knows what [the singers] say except themselves alone." Cheerfully, Durán heard the same songs and found nothing to reprehend "in general," though he conceded that the texts were so obscure that he could not understand them unless they were "explained" (conferido) by the natives.42
Exasperated, Sahagún noted in 1576 that the situation had worsened: "[ . . . ] they sing when they wish [ . . . ]. This continues; every day it grows worse [ . . . ]."43 And in the only work of Sahagún's published in his lifetime, a psalmody of Christian texts designed to replace the dangerous netotiliztli, he inveighed against the native songs in explicit terms, writing in his prefatory remarks that
Yet at what appeared to be the height of its intensity, the native cantares movement was apparently losing momentum. Rather, it was about to change. When it resurfaced in the 1600s the oldstyle stanzas had become coplas in the Spanish mode, though unreported remnants of the older material, much reduced, may well have survived.45 A surprisingly late performance of a michcuicatl 'fish song', evidently netotiliztli, judging by the title, is mentioned in Chimalpahin's Journal for the year 1593.46 The entry is suspect, however, in that the future historian Chimalpahin was only fourteen years old at the time.
What does Sahagún mean when he speaks of tecpas, persuasion, war, and other business that is not good? And why did the situation seem to be "growing worse every day"? Evidently the tecpas (Nahuatl tecpan 'palace') are the houses of the native gobernadores. Durán, more credulous than Sahagún, speaks defensively of these native lords presiding in their tecpas:
As for the "persuasion" and the "worsening," noted by Sahagún, the prolific (and devoutly Catholic) nineteenth-century documentarian Joaquín García Icazbalceta was able to suggest an answer:
Icazbalceta's approach to this matter was to be echoed in succeeding generations by writers who would find that "many examples confirm the observations and pessimism of Sahagún" and "few ecclesiastics were so perceptive, or so outspoken,"49 while historians in the second half of the twentieth century, nevertheless, became more interested in the mechanics of acculturation than in the politics of resistance.
As for Sahagún's "war or other business that is not good," the songs themselves provide a few clues. The reader may take a look at Romances XVII and XXX and decide whether either of these pieces might be read as a call to arms. And note unmistakable allusions to the forbidden topic of human sacrifice in song XXX (ytzimiquixochitli 'flower of knife death'), song VIII (tiçȃtl a y■huitl 'chalk ah! and plumes'), and song XXXVI (motiçȃyo ye [mi]huiyon 'your chalk, [your] plumes'), comparing the same terms in the documentation so scrupulously prepared by Sahagún: jtzimjqujzxuchitl 'flower of knife death' (FC 6:171:30) and ytiçaio yujio 'his chalk and plumes' (FC 2:48:7).
Passages in the netotiliztli that appear to be exercises in philosophy - or, as Alva Ixtlilxóchitl might say, moral philosophy - are expressions of the war ethic and the concept of the warrior as icnotl 'bereaved one' resolved in favor of war (not peace) when the netotiliztli are considered as a whole.50 Moreover, echoes of the Spanish Conquest are heard throughout the corpus.51
These elements, taken together, accord with a minimum definition of what may be called revitalization. Although it came too late to incite rebellion, the Cantares movement, to give it a name, staged a recurring drama of resistance during the ostensibly pacific third quarter of the sixteenth century, celebrating a war ethic at odds with the newly dominant culture.52 If it is understood that in this drama dead warriors are being brought back from the other world - a subject to be explored in the following essay - it will be seen that the songs of the Romances and the Cantares, cryptic though they may be, exhibit a signal feature of the classic revitalization movement; that is, a movement among at least a segment of a subdued population that appeals to the supernatural when the opportunity for open rebellion has passed.53
Out of the ashes of the Cantares movement arose two of the potent symbols of modern Mexican consciousness as it began to take shape in the first half of the 1600s. These were the flower world of Juan Diego - Saint Juan Diego as of the year 2002 - crafted by cleric Laso de la Vega in 1649; and the poet-king Nezahualcoyotl - namesake of a modern city and subject of several twentieth-century biographies - created by the historian Alva Ixtlilxóchitl in 1618. Both are traceable to the netotiliztli: Juan Diego's literary setting to the Cantares Mexicanos, as mentioned above; and the poet-king to the codex Romances.
Although the Cantares contains songs that have been ascribed to Nezahualcoyotl, attention in this regard centers on the Romances, with its suggestive glosses and the pedigree that seems to have been given to it by Pomar and Ixtlilxóchitl. Pomar, who apparently attached the Romances to his Relación, claimed that the religious views of Nezahualcoyotl were to be found in the "ancient songs" (i.e., the Romances?), and Ixtlilxóchitl stated, helpfully, that King Nezahualcoyotl had composed "sixty and some" songs addressed to the Creator (again, the Romances?).54 In fact the Romances has thirty-six songs, thirty-seven if XXIX and XXIX-A are counted as two. Yet Pomar considered the "ancient songs" to be in "fragments"; and the modern scholar Garibay, seeing discrete thematic units, divided the Romances into sixty songs, several of which he divided further in his commentaries.
Nezahualcoyotl's role in modern nation-building - in any event a phenomenon to be viewed through the lens of mythography rather than history - may be better understood in relation to the parallel case of Peru. Like Mexico, Peru faced an overwhelming native heritage, dangerous but also ripe with opportunity. The challenge, for nation builders, was to demonize native resistance while simultaneously integrating the native past. To construct the necessary story, two figures were required.
For Peru these two were Viracocha Inca, eighth ruler of Cuzco but first of the actual emperors (ruled ca. 1402?-40), and Atahualpa, last of the de facto emperors (ruled 1525-33). From legends that came to light in the 1500s we learn that Viracocha Inca was a proto-monotheist who predicted the Spanish Conquest and, with it, the triumph of Christianity.55 Atahualpa, on the other hand, had become a figure of shame, who "murdered left and right" and "came to an evil end as punishment for his cruelty and bloodshed."56
For Mexico the two corresponding figures ought to have been Montezuma I, fifth ruler of Tenochtitlan but first of the actual emperors (ruled 1439-68), and Montezuma II, last of the de facto emperors (ruled 1502-20). In fact Montezuma II did collect the blame. However, since the mythologizing of a scapegoat had been the first order of business, "Montezuma" was no longer an available name by the time the first notice of the prophet appeared, in 1582 (in Pomar's Relación). If for no other reason than to avoid ambiguity, the paradigm shifted sideways, substituting for Montezuma I the only major figure among his contemporaries. This was the ruler of Texcoco, Nezahualcoyotl (ruled 1431-72), staunch ally of Tenochtitlan and relative of Montezuma I - some said his brother.57
Montezuma II, demonized in sixteenth-century legend and never rehabilitated in mainstream literary lore, had now become the "weeping," "fainting" coward who had allowed the empire to fall;58 or, in stronger terms, the arrogant despot "drunk with his own power," addicted to human sacrifice, who ignored warnings issued by the God of Christianity.59 Nezahualcoyotl, on the other hand, objected to the worship of "devils," practiced a virtually Christian brand of monotheism, and predicted the arrival of the Spaniards precisely to the year - 1 Reed (i.e., 1518).60
These paradigmatic figures are not without duplicates and further ramifications. The Inca Mayta Capac, for example, rivals Atahualpa as a legendary barbarian. Topiltzin Quetzalcoatl is better known than Nezahualcoyotl as a proto-monotheist. Spaniards in the sixteenth century identified themselves with Viracocha Inca and with Topiltzin Quetzalcoatl.
The special virtue of Nezahualcoyotl, the quality that makes him modern even for the twenty-first century, lies in his role as poet and humanist.61 Alva Ixtlilxóchitl spoke for future generations when he declared that the songs of Nezahualcoyotl contained phrases to rival "the divine Plato and other great philosophers."62 Equally forward-looking, Ixtlilxóchitl's immediate predecessor, Mexico's first modern historian,63 Juan de Torquemada, could write:
Thus in one breath Torquemada delivers the pastoral diction of European poetry and a hint of the Old World carpe diem theme, features derived from the Greeks (Theocritus, Epicurus) and conveyed by the Romans (Virgil, Lucretius), conferring the unmistakable badge of humanism.64 Xochitl mamani in huehuetitlan actually means 'flowers stand [as a group, over an area] beside the drum' (cf. Romances 13v:3). Torquemada has read huehuetitlan 'beside the drum' as ahuehuetitlan 'beside the cypresses' and has added Virgilian 'copses' for good measure, 'freshening' the flowers to make room for Lucretius. Never mind, for the moment, what these 'flowers' might mean. Nezahualcoyotl, remade and perdurable, has entered the narrative of the new culture.