Dante’s heritage: questioning the multi-layered model of the Mesoamerican universe Jesper Nielsen! & Toke Sellner Reunert” Ancient Meoamericans are generally thought to have imagined the universe stacked in vertical ayers, not unlike the cosmic layers of Dantes Comedy. Dismantling this model, our authors show it 10 be based upon a post-conquest European-Aztec hybrid. This penetrating critique tracks the history of the hybrid cosmos from its first appearance through its resilient repetition until today. Keywords: Pre-Columbian Mesoamerica, Colonial Mesoamerica, cosmology, European influence, Franciscans Introduction A recurring methodological discussion in Mesoamerican research has centred on the extant use of ethnohiscorical and ethnographic analogies (Trigger 1981; Quilter 1996). Although we find this approach both inevitable and productive, the present study will emphasise how important it is to investigate and trace the development of any cultural clement in time and space with extreme care, In the present study we will suggest that a generalisation has taken place regarding the idea of a multi-layered Mesoamerican universe, showing that this cosmic structure has been inferred primarily from post-Columbian central Mexican sources and not from pre-Columbian evidence such as hieroglyphic texts or iconography. Secondly, and more importantly, we put forward the hypothesis that the notion of a multi-layered universe is not genuinely an autochthonous Mesoamerican concept, and that it was only introduced into the area in the sixteenth century, and ultimately derives from European visions of the cosmos, A main objective of our research has been to examine under which circumstances and by whom, sixteenth-century sources on Mesoamerican religion and cosmography were composed. Layers or regions: the topography of the otherworld ‘The early Colonial K’iche’ Maya manuscript known as the Popol Wj describes how the Hero Twins have to face particular ordeals in six houses in the underworld of Xib’alb’a (Christenson 2003: 160-74). Each house is named according to the particular ordeal it "Department of American Indian Languages and Cultures, Insitute of Cross-Culeuntl and Regional Studies, University of Copenhagen, Artllerive; 86, 2300 Copenhagen S., Denmark (Email: jnaelien@hurnkalk) Gronneboj 35 st. t, 2720 Vanlose, Denmark (Email: toke@reunere dk) Received: 14 May 2008; Revised: 18 June 2008; Accepted: 3 September 2008 ANTIQUITY 83 (2009): 399-413 399 ResearchDante's heritage: questioning the multi-layered model of the Mesoamerican universe ‘ ORs Yo’ a ey 8 Figure 1. The well-known model of the Aztec multi-layered universe based on Codex Vaticanus A folios Te-2r (adapted from Nicholson 1971: Figure 7). presents to the protagonists, For instance, the second house is called ‘Blade House’ and is filled with sharp cutting blades, In the description of the setting, it appears thar all houses are arranged on the same level, as though chambers in the cavernous Xib’alb’a, The geography of the Popol Wj underworld in many ways resembles the layout of the Aztec underworld according to the Codex Vaticanus A (c. 1566-89, Glass & Robertson 1975: 186; Miller & Taube 1993: 177-8) (Figure 1). In the Vaticanus A, the first realm beneath the carth is called ‘Water Passage’, while the second is named ‘Where the Hills Clash Together’ (Nicholson 1971: 406-8), In the Popol Wij the Hero Twins begin their journey to Xib'alb'a by passing through great river canyons, that is, narrow spaces where the rocks come together, and they cross the big rivers called Pus and Blood. The realms named ‘Obsidian Knife Hill’ and ‘Where Someone Is Shot with Arrows’ in the Varicamus A can be compared to the ‘Blade House’ in the K’iche' epic, in the sense that they are places where blades inflict pain on humans. The Aztecs called the eighth realm of the underworld “Where Someone's Heart Is Eaten’ and illustrated this place by a carnivore devouring a human heart. The obvious K’iche’ parallel is the ‘House of Jaguars’. Buc while the K’iche’ highland Maya of the mid- sixteenth century presents an underworld horizontally divided into a number of departments or regions, the contemporaneous Aztec Vaticanus A places these departments on top of each other, forming a vertically arranged underworld of multiple layers. Is this simply an example of two alternative Mesoamerican concepts? Or can the difference best be explained by other historical developments and cultural mechanisms? As indicated, ic cannot be verified that the vertical multi-layered structure is an indigenous pre-Columbian model of the cosmos. Thus, all known sources that make references to such a multi-layered cosmological structure are of post-Columbian origin. We therefore suggest that the concept ofa multi-layered universe was introduced by the Spanish intruders, and more specifically the Franciscans and their Dantean world-view with its nine layers in both heaven and underworld. Yer today, the existence of a pre-Columbian multi-layered cosmos is taken for granted. 400Jesper Nielsen and Toke Sellner Reunert Common knowledge and where it came from The majority of textbooks available describe how the ancient Maya and Aztecs divided the heavens and the underworld into a number of layers (e.g. Evans 2004), most often 9 layers in the underworld and 13 layers in the heavens. Miller and Taube (1993: 177-8) refer to Codex Vaticanus A as the source ‘where the 9-13 scheme receives its most explicit and. ample representation’. They add, however, that: ‘The Maya certainly perceived layers of both Underworld and upper world but the notion of nine levels of the Underworld is not specific or universal for the Maya, nor is it for either the Mixtecs or the Zapotecs. This note of caution is important, as it turns out that it is almost exclusively Codex Vaticanus A that is referred to, if any source is mentioned at all, in the broad syntheses of Mesoamerican multi-layered cosmography. Seler was the first to present a lengthy discussion of central Mexican cosmography, including the notion of stratified layers in the heavens and in the underworld. However, he also noted that: ‘In she same way as the earth, this flat, two-dimensional shield (of the underworld] was cur up into nine regions corresponding to the four cardinal directions. Thus, the underworld was divided into nine regions’ (Seler [1923] 1996: 9), and emphasised that there are a variety of ways to perceive the layout of the underworld (Seler (1923] 1996: 3-23). Despite Seler’s nuanced analysis, the Vaticanus A scheme in time has become the standard for understanding ancient central Mexican cosmography. In 1961, Krickeberg’s synthesis of the cosmology of the Postclassic cultures further established what would become the unquestioned fact that the Aztecs believed in a cosmic structure arranged vertically in layers of 9 and 13 (Krickeberg er af, [1961] 1968: 39-40; sce also Lopez Austin 1997: 16-17, 106-8). From the Maya area we have no carly commentators who depict or describe the universe in a way similar to that of the Codex Vaticanus A. For example, Landa in his Relacién (c. 1566) does not mention layers (Tozzer 1941). It would seem that the idea of a 9-13 layer model among the Classic Maya was first advanced by J. Eric $. Thompson. In 1934 he suggested that: ‘it is possible shat the various groups of Maya direction gods |...) were considered to be on different celestial or terrestrial planes (Thompson 1934: 216, emphasis added). Thus, directional gods were transferred to vertical layers on a hypothetical basis. ‘Twenty years later he was more certain: ‘They appear to have believed that the sky was divided into thirteen compartments, in which of each certain gods resided. These may have been thought 10 be arranged as that number of horizontal layers one above the other’ (Thompson 1954: 225-6, emphasis added). Finally, in Maya History and Religion (1970) Thompson left no room for doubt: ‘there are thirteen layers of the skies [...] just as there are nine layers of the underworld’ (1970: 280, emphasis added). Eventually, the idea of a multi-layered cosmos became so deeply embedded in Mesoamerican research that it was automatically accepted as a genuine pre-Columbian concept (e.g. Villa Rojas in Leén-Portilla 1988: 135). The cosmic model soon influenced interpretations of Mesoamerican architecture. Thus, Klein (1975), drawing on Krickeberg’s research, and Van Zantwijk (1981) published interpretations of temple pyramids based on the references to layered cosmologies documented in Colonial sources. The assumed nine layers of the underworld are also alleged to be represented in the architecture of certain Maya sites (Carlson 1981: 154). We must stress that we do not 401 ResearchDante’ heritage: questioning the multi-layered model of the Mesoamerican universe per se reject the association between the underworld, the number nine and architecture. Thus, the so-called Twin Pyramid groups of Tikal, which are thought to be architectural loorwayed buildings on their south side (Ashmore 1991), and are pethaps emulating the series of horizontally arranged houses of the underworld known from the Popol Wj. cosmograms, include nine Mesoamerican visions of the universe In the Mesoamerican data, it readily becomes apparent that a four-part image of the universe is by far the dominant scheme. This horizontal structure, based on a central middle point and the four world directions, can be traced further back into pre-Columbian Mesoamerica, from the Postclassic and as far back as to the Preclassic (Nicholson 1971: 403-6; Freidel er al. 1993: 71-3, 128-31; Miller & Taube 1993: 77-8, 186-7). A well-known representation of this horizontally divided cosmos is found on the first page of the Mixtec Codex Fejérvdry Mayer (Figure 2), where the god of fire is shown standing at the centre of the earth, while two gods are associated with each cardinal direction, thus amounting to nine gods (see also the Maya Codex Madrid, pp. 75-6). The gods and trees of the four directions are framed by a cross pattéc with the arms of the cross oriented toward the four cardinal directions, but counting the intercardinal loops and the centre the number nine is rendered once more (Boone 2007: 114-7). Thus, we have nine horizontal divisions of the cosmos. In contrast, pre-Columbian iconography and inscriptions do not make any unequivocal references to a layered universe. Deities, supernatural beings and places, in particular among the Classic Maya, often appear with numerals 1, 3, 5, 7, 9 and 13 prefixed, but there is currently no evidence that any of these numerals, and more specifically 7, 9 and 13, can be associated with vertical layers rather than horizontal regions of the cosmos, As already mentioned, the depiction of the Mesoamerican cosmos most commonly referred to is the illustration of the layered universe in the Vaticanus A (folios Iv-2r), However, if we compare the Vaticanus A representation with the Franciscan friar Bernardino de Sahagiin’ derailed description of the Aztec underworld we find some notable discrepancies. Thus, the latter has no reference to layers, and the text rather provides us with the impression of moving through a continuous horizontal landscape. If we consult the appendix of Book 3 of the Florentine Codex (or Historia General), we tead that: ‘all who died went to [one of | three places when they died, The first place was there in the place of the dead |Mictlan]'. A prayer accompanied the deceased and the goods placed in the grave: ‘Here is wherewith thou wilt pass where the mountains come together. And bere is wherewith thou wilt pass by the road which the serpent watcheth. And here is wherewith thou wilt pass by the blue lizard, the xochitonal. And here is wherewith thou wilt travel the eight deserts. And here is wherewith thou wilt cross the eight bills, Here is wherewith thow wilt pass the place of the obsidian-bladed winds [...). And also they caused him to take with him a little dog, a yellow one {...]. It was said that it would take [the dead one] across the place of the nine rivers in the place of the dead. And [...] there was arrival with Mictlan tecutl [...]. And there in the nine places of the dead, in that place there was complete disappearance. (Anderson & Dibble 1978: 41-4) 402Jesper Nielsen and Toke Sellner Reunert Figure 2. The casmograra of Codex Fejérutry Mayer, p. 1 (© National Museums Liverpool, World Museum Liverpoo)) In this description there is no reference to layers or to a downward movement from one place to another. The journey of the dead to their final destination, described as the ninth place, could equally well - or perhaps better ~ be interpreted as a journcy through regions situared on the same level, and termin: tral location. Those who did not arrive in Mictlan came to Tlalocan or 40 the home of the sun, in heaven’ (Anderson & Dibble 1978: 47-9). Again we find no indication of a multitude of layers, buta rather basic contrast between ‘down’ and ‘up’. Another of Sahagdn’s texts, the Primeros Memoriales contains a reference to chicunavatenco’, ‘the ninth underworld’, but this location is not explicitly described as being situated below other underworlds (Sa hagtin 1997; 177-8; see also Anderson 1988: 156-9). A much later source, Ruiz. de Alarcén's Treatise on the Heathen Superstitions (1629), also provides important information. Summarising some of Ruiz de Alarcén’s observations, Andrews and Hassig conclude that: ‘The native view of the supernatural world persisted in 403 ResearchDantes heritage: questioning the multi-layered model of the Mesoamerican tniverse the concepts of a celestial realm and an underworld, The celestial realm Topan (‘Above-us’) also called Chiuenauh Topan (“Nine-Topan") and Chiuenauhtlanepaniuhean (“Nine-layering- place’), has nine levels, although these are not discussed separately, nor are internal distinctions drawn’ (Andrews & Hassig 1984: 21). The key term is Chiucnauhtlanepaniuhcan, and Ruiz de Alarcén originally cranslated this as ‘a las nueve juntas o emparejamientos’ (1984: 224). Andrews and Hassig consider thisa misinterpretation, hence their cranslation ‘Nine-layering- place’. However, no part of the word can be translated as ‘layering’ or ‘layer’. ‘Nepaniul’ is a form of the verb ‘nepanoa’, meaning juntarse una cosa con otra’ (Karttunen 1992: 169). Thus, the term simply designates a place where nine entities are united or fitted together (Una Canger, pers. comm. 2007), and Ruiz de Alarcén’s original translation is not erroneous. On the contrary, Andrews and Hassig’ interpretation of the term may be coloured by the expectations generated by ‘common knowledge’, and Chiucnauhtlanepaniuhcan could equally well refer to a horizontal division of the heavens. As mentioned, none of the early Colonial sources on the Maya make reference to layers, whereas the importance of the four cardinal directions and the centre is referred to in several instances (Tozzer 1941: 135-9; Christenson 2003: 65). Only in much later Yukatek Maya sources do we find references to a layered structure of the cosmos. In the Book of Chilam Balam of Chumayel we encounter the ‘9-God’ and the ‘13-God’, and they appear in the context of the thirteen fas or ‘layers’ of the heavens. The same source mentions the ‘the seventh stratum of the earth’ and ‘the four layers of stars’ (Roys 1933: 31-2, 99-101). Ir would thus seem that the link berween the 9 and 13 gods and cosmic layers derive from the Chumayel manuscripr, and thar modern scholars have assumed that a similar belief existed in pre-Columbian times. However, it is important to stress that by the time the Chilam Balam of Chumayel was written in the late eighteenth century, Yukatek Maya culture had long been influenced by Euro-Christian concepts and genres, as is evident throughout large parts of the book (Mignolo 1995: 204-7; Graham in press). An influence from European cosmology is also apparent in the Book of Chilam Balam of Kawa, which includes a European geocentric, spherical model of the universe (Bricker & Miram 2002: 92). According to Bricker and Miram the Maya terms for ‘layers’: ‘tas and yal refer not only to the “layers” of the sky in these Books of Chilam Balam but also to the celestial spheres of the European universe in the Kaua’ (Bricker & Miram 2002: 36). Altar 3 from the Late Classic Maya site of Altar de los Reyes (Campeche, Mexico) bears a hieroglyphic text referring to the ‘divine lands, the 13 lands in combination with 13 Emblem Glyphs of Maya polities, strongly suggesting that the Maya of this region acknowledged 13 subdivisions or regions of the earth (Grube 2003). It is possible that the geography of the underworld mirrored a similar horizontal structure coupled with the number 13. Knab’s account of the geography of the underworld from San Miguel Tzinacapan (a Nahua-speaking community in Puebla, Mexico) supports this hypothesis (Knab 1991, 1993), According to the shamans of the village, the underworld (called “Talocan’) in many ways mirrors the earthly level and its organisation. The underworld is organised according to the four cardinal directions, and one of Knab’s informants relates that inside the underworld there ‘are fourteen of everything |...]. Thirteen outside the center for us, and one of everything inside for the Lords’ (Knab 1993: 63). In Sahagiin’ Primeros Memoriales the numbers associated with geography of the underworld were 8 and 9, while among the San Miguel Tzinacapan shamans itis the 404Jesper Nielsen and Toke Sellner Reunert number 14 that has gained importance in the division of the underworldly terrain, and we hear of 14 rivers, lakes, hills fields, roads, county seats etc. in the underworld (Knab 1991: 47). References to layers, however, are absent, The above-mentioned groups of 9 and 13 deities may thus originally have resided over horizontally arranged regions rather than layers. To summarise: the relatively few sources describing a multi-layered universe arc all of post-Columbian origin, and are almost certainly influenced by Euro-Christian concepts regarding the structure of the cosmos. Far more common, in both pre- and post-Columbian sources, is an emphasis on the cardinal directions, the centre and a basic three-tiered cosmos. This is surprising if the multi-layered cosmos was indeed part of a common and widespread pre-Columbian world-view, which would have further facilitated a merging with the Euro- Christian cosmology. Dante’s universe ‘Ac the time of the conquest, the European view of the universe was predominantely that of the Franciscan and Dominican friars and equated to a Dantean cosmology. Although Dante Alighieri (1265-1321) did not belong to either the Franciscan or the Dominican order himself, he was strongly influenced by both. During his youth Dante attended philosophical and theological schools and became familiar with the central Franciscan and Dominican teachings (Hawkins 2007). Dante wrote his masterpiece, the so-called Divine Comedy, from 1308-1321, i.e. some 200 years before the Spanish invasion of Mesoamerica. Nevertheless, his description of the geography of Heaven and Hell was widespread in late fifteenth and early sixtcenth-century Europe, where the Franciscans were active in both translating and commenting on the Comedy (Iriarte 1984: 131). During this period the Comedy had tremendous influence on Spanish theological poetry (Friederich 1949: 45-7), and reached the status of a quasi-biblical text, based on its own claim to have been inspired by a vision. In the Comedy, Dante travels through the three realms of the dead (Inferno, Purgatory and Heaven). The Inferno consists of nine stacked circles, resembling an inverted pyramid, In each of the nine circles, the dead suffer according to a gradual increase in wickedness, culminating at the centre of the earth, where Satan resides. While this underworld is placed below the holy city of Jerusalem, Purgatory is found on the exact opposite side of the world, The earthly Purgatory consists of a mountain with seven terraces, each corresponding to one of the deadly sins, but including the Antepurgatory and the Terrestrial Paradise it has nine layers, thus corresponding to the nine circles of Hell, and the nine spheres of Paradise. fier climbing the seven terraces, Dante reaches the Garden of Eden, and can now approach the heavenly spheres and the residence of God (Figure 3). Dante's heaven represents the medieval equivalent of Aristotle’s nine heavenly spheres, and adopting this model, Dante formulated a theological-astronomical geography of the cosmos, consis the underworld, seven or nine layers on earth, and nine layers of the heavens. We know that Dante’s work did reach Mexico in the early Colonial period. The inventories of the library of the school in Tlatelolco of 1572, 1574 and 1582 do not mention the Comedy (Mathes 1982) but a detailed invoice from a shipment of books from Sevilla to San Juan de Ulda, Veracruz, in 1600 refers to two copies of Dante in Italian (Green & Leonard 1941: 12). Furthermore, a sixteenth-century copy of the Comedy is held in the National Library ing of nine layers of 405 ResearchDante’ heritage: questionin uulti-layered model of the Mesoamerican universe orence (1.465) in the cathedral of Santa Maria del Figure 3. Domenico di Micheline’ fresco The Comedy Illuminatin Florence, thal. Nove the teraces of Purgatory and Garden of Eden in the background, as well asthe divisions of Ineavens (©) Photo SCALA, Florence in Mexico City and may well originally have belonged to one of the convent libraries (Pablo Escalante Gonzalo, pers. comm. 2008) What is crucial to our argument, however, is not whether the Franciscan friars in Mexico had one or several copies of Dante's Comedy in their possession, but that they were familiar with the Dantean multi-layered cosmology, and that this world-view was a natural point of counts of the geography of the cosmos. 9 the book Rhetorica Christiana was printed in Perugia, Italy, written by the Mexican-born Franciscan Diego de Valadés (153. possibly a native Tlaxcalan woman (Edgerton 2001: 114), ‘While Valadéss main subject was to explain in detail the art of literary rhetoric — how language can most effectively serve the reaching of Christian doctrine — he |...] believed that the same techniques and intention were erton 2001: 116), One of the illustrations in the Rhetorica shows Valadés’ mentor Friar Pedro de Gante (c. 1480-1572) preaching to the reference in the interpretation of the local In 15: 322), son of a Spanish conquistador and equally applicable t0 the visual arts’ (Edg Indians by means of pictures. A text accompanying the same illustration states: “Because the Indians lack letters, it was necessary to teach them by means of some illustrations’ (cited after Edgerton 2001: 116). According to Ricard, another friar, one Luis Caldera who did not 106Jesper Nielsen and Toke Sellner Reunert know the language of the Indians he was evangelising, was using identical visual means of persuasion: ‘Carrying large pictures representing the sacraments, the catechism, heaven, hell, and purgatory, he wene from village ro village’ (Ricard 1966: 104). Some of the pictures used for teaching the catechism were also represented in the Valadés’ rhus, three full-page sized images are showing heaven, earth and hell as divided into layers, much like the Dantean layout of cosmos. One of these illustrations shows God (with Jesus reclining in his lap) sitting on his throne inside a cloud at the uppermost layer, while the devil being worshipped by sinners is placed at the bottom of the page (Figure 4). The layout of this illustration and the utilisation of it by the friars sum up one of the crucial points of this article. It seems quite probable thar the multi-layered world-view was taught and referred to in the religious schools of New Spain, and thus reached the native literati that were working for the friars or independently. Rhetorica. ‘| Painting the universe — but whose universe? We can now approach the question of how and why the vertical structure was adopted in New Spain. To provide an answer, we need to examine the carly Colonial context in which documents such as the Codex Vaticanus A were produced. First, it is important to recognise the biases of the Europeans who recorded their impressions of the New World. The friars primarily worked froma preconceived model of the pagan ‘alternative’ to Christianity, based on Greco-Roman religion. ‘Todoroy points to a series of equivalences between the Aztec and Roman gods in Sahaguin’s description of the Aztec deities, and refers to one example where the goddess Chicomecoatl is called another Ceres (Todorov 1984: 231-3; see also Keen 1971). However, images like the cosmogram of the Vaticanus A should not be seen as a result of Franciscan attempts to conceptualise the Mesoamerican cosmography according to a Euro-Pagan tradition, Rather, we suggest thar such images are a result of the appropriation of Euro-Christian ideas by well-educated Nahuas. Many of the documents produced in the early Colonial period were not only the work of the Franciscans themselves. Indian scribes and informants thar had been taught by the friars at schools such as the Colegio de § Cruz in Tlatelolco, central Mexico, contributed considerably co the manuscripts, and not least ro the production of images (e.g. Ricard 1966; Kobayashi 1985; Bremer 2003). The Tarelolco-school opened in 1536, and here the sons of the Indian nobility, the colegiales, were taught grammar, geometry, astronomy and elemental theology. Some colegiales were also trained in the arts such as: ‘the figural styles and iconographic system of late Gothic and Renaissance Europe’ (Boone 2003: 145). Bremer describes the students of the Franciscan schools as hybrids with a bicultural education, being ‘neither natives nor Europeans’ (2003: 20), and he points out how Sahagan's Historia General was the work not only of Sahagin himself, but also of ‘he hybrid colegiales, Indians by birth and quasi-Europeans by education and training. Escalante Gonzalbo provides abundant examples of how native artists copied images or altered images from the Bible or from books on natural history, and also noted the strong European influence in the illustrations of the Historia General (Escalante Gonzalbo 2003; see also Peterson 1988). Clearly, the Franciscan schools provided an intellectual environment in which an almost inevitable blending of old and new was the norm. It was in nt 407 ResearchDantes heritage: questioning the multi-layered model of the Mesoamerican wniverse Che ay SUT CK ven, earth and bell, divided into layers (1579). (© Benson Latin American 408Jesper Nielsen and Toke Sellner Reunert es Se were SX Figure 5. The first two scenes of the sacred narnatve of the Selden Roll (MS. Arch. Selden A. 72(3), © Bodleian Library Osford University) this bi-culeural setting that many of our primary sources on the nature of the Mesoamerican cosmic structure were formed, in text and image. ‘The cosmological model of the Vaticanus A fits into the pattern of production of syncretic religious images in the early Colonial period thar we have outlined above. The original cosmogram runs over two pages (Iv-2r) and shows the Aztec creator god ‘Tonacatecuhtli seated in the thirteenth and uppermost layer, Omeyocan (Figure 1). Below, in layer 2, we find the moon, just above the earth, Tlalticpac, which constitutes the first layer of the heavens as well as of the underworld. The ninth and final realm of the under- world is Mictlan, the ultimate station of the dead. Another representation of the heavenly layers can be found in the Selden Roll, a sixteenth-century Mixtec manuscript (Burland 1955) (Figure 5). The first part of the codex shows nine layers, with the sun and moon occupying the first, lower level. A path, marked by foot prints, penetrates the sequence of layers marked with stars, In the ninth and uppermost layer the culture hero Lord 9 Wind is seated between a man and a woman both named 1 Deer. couple is equivalent to the Aztec ancestral couple, the gods of creation and procreation Omerecuheli and Omecihuatl. The primordial pair was said to reside in Omeyocan, a mythical place of duality, birth and descent (Miller & Taube 1993: 40-41, 127-8, 154). Interestingly, in pre-Columbian times primordial couples found in most of Mesoamerica, 409 ResearchDantes heritage: questioning the multi-layered model of the Mesoamerican tniverse were associated with caves and the earth rather than the heavens (Nielsen & Brady 2006), In contrast to the Vaticanus A scene we do not find a corresponding image of the underworld in the Selden Roll. Instead, a row of footprints enters the open maw of the earth monster and reaches Chicomoztoc (‘Seven Caves’). Thus, the Selden Roll does not present the underworld as a mirror image of the layers of the sky, but couples the latter with an Aztec mythic place name associated with the number seven. The seven caves are situated within a mountain and are arranged at the same level, much like the houses or caves of the Popol Wij. In fact, the multi-layered heaven of the Selden Roll represent an anomaly in a composition and arrangement that in many other respects reflect pre-Columbian concepts and conventions. The placement of the creator couple in the upper layer also seems co be heavily inspired by Euro-Christian and Dantean_ imagery. couple constitutes a near perfect parallel to Adam and Eve as they are standing in the Garden of Eden at the summit of Mount Purgatory in the famous fresco The Comedy Iluminating Florence by di Michelino (1417-1491) (Figure 4). Dating to 1465 the fresco shows Dante holding the Divine Comedy. Behind Dante the terraces of Purgatory with Adam and Eve at the top and the heavenly layers can be discerned. Another depiction of the Euro-Christian cs from Oresme’s Book of the Bibliotheque Nationale Heaven and the World from 1377, which shows god seated in the uppermost of nine layers of stars, with the moon occupying the lowest level (Figure 6). Both the Vaticanus A and the Selden Roll thus evidently merge pre-Columbian elements and deities with European and Christian concepts, and parallels seem to have been sought: native creator gods and ancestral couples replace Adam and Eve or the Almighty; Omeyocan, the mystical abode of the first owo humans is likened to Paradise, and the original division of heaven and underworld in a number of horizontal regions are transformed into a vertical structure akin to the illustrations the native artists and colegiales would have seen during teachings and sermons and copied from volumes in the Francis The Aztec ancestral Figure 6. It World (1377) by Nicole Oresme: de France, Paris) imination from the Book of the Heaven and the Cosmos com n libraries. None of the two cosmological layouts discussed here are therefore exact copies, but rather reinterpretations, of the Dantean cosmology. The illuscrations of the Vaticanus A and the Selden Roll are ‘neither native nor European’ to use Bremer’s phrase (2003: 20); they both reflect a Colonial period syncretism of a native central Mexican concept of the universe with the Euro-Christian Dantean cosmology. Conclusions Reviewing the evidence bearing on the Mesoamerican structure of the cosmos, we have argued that the long-assumed multi-layered model cannot be maintained. Rather, our 410Jesper Nielsen and Toke Sellner Reunert examination has revealed that a basic three-tiered model combined with a strong emphasis on the horizontal divisions of each layer is more likely to have been the dominant scheme before the Spanish invasion and the introduction of Euro-Christian ideas relating to the cosmos. We have also shown how the two crucial images from the Codex Vé the Selden Roll can be explained as the product of culturally hybrid native artists. We do nor reject the possibility of a vertically scructured universe (beyond the three basic layers) in pre-Columbian Mesoamerica. What we argue is that all evidence pointing to a belief in a multi-layered universe is post-Columbian and produced in a context where Euro-Christian ideas were deliberately imported, taught and adopted. The multi-layered Mesoamerican universe is thus almost certainly a hybrid of original pre-Columbian and imported Euro-Christian visions of the cosmos. Our conclusions exemplify how at times ‘common knowledge’ and generalised terms and concepts need to be reviewed. Many arguments, scientific narratives and ‘wuths’ clearly and unsurprisingly consist of several layers of preconceived assumptions and generalisations. anus A and Acknowledgements ‘We would like to express our thanks to a number of colleagues who have been helpful during the process of researching and writing this article. This includes Una Canger, Pablo Escalante Gonzalbo, Mectelise Fritz Hansen, Elizabeth Graham, Christophe Helmke, Stephen Houston, Bodil Liljefors Persson, Jorgen Nybo Rasmussen, Karl Taube and two anonymous referees. Any errors and misinterpretations, however, remain oursole responsibility References ~ 2007. Gyete of time and meaning in the Mexican books ANDERSON, AJ.O. 1988, Sahagéin’s informants on the fie Aan CT) ey Tes Det nature of Tlalocan, in JJ. Klor de Alva, HB. BreMeR, T'S, 2003. Reading the Sahagéin dialogues, in J.B Schwaller (ed.) Sehagiin at 500: essays on the quincentenary of the birth of Fr. Bernardino de Sabagiin: 11-29. Berkeley (CA): Academy of Nicholon & E. Quifiones Keber (ed.) The work off Bernardino de Sahagin-pioneer etluographer of sixteenth-century Aztec Mexico: 151-60, New York (NY) & Austin (TX): Instirute for Mesoamerican, ‘American Franciscans History. Studies, University at Albany, State University of BRICKER, V.R. & H. MIRAM, 2002. An encounter of two New York & University of Texas Press. worlds: the Book of Chilam Balam of Kaua (Publication 68). New Orleans (LA): Middle American Research Institute, Tulane Universi ANDERSON, AJ.O. & C.E. Distt (transl). 1978 Florentine Codex — general history of the things of New Spain by Fray Bernardino de Sahagtin. Book 3: the BROTHERSTON, G. 1995, Painted books from Mexico. origins of the gods. Santa Fe (NM): School of London: British Museum Press. American Research & University of Utah Press. Buriann, CA. 1955. The Selden Roll — an ancient ANDREWS, J.R. & R. HASSIG. 1984, Treatise on the Mexican picture manuscript in the Bodleian Library Ineathen superstitions that today live among the «at Oxford (Monumenta Americana 2). Berlin: Gebr. Indians native to this New Spain, 1629 by Hernando Mann. Ruiz de Alarcén, Norman (OK): University of CARLSON, J.B, 1981. A geomantic model for the Oklahoma Press. inferprecacion Of Mesoamerican Stes: ai esti ti Asttvore, W. 1991. Site-planning principles and cross-cultural comparison, in E.P Benson (ed.) concepts of directionality among the ancient Maya. Mesoamerican sites and world-vieus: 143-215. Latin American Antiquity 2: 199-226. Washington (DC): Dumbarton Oaks. Boone, E.H. 2003. The multilingual bivisual world of | CHRISTENSON, A.J. 2003. Popo! Vis, the sacred book of Sahagiin's Mexico, in J.B. Schwaller (ed.) Sahagiin at the Maya: the great classic of central American 500+ esays om che quincentenary af the birch of Fr: spirituality. New Nltesford: © Books. Bernardino de Sahagiin: 137-66. Berkeley (CA): Codex Fejérviry Mayer. 1971. Graz: Akademische Academy of American Franciscan History. lagsanscat 4n ResearchDante’ heritage: questioning the multi-layered model of the Mesoamerican universe Codex Madrid (Codex Tio-cortesianus). 1967, Graz: ‘Akademische Druck-und Verlagsanstale Codex Vaticanus A (or Codex Vaticanus 3738). 1979. ‘Graz: Akademische Druck-und Verlagsanstat. EDGERTON, S.¥. 2001. Theaters of conversion: religions architecture and Indian artists in colonial Mexico Albuquerque (NM): University of New Mexico Press. ESCALANTE GONZALBO, P. 2003. The painters of Sahagin's manuscripts: mediators berween two worlds, in JE Schwaller (ed.) Sabugiin at 500: essays on the quincentenary of the birth of Fx. Bernardino de Saag: 167-91. Berkeley (CA): Academy of American Franciscan Histo EVANS, $.T. 2004. Ancient Mexico and Central America archaeology and culture history. London: Thames & Hudson. FaetpeL, D., L. ScHELE & J. PARKER. 1993. Maye cosmos: three thousand years on the shaman’ path. New York: Quill, William Morrow. Faiepenict!, W.P. 1949. Dante through the centuries. Comparative Literature 1: 44-54, GLASS, JB, & D. ROBERTSON. 1975. A census of native idle American pictorial manuscripts, in HF, Cline (ed.) Handbook of middle American Indians Volume 14, Part 3: 81-252. Austin (TX): University of Texas Press GRAHAM, E. In press, Close encountets, in LG, Cecil & T.W, Pugh (ed.) Maya worldviews at the conguest. Niwor (CO): University of Colorado Press. Gti, O.H. & LA, Leonann, 1941, On the Mexican booktrade in 1600: a chapter in euleural history: Hispanic Review 9: 1-40. Gnune, N. 2003. Appendix 2: epigraphic analysis of Altar 3 of Altar de los Reyes, in I. Spraje (ed.) Arclracological reconnaissance in southeastern Campeche, México: 2002 field season. Available at wwewamsiong/reports. Hawkins, PS. 2007. Dante and the Bible, in R. Jacof® ed.) The Cambridge companion to Dante: 125-40. Cambridge: Cambridge University Press. IRIARTE, L. 1984. Der Frunziskusorden: Handbuch der franziskanischen Ordensgeschichte. Alconting: Verlag det Bauerischen Kapuriner, KARTTUNEN, B. 1992. An analytical dictionary of ‘Nahuatl. Norman (OK): University of Oklahoma Press. KEEN, B. 1971. The Aztec image in Western hough. New Brunswick (NJ): Rurgers University Press KLEIN, C. 1975. Post-Classic Mexican death imagery as a sign of cyclic completion, in F.P. Benson (ed) Death and the afterlife in pre-Columbian America: 69-85. Washington (DC): Dumbarton Oaks KNB, TJ. 1991. Geografia del inframundo. Eieudios de Cultura Nabuatl 21: 31-57. ~ 1993. war of witches a journey into the underworld of the contemporary Aztecs. New York (NY): Harper Collins. Koaavasti, IM. 1985. La educccin comom conguitta (empresa francixcana en México). Segunda edici6n. Mexico City: Colegio de México. Krickeserc, W., H. TRinsory, W. MOttER & O. Zeke. 1968 [published in German in 1961]. Pre-Columbian American religions, London: Weidenfeld & Nicolson. LeON-PoRTILLA, M. 1988. Time and reality in the thought of the Maya (2"4 enlarged edition. by Alfonso Villa Rojas). Norman (OK): ‘of Oklahoma Press, Lopez AUSTIN, A. 1997. Tamoanchan, Tlalocan: places of mist. Niwot (CO): University of Colorado Press. Marites, M. 1982. Santa Cruz de Tlatelolco: la primera biblioteca académica de las Américas. Mexico City: Archivo Historico Diplomitico Mexicano, Secretaria de Relaciones Exterior MIGNOLO, W.D. 1995. The darker side of the Renaissance: literacy, territoriality, and colonization. ‘Ann Arbor (MI): University of Michigan Press. MILLER, M. & K. TaUe, 1993. Am illustrated dictionary of the gods and symbol of ancient Mexico and the ‘Maya. London: Thames & Hudson. NICHOLSON, H.B. 1971. Religion in pre-Hispanic Central Mexico, in G.F Ekholm 8 L. Bernal (ed,) Handbook of middle American Indians. Volume 10, Part 1: 395-446, Austin (TX): University of Texas Press. NIFISEN, J. & JAE, BRADY. 2006, The couple in the cave: origin iconography on a ceramic vessel from Los Naranjos, Honduras. Ancient Mesoamerica 17 203-17, PereRsoN, J.F 1988. The ‘Florentine Codex’ imagery and the Colonial “Tlacuilo’, in J.J. Klor de Alva, HB, Nicholson & E. Quifiones Keber (ed.) The work of Bernardino de Sahagun: pioneer elnagrapher of sixtcenth-century Aztec Mexico: 273-93. New York (NY) & Austin (TX): Institute for Mesoamerican Studies, University of Albany, Stace University of New York 8 University of Texas Press. QUIETER, J. 196, Continuity and disjunesion in. pre-Columbian art and culture. RES 29130: 303-17, Rican, R, 1966. The spiral conguest of Mexico. Berkeley (CA): University of California Press. Roys, RL. 1933. The Book of Chilam Balam of ‘Chummayel. Washington (DC): Carnegie Institution of Washington, SAHAGUN, B. DE, 1997. Primeros Memoriales (srans, and notes by T.D. Sullivan), Norman (OK) & Madrid: University of Oklahoma Press & Patrimonio Nacional and Real Academia de la Historia. 412Jesper Nielsen and Toke Sellner Reunert Sein, E. 1996 [published in German in 1923]. The ‘ToDoRav, T. 1984. The conquest of America: the ‘world view of the ancient Mexicans, in FE. question of the osher. New York (NY): ‘Comparato (ed.) Collected works in Mesoamerican Harper & Row. linguistics and archaeology. Volume 5: 3-23, Tozzer, AM. 1941. Land relaciin de ls cos de Lancaster (CA): Labyrinthos, Yucatén ~ a translation (Papers of the Peabody THOMSON, JES. 1934. Sky bearers, colors, and ‘Museum of American Archaeology and Frhnology directions in Maya and Mexican religion. Carnegie 18). Cambridge (MA): Peabody Museum of Institution Conzribusions to American Archaeology American Archaeology and Ethnology. é 2(10)}<209-42 TRIGGER, B.G. 1981. Archaeology and the ~ 1954, The rise and fall of Maya civilization, Norman ‘ethnographic present. Anthropolagica 2 (OR): University of Oklahoma Press. Vatabts, D. 1579. Rbetorice Christiana, Perugi ~ 1970, Maye history and religion, Norman (OK): Van ZANTWHK, R. 1981. The Great Temple oF University of Oklahoma Press. “Tenochtitlan: model of Aacec cosmovision, in EP Benson (ed.) Mesoamerican sites and world-views: 71-86. Washington (DC): Dumbarton Oaks. 413