Читать книгу The Codex Mendoza: new insights - Barbara E. Mundy - Страница 9

CHAPTER 2

The Painting Materials of Codex Mendoza

Davide Domenici

Università di Bologna

Chiara Grazia

Centro di Eccellenza SMAArt

(Scientific Methodologies applied to Archaeology and Art)

Università di Perugia

David Buti

CNR-ISPC (Istituto di Scienze del Patrimonio Culturale)

Laura Cartechini

CNR–SCITEC (Istituto di Scienze e Tecnologie Chimiche "Giulio Natta")

Francesca Rosi

CNR–SCITEC (Istituto di Scienze e Tecnologie

Chimiche "Giulio Natta")

Francesca Gabrieli

CNR–SCITEC (Istituto di Scienze e Tecnologie Chimiche "Giulio Natta")

Virginia M. Lladó-Buisán

The Bodleian Libraries, University of Oxford

Aldo Romani

Centro di Eccellenza SMAArt (Scientific Methodologies applied to Archaeology and Art)

Università di Perugia

Antonio Sgamellotti

Accademia dei Lincei

Costanza Miliani

CNR-ISPC (Istituto di Scienze del Patrimonio Culturale)

Introduction

What we present herein is the first scientific study of the Codex Mendoza’s painting materials, an accomplishment long awaited by Mesoamericanist scholars. The study has been carried out within the framework of a wider research project using an integrated array of spectroscopic instruments of the European mobile facility MOLAB, of the Center of Excellence Scientific Methodologies Applied to Archaeology and Art of the University of Perugia and the Institute of Molecular Science and Technologies. This study seeks to describe the chemical characterization of the painting materials used in the production of pictorial manuscripts in pre-Hispanic and colonial Mesoamerica (Miliani et al. 2010; Domenici et al. 2014; Brunetti et al. 2016). In November 2013, within this framework and thanks to the financial support of the European project CHARISMA, a MOLAB non-invasive campaign was performed on five Mesoamerican manuscripts held at the Bodleian Library in Oxford; these included the codices Laud (or Mictlan), Bodley (or Ñuu Tnoo-Ndisi Nuu), Selden (or Añute), Selden Roll, and Mendoza.1 Besides the work carried out in situ on the codices, the scientific project has been constantly supported by studies carried out in the laboratory on reference materials and pictorial models which were created based on the historical sources (Buti 2012; Grazia 2015). The research strategy applied to the Codex Mendoza has been based both on our own inspection of the manuscript at the Bodleian Libraries as well as on observations made by scholars that previously worked on it, who studied the painting technique, colors, and materials involved in making the Codex Mendoza.

Previous observations on the colors of the Codex Mendoza

The importance of considering the painting materials used by the painters of the Codex Mendoza was first stressed in 1938 by James Cooper Clark. Because he had no possibility of actually studying the materials of the manuscript, in his famous work, he simply listed the known native pigments used by ancient Mesoamerican painters as they had been recorded by Friar Bernardino de Sahagún in Book XI of his Florentine Codex (Clark 1938, 9). Two decades later, Donald Robertson published his detailed observations on the painting technique employed in the codex; he noted the coexistence of two different ways of applying colors: the flat washes, which he considered to be the traditional style of pre-colonial origin, and the use of one or more tones of a same color in order to create shades suggesting plasticity and tridimensionality, a trait that he considered to be a European-derived innovation. He also noted that these innovative traits are more abundant in the third part of the codex, the one with no pre-colonial models. In this section, they appear together with other novel elements such as the use of European perspective and the more realistic, “willowy” human figures engaging in a wide range of actions and facing right (that is, following the European reading order from left to right). Despite these differences, Robertson ([1959] 1994, 102–7) supposed that a single painter worked on the codex.

Building on Robertson’s observations, Kathleen Steward Howe (1992), who studied the codex as a member of the research group that produced the outstanding four volume study edited by Patricia Rieff Anawalt and Frances F. Berdan (1992), agreed on the presence of a single master painter. However, she suggested that he could have worked in a workshop, since variation in the technique of color application indicated that colors could have been applied by different collaborators (Howe 1992, 26). She noted that, even if the colors are mostly flat washes in Part 1, a brown color has been superimposed on an underlying yellow layer on the temples’ thatches (28). In Part 2, yellow and blue are always applied as flat washes, reds and yellows are blended to produce an “amber color”, and green is shaded to indicate modeling in warrior costumes and feathers; according to her interpretation, this green shading is a European trait. Howe also noted that the black frame lines were always painted before coloring, while the black elements within costumes were traced before the color if the costume is blue or orange, but after if the color is yellow or red. In her opinion, such details show that a group of people, working in a specific order, painted the manuscript. According to her observations, the painter who used the black color did not paint any other area and the painters who applied the colors were mostly acting according to pre-colonial modes; only the one who painted the green color seemed to be more influenced by European modes, maybe because he or she was younger in age. She observed that the same collective work was performed in Part 3, where the figures show a more evident volumetric intention. The men’s mantles, for example, show grey shades traced by the same painter who traced the frame lines of the drawing (presumably the principal artist) (28-29). In her words, “The use of color in the Codex Mendoza reflects both indigenous conventions and European conventions. The consistent use of a particular system of color application for individual pigments indicates a workshop system. The choice of the system of pigment application made by individuals within the workshop would seem to reflect the strength of their tie to the indigenous manuscript tradition rather than the newly introduced European system” (29). Nevertheless, “The amalgam of European and indigenous elements produced by the Mendoza artist speaks eloquently of both the speed of acculturation and resilience of the pictorial manuscript tradition” (25).

In several articles, Juan José Batalla Rosado (2007a; 2007b) proposed that the same “master painter” painted folios 6r-11v of the Matrícula de Tributos and the whole “tribute section” of the Codex Mendoza (folios 31r-42r). Batalla Rosado does not refer to colors in his analysis, as he mostly focused on drawing style. As such, his hypothesis of a single painter for the Codex Mendoza (or, at least, for the “tribute section”) only refers to the black line drawings and does not contradict the possibility of different painters applying different colors, as suggested earlier. Norman Hammond (2005) also provided interesting comments on the visual appearance of colors in Codex Mendoza and on their rendering in published facsimiles; he also suggested that the red color was an organic one, that is, cochineal or annatto.

More recently, Jorge Gómez Tejada (2012) made detailed observations on the material appearance of the colors of the Codex Mendoza, further stressing the explicit and intentional coexistence of indigenous pictorial modes and Europeanizing materials and techniques. In his opinion, the three parts of Codex Mendoza were drawn and painted as part of a single and coherent project by two identifiable artistic hands, whose work intermixes in the codex and even on the same page, suggesting that the two artists worked in the same workshop, as also suggested by the general coherence of the pool of painting materials across the three sections of the manuscript (Gómez Tejada 2012, 142–57; Chapter 5 of this volume). Although he agrees that some traits (such as the rendering of objects hit by lateral light) are European-derived, Gómez Tejada interestingly observed that the shading of color in order to render volumetric forms as well as the differential saturation of colors (shown in depictions of jade and water) do not necessarily indicate a European influence. He drew this conclusion because similar visual resources can be observed on pre-colonial artifacts such as Codex Borgia and in Maya mural paintings, even in examples dating back to the Preclassic period. In other words, according to Gómez Tejada, flat washes were not the only essential trait of pre-colonial scribal art.

Based on naked-eye examinations of the original manuscript at the Bodleian Libraries and personal communictions with Diana Magaloni Kerpel, Gómez Tejada proposed that the two painters used a constant and restricted palette composed by lamp black, a white paint he calls tízatl (often mixed with other colors), an organic yellow he identifies as zacatlaxcalli, cochineal red, a purple he identifies as camopalli, brown ochres, and Maya blue. He also observed that these colors were sometimes superimposed over one another and mixed. As for the alphabetic texts, he recognized the use of ferrogallic inks (sometimes also used to correct details of the images) and he supposed that the red color used to trace the alphabetic transcription of the year signs on page 1v is minium (that is, red lead). In sum, Gómez Tejada (2012, 136) stressed that the Codex Mendoza was painted as a single and coherent project by at least two artists who consciously mixed native and European materials, techniques, and modes of representations “as conscious and powerful rhetorical devices that are employed to meet the narrative priorities of the artists and authors of the work and not merely approached or emulated”.

Our own inspection of the manuscript confirmed some of the previous observations by Howe and Gómez Tejada, such as the quite homogeneous palette used on the whole codex, even with varying degrees of frequency of specific color use. All three parts, for example employ the superposition of grey over pink, even if this color is much more common in Part 3 than in Parts 1 and 2. Color intensity is often used to represent volume and tridimensionality (as in the green mountain toponyms), but also to provide a visual representation of iridescent colors (as on feathers, greenstones, and watery bodies), a key element in Nahua aesthetics (Russo 2011). As is first stressed by Gómez Tejada (2012, 108–9) in his observation about the depiction of feathers in Codex Borgia, this should not necessarily be seen as a colonial innovation. Nevertheless, it is also clear that not all colors were used this way in order to represent volume. For example, if most of the green mountain toponyms and cactuses of Part 1 show the use of differential color intensity, that same cannot be said for red (e.g. folio 8r) or grey (e.g. folio 16v) mountains in the same part. In any case, it is clear that red is sometimes used with a decreasing intensity in order to represent color shades, such as in the flames of burning temples (where it is superimposed on a yellow layer) or in the spondylus shells (e.g. folio 12r). Brown is also used in different tones to represent iridescence, such as in the feather of the eagles on folios 2r or 5v. In blue areas, the varying intensity is evident, but it is not always clear to which degree this was intentional (e.g. in the turquoise year bands). These variations in color use suggest that Howe could have been right in imagining different painters applying different colors; she was also probably correct in noting that the green color is the one that shows more innovative uses, such as the representation of objects lit by a lateral light.

The different colors seem to have had different degrees of liquidity: observing the reverse side of the painted pages, it is clear that the red color was the one that passed through the page more than any other. The other color that often passed through the page is yellow (one of the two yellows used in the manuscript, see below), as can be observed on the reverse of the yellow shields in Part 1. These different degrees of liquidity could have had some causal relationship with the differential painting order observed by Howe in the warriors’ costumes in Part 2, where inner black elements were traced after coloring only if the color was precisely red or yellow; perhaps, being more liquid, these colors would have smudged the underlying black areas.

The MOLAB campaign: methods, strategy, and limitations

Due to logistic limitations resulting from the need to perform analyses with various instruments on five different codices as well as to minimize the handling of the manuscripts, a few representative folios of Codex Mendoza were selected for study. The selection was guided by the goal to measure the different colors used in the manuscript in at least one folio per part in order to eventually detect changes in the palette within the different parts of the manuscript. The selected folios were 1v, 2r, 37r, 63r, and 64r. The strict selection of folios and the relatively low number of measured color areas means that our analyses cannot be expected to reveal any slight differences in color use or a precise mapping of the distribution of specific painting materials throughout the manuscript; achieving such objectives would require a much longer and more detailed analytical campaign. Nevertheless, the analyses allowed us to define the chemical characterization of the main colors of the palette used by the painters of the Codex Mendoza. When compared with the data recovered from other pre-colonial and colonial manuscripts studied in our project, these findings also allow us to evaluated the degree to which the Codex Mendoza’s painting technology was innovative as compared with traditional painting techniques of pre-colonial origin.

Such non-invasive analyses provide previously unthinkable data in the scientific study of rare and precious manuscripts that cannot be sampled for obvious reasons. However, it is also important to stress that such techniques also have some inherent limitations. For example, the employed techniques can reveal the chemical composition in a certain color area but not their absolute amount; moreover, when dealing with organic colorants, it is easier to characterize blue and red dyes than yellow and orange ones.

The following spectroscopic methods and imaging techniques were used in situ in a complementary fashion: X-ray fluorescence (XRF), reflection mid-FTIR, UV-Vis reflection and emission, micro-Raman, digital microscopy, and digital NIR imaging. Experimental details can be found in a previous publication (Domenici et al. 2017).

Results

White areas (paper)

The mid-FTIR spectra collected on the white ground show the signals related to the cellulose paper, as well as small and variable amounts of Ca, Fe, S, and K detected by XRF. The absence of any significant signal confirms what was readily detactable by naked-eye observation: most of the white areas of the images were simply left unpainted. Unfortunately, none of the rare areas where a white painted color was employed were measured. Nevertheless, the fact that calcium carbonate was detected as a component of a grey area (see below) strongly suggest that the white color of the Codex Mendoza painters’ palette was indeed calcium carbonate (called tizatl in colonial Nahuatl texts) (Dupey García 2016), as suggested by Gómez Tejada (2012).

Black and red inks in alphabetic writing

The black ink used for alphabetic writing (measured on all the selected pages) shows an elemental profile with iron as the main element and with variable amounts of zinc and manganese. This combination clearly indicates the use of an iron gall ink, a material of European origin often found in colonial Mesoamerican manuscripts. In some instances, gall ink was used to add small details to the pictorial imagery, such as on folio 63r, where it was detected on the tip of the spear of the young warrior on the left. Gall ink was also detected in the Spanish gloss of the year signs on folio 1v, previously thought to have been traced with lamp black.

The results obtained from the red color used in the Nahuatl gloss of the same calendric signs on page 1v were more unexpected, given that it was previously thought that it was composed of minium (Gómez Tejada 2012, 88). Elemental analysis combined with UV-Vis spectroscopy allowed for the straightforward identification of cinnabar (HgS). The use of cinnabar (also known as “vermillion”) as a colonial painting material is not unique since it was also used on the Codex Florentinus (Giorgi, Chelazzi, and Magaloni Kerpel 2014, 165) and the Beinecke Map (Magaloni Kerpel 2012, 83; Newman and Derrick 2012, 96); however, this is the first time that it has been proved that cinnabar was used as a writing material.

Black and grey

The black frame lines and the black areas within the images of the Codex Mendoza (measured on all the selected pages) are all made of a vegetable-derived carbon black (apart from the abovementioned small corrections with gall ink). Vegetal carbon black (known as “tlilli” or “tlilliocotl” in Nahuatl) is the only material that our analyses showed to be truly pan-Mesoamerican, since it was identified in all the manuscripts measured thus far (Domenici et al. 2014; Dupey García 2015, 152–53; 2016). Most grey areas in the Codex Mendoza were simply painted with a diluted form of carbon black. The only exception to the pattern so far described was detected in the smoke volutes on the lower left part of folio 63r, where lead was detected; this is the only point in the painted imagery where a non-traditional Mesoamerican material was detected. Since this lead grey was identified in a single point in the whole manuscript, it is difficult to say if it was a common color in the painters’ palette or if it represents a unique occurrence.

In various instances, carbon black was superimposed over a pink color, especially to paint the pinkish-grey flesh of people: our measurement on such pink-grey areas (folio 63r: body of upper left figure; folio 64r: body of the porter, or tlamemeh, in the upper register, and the leg of the lower right warrior) showed the presence of cochineal, which was consistent with other pink areas in the manuscript (see below). Calcium carbonate or tizatl (not detected in purely grey areas), which was probably used to lighten the color, was also detected. The superimposition of a grey layer over the pink color of the skin of some individuals is interesting, since a similar practice is observed in other Mesoamerican paintings, such as the victorious warriors of the Battle Mural at Cacaxtla (Brittenham and Magaloni Kerpel 2017, 67–70) or some of the individuals portrayed in the Chihk Nahb murals in Calakmul (Brittenham, personal communication; P. Baglioni and R. Giorgi, personal communication). This technique is probably intended to reproduce the black body paint applied over the skin, as also noted by Gómez Tejada (2012, 107).

Red and pink

An insect-derived anthraquinone lake pigment has been detected in all the red areas of the painted figures (folio 2r: lower left shield; folio 37r: right warrior’s cap and shield; folio 63r: teponaztli holder; and folio 64r: mantle of lower right warrior) by means of UV-Vis reflectance and emission spectroscopy. The same anthraquinone lake pigment was also used to paint pink areas, both in a lighter and a darker hue (folio 63r: ladle censer and folio 64r: border of the central mantle in the third register). Considering the historical period and the geographical origin of the codex, cochineal is the most likely insect source of this pigment. The use of cochineal as a codex painting material has been detected in most of analyzed codices (Domenici et al. 2017) and it is clearly reported in historical sources, in which it is called tlapalli. Tlacuahuac tlapalli could be the name of a cochineal lake, while terms such as tlapalnextli and camopalli seem to refer to specific chromatic variants of cochineal lakes (Dupey García 2009; 2015, 153; 2016).

Although frequently negligible, an increase of potassium related to the paper support was detected. This increase may possibly be due to the thinness of the painting layer and, tentatively, can be associated with the practice of making lake pigments by adding potash alum (AlK(SO4)2∙12H2O); these are well-documented in European pigment recipes (Kirby, Spring, and Higgitt 2005) and also reported in pre-Columbian historical references (Dupey García 2015, 153; 2016). As previously noted, this cochineal-based painting material seems to have been quite liquid, since it bleeds through the paper sheet more than any other color in the manuscript.

Yellows, oranges, and browns

Two different types of yellow compounds have been found in the Codex Mendoza. The first is arsenic-based, as indicated by XRF performed on some yellow areas (folio 2r: central shield and Tenoch’s icpalli and folio 37r: headdress of the left warrior, gourd, and mat box). The compound is most probably orpiment (As2S3), as identified by Raman spectroscopy in other codices (Buti et al. 2018; Domenici et al. 2017; Brunetti et al. 2016) The second yellow (foli 63r: astronomer’s seat and woman’s face in third register and folio 64r: costume of the central lower warrior) is an organic dye; we concluded it was organic in these areas because XRF analysis did not reveal any key element characteristic of inorganic pigments. Neither XRF nor FTIR provided indication of an inorganic support for the dye: XRF analyses showed a negligible amount of potassium (although this could be due to the low layer thickness) and FTIR showed the absence of any infrared signal of clay, although the hindering effect of the cellulosic support cannot be overlooked. These observations suggest that the organic color was not used in these areas as a lake pigment or as a hybrid pigment (Buti 2012; Domenici et al. 2017).

Unfortunately, it is not easy to distinguish orpiment from the yellow dye with simple naked-eye observation, since the color hues of the two painting materials seem to overlap (especially in the lighter yellow hues); these similarities are probably related to specific painting techniques such as different dilutions, etc. Nevertheless, the darker hue of the warrior’s costume in the lower register of folio 64r seems to be characteristic of the yellow dye. While the use of orpiment is not recorded in any colonial historical source, various organic yellows/oranges are mentioned with names that probably refer both to the organic dye and to the lake from which it was obtained. Zacatlaxcalli was a yellow obtained from Cuscuta sp. plants, while xochipalli was a more orange tint obtained from Cosmos sulphureus (Dupey García 2015, 254; 2016). Zacatlaxcalli is a likely source of the non-orpiment yellow of the Codex Mendoza, as previously suggested by Gómez Tejada.

In the brownish/orangeish areas on folio 37r (cuexcómatl) and folio 63r (teponaztli), XRF analysis and UV-Vis reflection indicated the presence of an iron-based pigment, possibly rich in goethite. XRF also showed a higher signal of arsenic in these same areas (and of manganese only on folio 63r), but it is not clear if this indicates that the ochre was mixed with orpiment or if arsenic naturally occurred within the ochre.

In some dark brown areas (folio 2r: eagle and darker hue of tetl sign; folio 63r: wooden stick in front of the hands of the beaten individual; and folio 64r: feathers of the lower right man), XRF spectra show signal of iron and (except on folio 64) manganese, an indication of the presence of manganese-based compounds (such as oxides and/or hydroxydes) which frequently occur along with iron oxides in natural earth pigments. The reflectance profile does not give any indication of their molecular composition.

Other brown areas, mostly with light hues also including the “fleshy” color used for human skin (folio 2r: skin of the upper right man and folio 64r: left mantle in the third register and leg skin of the central prisoner in the fourth register), were painted with organic dyes, as indirectly indicated by the low iron content and the increase of potassium. These characteristics are usually associated with the presence of a lake pigment, although these components do not represent conclusive proof of such a pigment. In some cases, as in the light hue of tetl on folio 2r, finding a low signal of iron does not allow us to conclusively determine if the paint in question was a dye or a diluted earth pigment, though the latter of these is more probable.

Sometimes, it seems that earth pigments and dyes were superimposed over one another: for example, the canoe on folio 63r, seems to have been painted with the light brown dye and then the dark shade on its bottom was probably added with an earth pigment.

Blue

All blue areas of the Codex Mendoza (measured on all selected pages) were painted with Maya blue, the well-known organic-inorganic hybrid pigment that was obtained by heating a mixture of indigo and palygorskite, whose presence has been detected by both FTIR and UV-Vis reflectance. The different degrees of darkness observable in the manuscript’s blue areas did not result in different instrumental readings. These variations in darkness must have been obtained by means of different dilutions and/or the repeated overlaying of the color, which nevertheless always shows a dense, covering aspect.

It seems that Maya blue was not explicitly mentioned in historical sources. However, it has recently been convincingly proposed that the term texotli could precisely refer to hybrid organic-inorganic blue pigments, including the one we call Maya blue (Dupey García 2010, 83–85; 2015, 155; 2016; Magaloni Kerpel 2012, 71).

Green

As expected on the basis of the two different yellows previously commented, two different greens were detected on the manuscript, both of them obtained by mixing or superimposing Maya blue and one of the two yellow colors used in the manuscript. In fact, in all green areas, mid-FTIR analysis detected palygorskite, the inorganic substrate of Maya blue and UV-Vis reflectance allowed the conclusive identification of the pigment. The dense, covering appearance of both the Codex Mendoza’s greens is thus clearly related to the use of Maya blue.

The yellow chromatic component of the first of the two greens (folio 2r: thatch of the upper temple and folio 37r: darker area of the headdress’s feathers) is orpiment. The second green contains the yellow dye used for some of the yellow areas. This green was detected on the cactus on folio 2r and in both shades of the green feathers of the central warrior, in the third register of folio 64r. On the reeds on folio 2r and on the lighter shade of the headdress feathers on folio 37r, As is present only in low amounts. We cannot state with certainty which of the two greens was used, since the low amount of As could be due to the thin layer of highly diluted coloring material.

Green colors mentioned in historical sources are quiltic and yappalli, both of them obtained by mixing zacatlaxcalli and texotli, that is a yellow organic dye and Maya Blue, with the latter being a darker, even brownish one (Dupey García 2015, 156; 2016). Quiltic is then the most likely candidate as the Nahuatl name of one of the green colors detected on Codex Mendoza.

Discussion

In general terms, the pictorial imagery of the Codex Mendoza shows that a quite homogenous palette composed mostly of traditional painting materials of obvious pre-colonial origin was used. This is demonstrated by the use of carbon black (both for frame lines and painted areas), calcium carbonate, cochineal lake (for red and pink painted areas), Maya blue, and yellow/brown dyes. We would need a much higher number of measured points to map in detail the actual distribution of these coloring materials within the manuscript, but our results do not show clear evidence of changes in the palette. For example, although a yellow dye to paint yellow areas was only detected in Part 3, the same yellow dye was also detected as a chromatic component of green colors in both Parts 1 and 2.

All of the above-mentioned materials have been detected in the Borgia Group codices, the most comparatively relevant group of pre-Hispanic manuscripts, since we unfortunately lack any secure pre-colonial codex from the Basin of Mexico (Miliani et al. 2012; Domenici et al. 2014; 2018; 2017; Buti et al. 2018). A scientific analysis of the colors used in the Matrícula de Tributos, whose dating is debated, would obviously be of great comparative interest in order to ascertain its degree of material affinity with the Codex Mendoza. It is important to note that these same materials were also used to paint the Codex Borbonicus, the early colonial manuscript from the Basin of Mexico whose palette does not show any clear evidence of innovative colonial technological traits (Pottier et al. 2019).

Apart from these traditional materials, the Codex Mendoza also contains some colors that deserve a more detailed discussion, since their status as “traditional” or “innovative” is controversial. The first of these colors is arsenic trisulphide (orpiment), which was detected only in Parts 1 and 2 and also as a component of green colors. Given the limited number of measured points, we would be extremely cautious in excluding the possibility that it was also used in Part 3. We have previously detected this yellow inorganic color on codices Laud, Fejérváry-Mayer, Cospi verso, and Nuttall recto, thus undoubtedly demonstrating its usage in pre-colonial codex painting (Miliani et al. 2012; Domenici et al. 2014; 2018; 2017; Buti et al. 2018). Nevertheless, its usage seems to have been fairly restricted, both spatially and chronologically. On one hand, codices Laud, Fejérváry-Mayer, and Cospi verso (which share some very specific and distinctive stylistic and thematic traits) were arguably painted in the same region, probably the Tehuacan area (Álvarez Icaza Longoria 2014). Although the Codex Nuttall was arguably painted in the Tilantongo area of the Mixtec region, it shows both technological and stylistic similarities with the abovementioned codices, perhaps reflecting pre-colonial interaction across the current Oaxaca-Puebla border. On the other hand, the fact that orpiment appears only on the later-painted sides of both codices Cospi and Nuttall suggests that its introduction onto the painters’ palette was a relatively late phenomenon in pre-Hispanic times. Due to the lack of pre-colonial manuscripts from the Basin of Mexico, we do not know if orpiment was used by local painters before the conquest. However, the absence of orpiment on codices Borgia, Cospi recto, and Vaticanus B as well as on the early colonial Codex Borbonicus seems to suggest that it may not have been used before this period. If this hypothesis is correct, the usage of orpiment on Codex Mendoza should be interpreted as a colonial innovation and perhaps as an early manifestation of a process of technological transformation which led to the quite common usage of orpiment in post-1550 colonial manuscripts from the Basin of Mexico, such as Codex De la Cruz-Badianus, the Beinecke Map, and Codex Florentinus (Newman and Derrick 2012; Giorgi, Chelazzi, and Magaloni Kerpel 2014; Zetina et al. 2011).

We discovered a similar problem related to the usage of orangeish/brown ochers and iron-based pigments. Our analyses did not detect any usage of ochre in Central Mexican or pre-colonial Mixtec manuscripts. On the other hand, we know that ochers such as goethite were commonly used in pre-colonial Mexica mural painting (López Luján and Chiari 2012, 333). Thus, according to presently available data, the introduction of ochers onto the codex painters’ palette seems to be a Central Mexican colonial innovation. This adition broke those cultural norms that made the pre-colonial color palette of codex painters clearly distinguishable from that used in mural painting, which was almost completely devoid of any purely inorganic pigment. The growth of this trend is shown by the common usage of ochers and other inorganic pigments in manuscripts dated to the second half of the sixteenth century, such as Codex De La Cruz Badianus, Codex Florentinus, and the Beinecke Map as well as in various maps found in Relaciones geográficas (Haude 1998; Newman and Derrick 2012; Giorgi, Chelazzi, and Magaloni Kerpel 2014; Domenici et al. 2018; Zetina et al. 2011).

A specific goethite-based brown/orangeish pigment detected on the Codex Mendoza deserves further explanation. As mentioned above, this pigment contains arsenic, but it is difficult to say if this last element indicates the mixture of ochre with orpiment or if it was contained in the natural earth. Nevertheless, it is extremely interesting to note that a very similar, if not identical, material has been detected on the Beinecke Map, where it was interpreted by Magaloni Kerpel (2012, 87) as evidence of the use of orpiment. On the other hand, Newman and Derrick (2012, 97) apparently favored the idea that arsenic naturally occurs within the goethite.

The last innovative element is cinnabar. Despite being widely used in pre-Hispanic Mesoamerica for various purposes, according to our current data, this material was not used as a painting material on pre-colonial codices. On the other hand, cinnabar has been detected on colonial manuscripts such as the Beinecke Map, Codex Florentinus, and the Meztitlan Map (Haude 1998; Newman and Derrick 2012; Giorgi, Chelazzi, and Magaloni Kerpel 2014; Domenici, Miliani, and Sgamellotti 2019; Domenici et al. 2018). Interestingly enough, in these post-1560 manuscripts cinnabar was used as a painting material, while in the Codex Mendoza its usage was restricted to a written gloss. This limited usage suggests that it was a fairly recent introduction, still not accepted as a component of a proper pictorial palette, and confined to the realm of alphabetic writing shared with the European-introduced gall inks.

When paired with cultural and art historical analysis, the diverse implications of these elements can contribute to our understanding of how the Codex Mendoza was manufactured. First of all, the almost complete homogeneity of the palette in the three parts of the codex and the recurring variations in the patterns of color application strongly suggests that it was painted in a single workshop, probably by two or more painters following a working procedure of pre-colonial origin. These artists primarily painted in a traditional style using a mostly traditional palette, but also introduced a restricted set of new stylistic and technological traits.

It has been proposed that Mesoamerican painters may have selected their painting materials based on the symbolic attributes of the specific items they were depicting (e.g., Magaloni Kerpel 2011; 2012; 2014; Giorgi, Chelazzi, and Magaloni Kerpel 2014). Such an assertion does not seem to be fully confirmed by our analyses of pre-colonial manuscripts, given that a single painting material is often used all over a whole codex (especially in the case of black, red, and blue items) and variations in this pattern seem to depend more on chromatic rather than on material attributes of paints (as is the case of the yellow/orange/brown and green items). A similar pattern seems to characterize the Codex Mendoza; no clear usage of specific painting materials seems to be linked to the symbolic qualities of the depicted items and different depictions of the same item (e.g. a yellow shield) can be painted either with an organic or an inorganic color. Again, yellow and brownish colors show a higher compositional variance, while the same black, red, and blue painting materials are used all over the manuscript. Similarly, the purported symbolic pairing of local lamp black and non-local minium in two parallel glosses on folio 1v (Gómez Tejada 2012, 88–89) has been dismissed by our analyses, which showed that they were actually painted with gall ink and cinnabar. It is debatable whether cinnabar is traditional/non-traditional or local/non-local.

This said, we wish to clarify that we are not at all denying the fact that painting colors were imbued with deep symbolic significance derived from their materiality. But, in our opinion, this significance seems to be more generally linked to the brilliance of mostly organic painting materials used on pre-colonial codices. These materials were perceived as an aesthetic attribute expressing the “bursting” and “flowery” essence of manuscripts, which were intended to be chanted rather than read; thus, the manuscript pages were actual materializations of Mesoamerican “flowery speech” (Domenici 2016). This pattern seems to have been progressively broken during the early colonial times, when a host of local and foreign inorganic and more opaque pigments appeared on the codex painters’ palette, probably as a reflection of the changed context of production and performance of pictorial manuscripts increasingly manufactured and perceived as “books” in the European sense of the term (Domenici et al. 2018). With its still limited use of inorganic pigments and European-derived materials, the Codex Mendoza seems to stand at the very beginning of this process of early colonial material experimentation.

The rather limited set of colonial technological innovations identified on the Codex Mendoza leads us to the problem of the manuscript’s chronology, a topic that is also tackled on different basis and with different methodological approaches in other contributions in this book. As we saw, the Codex Mendoza palette seems to be more “innovative” than the one of the Codex Borbonicus, which was probably painted during the1530s; this assumption is based on the traditional dating of the Codex Mendoza to 1541-1542 (Robertson [1959] 1994, 87–90). On the other hand, the Codex Mendoza looks more technologically conservative than Codex De la Cruz Badianus, Beinecke Map, and Codex Florentinus. Despite some differences, the strong similarity between the palettes of Codex Mendoza and the Beinecke Map (arguably painted around 1565) is still noteworthy (Miller 2012, 3). Even if technological traits cannot be automatically taken as chronological indicators, our results suggest that the Codex Mendoza (more innovative than Codex Borbonicus and somewhat less innovative than the Beinecke Map) could have been painted somewhere between the 1530s and the 1560s. Within this rather ample time span, either the traditional dating of the manuscript to 1541-42 and the later date range of 1547-1552 convincingly proposed by Gómez Tejada (2012; see this volume too) would fit. In the absence of direct scientific dating of the actual manuscript, further specification cannot be achieved by means of purely technological comparisons between often loosely dated manuscripts. Thus, the task of more precisely dating Codex Mendoza should be more properly tackled by historical and art historical research.

Despite the inherent limitations of the analytical methods and strategies adopted in our analyses and of comparisons carried out among a sadly restricted corpus of extant manuscripts, these non-invasive scientific analyses allowed the first chemical characterization of Codex Mendoza’s palette. This is a wonderful snapshot of the early phases of the technological colonial encounter that took place in mid-sixteenth century Mexico.

Acknowledgements

The non-invasive MOLAB campaign on the five Mesoamerican manuscripts held at the Bodleian Library in Oxford was carried out thanks to the financial support of the European Project CHARISMA. Bruce Barker-Benfield (Bodleian Libraries), Marinita Stiglitz (Bodleian Libraries), and Noemi Mancinelli (SMAArt, Perugia) are gratefully acknowledged for their help during the in situ campaign. Our deepest thanks to Jorge Gómez Tejada who, besides inviting us to contribute to the present volume, thoroughly discussed different issues with us and carefully edited our paper. Claudia Brittenham read and discussed a draft of this text, providing insightful comments; Piero Baglioni and Rodorico Giorgi kindly answered our questions on the Calakmul murals.

References

Álvarez Icaza Longoria, María Isabel. 2014. “El Códice Laud, Su Tradición, Su Escuela, Sus Artistas.” Tesis doctoral, Universidad Nacional Autónoma de México.

Batalla Rosado, Juan José. 2007a. “Matrícula de tributos y Códice Mendoza: La autoría de un mismo ‘maestro de pintores’ para los folios 6-R a 11-V del primero y la totalidad del segundo.” Anales del Museo de América, no. 15: 9–20.

———. 2007b. “The Scribes Who Painted the Matrícula de Tributos and the Codex Mendoza.” Ancient Mesoamerica 18 (1): 31–51. https://doi.org/10.1017/S0956536107000077.

Berdan, Frances. 1992. “The Imperial Tribute Roll of the Codex Mendoza.” In The Codex Mendoza, edited by Frances Berdan and Patricia Rieff Anawalt, 2:55–79. Berkeley: University of California Press.

Brittenham, Claudia, and Diana Magaloni Kerpel. 2017. “The Eloquence of Color. Material and Meaning in the Cacaxtla Murals.” In Making Value, Making Meaning. Techné in the Pre-Columbian World, edited by Cathy Lynn Costin, 63–94. Washington D.C.: Dumbarton Oaks.

Brunetti, Brunetto Giovanni, Costanza Miliani, Francesca Rosi, B. Doherty, L. Monico, Aldo Romani, and Antonio Sgamellotti. 2016. “Non-Invasive Investigations of Paintings by Portable Instrumentation: The MOLAB Experience.” Topics in Current Chemistry 374 (1): 10. https://doi.org/10.1007/s41061-015-0008-9.

Buti, David. 2012. “Multi-Technique Approach for the in Situ Study of Ancient Manuscripts.” Tesis doctoral, Università degli Studi di Firenze.

Buti, David, Davide Domenici, Chiara Grazia, J. Ostapkowicz, S. Watts, Aldo Romani, F. Presciutti, Brunetto Giovanni Brunetti, Antonio Sgamellotti, and Costanza Miliani. 2018. “Further Insight into Mesoamerican Paint Technology: Unveiling the Colour Palette of the Pre-Columbian Codex Fejérváry-Mayer by Means of Non-Invasive Analysis.” Archaeometry 60 (4): 797–814. https://doi.org/10.1111/arcm.12341.

Clark, James Cooper. 1938. Codex Mendoza: The Mexican Manuscript Known as the “Collection of Mendoza” and Preserved in the Bodleian Library. Oxford, UK: Waterlow and Sons.

Domenici, Davide. 2016. “La memoria fiorita. Scrittura, memoria e materialita’ del colore nell’antica Mesoamerica.” Confluenze: Rivista di Studi Iberoamericani 8 (2): 161–80.

Domenici, Davide, David Buti, Costanza Miliani, Brunetto Giovanni Brunetti, Antonio Sgamellotti, and Bruno Brunetti. 2014. “The Colours of Indigenous Memory: Non-Invasive Analyses of Pre-Hispanic Mesoamerican Codices.” In Science and Art: The Painted Surface, edited by Antonio Sgamellotti, Brunetto Giovanni Brunetti, and Costanza Miliani, 94–119. Cambridge: Royal Society of Chemistry.

Domenici, Davide, David Buti, Costanza Miliani, and Antonio Sgamellotti. 2018. “Changing Colours in a Changing World. The Technology of Codex Painting in Early Colonial Mexico.” In Materia Americana. The ‘Body’ of Spanish American Images (16th to Mid-19th Centuries), edited by Gabriela Siracusano and Agustina Rodríguez Romero. Los Angeles: The Getty Foundation.

Domenici, Davide, Chiara Grazia, David Buti, Aldo Romani, Costanza Miliani, and Antonio Sgamellotti. 2017. “La cochinilla en la pintura de códices prehispánicos y coloniales.” In Rojo mexicano, edited by Georges Roque. Ciudad de México: CONACULTA.

Domenici, Davide, Costanza Miliani, and Antonio Sgamellotti. 2019. “Cultural and Historical Implications of Non-Destructive Analyses on Mesoamerican Codices in the Bodleian Libraries.” In Mesoamerican Manuscripts: New Scientific Approaches and Interpretations, edited by Maarten Jansen, Virginia M. Llado-Buisán, and Ludo Snijders. Brill: Leiden.

Dupey García, Élodie. en imprenta. “Aztec Reds. Investigating the Materiality of Color and Meaning in a Pre-Columbian Society.” In Essays in Global Color History: Interpreting the Ancient Spectrum, edited by Rachael Goldman. Piscataway: Gorgias Press.

———. 2009. “Du rouge de la cochenille à la transmission des savoirs traditionnels: Les multiples significations de tlapalli chez les anciens Nahua (Mexique).” In L’Antiquité en couleurs, edited by Marcelo Carastro, 207–22. Grenoble: Éditions Horos.

———. 2010. “Les couleurs dans les pratiques et les représentations des Nahuas du Mexique central (xive -xvie siècles).” Tesis doctoral, École Pratique des Hautes Études.

———. 2015. “El color en los códices prehispánicos del México Central: Identificación material, cualidad plástica y valor estético.” Revista Española de Antropología Americana 45 (1): 149–66.

———. 2016. “Los colores del Códice borbónico y el átlas en cromotipia de Ernest-Theodore Hamy.” Estudios de Cultura Náhuatl 52: 225+.

Giorgi, R., D. Chelazzi, and Diana Magaloni Kerpel. 2014. “Unveiling the Artistic Technique of the Florentine Codex: When the Old World and the New World Met.” In Science and Art: The Painted Surface, edited by Antonio Sgamellotti, Brunetto Giovanni Brunetti, and Costanza Miliani, 147–69. Cambridge: Royal Society of Chemistry.

Gómez Tejada, Jorge. 2012. “Making the Codex Mendoza, Constructing the Codex Mendoza: A Reconsideration of a 16th Century Mexican Manuscript.” Tesis doctoral, Yale University.

Grazia, Chiara. 2015. “Through the Shades of Time, from Pre-Columbian to Contemporary Art: Application of UV-Vis-NIR Reflectance and Fluorescence Spectroscopy to Understand Painting Materials.” Tesis doctoral, Università degli Studi di Perugia.

Hammond, Norman. 2005. “Clarifying Color in the Codex Mendoza.” Latin American Antiquity 16 (4): 467–68. https://doi.org/10.2307/30042510.

Haude, Mary Elizabeth. 1998. “Identification of Colorants on Maps from the Early Colonial Period of New Spain (Mexico).” Journal of the American Institute for Conservation 37 (3): 240–70. https://doi.org/10.1179/019713698806082822.

Howe, Kathleen Stewart. 1992. “The Relationship of Indigenous and European Styles in the Codex Mendoza.” In The Codex Mendoza, edited by Patricia Rieff Anawalt and Frances Berdan, 1:25–33. Berkeley, CA: University of California Press.

Kirby, Jo, Marika Spring, and Catherine Higgitt. 2005. “The Technology of Red Lake Pigment Manufacture: Study of the Dyestuff Substrate.” National Gallery Technical Bulletin 26: 71–87.

López Luján, Leonardo, and Giacomo Chiari. 2012. “Color in Monumental Mexica Sculpture.” RES: Anthropology and Aesthetics 61–62 (March): 330–42. https://doi.org/10.1086/RESvn1ms23647839.

Magaloni Kerpel, Diana. 2011. “Painters of the New World: The Process of Making the Florentine Codex.” In Colors between Two Worlds: The Florentine Codex of Bernardino de Sahagún, edited by Gerhard Wolf and Joseph Connors, 47–76. Florence: Kunsthistorisches Institut in Florenz Max-Plank-Institut - Villa I Tatti The Harvard University Center for Italian Renaissance Studies.

———. 2012. “The Traces of the Creative Process: Pictorial Materials and Techniques in the Beinecke Map.” In Painting a Map of Sixteenth-Century Mexico City: Land, Writing, and Native Rule, edited by Mary Ellen Miller and Barbara E. Mundy, 75–90. New Haven: Yale University Press.

———. 2014. Los colores del nuevo mundo: artistas, materiales y la creación del Códice florentino. Ciudad de México: Universidad Nacional Autónoma de México, Instituto de Investigaciones Estéticas.

Miliani, Costanza, Davide Domenici, Catia Clementi, F. Presciutti, Francesca Rosi, David Buti, Aldo Romani, L. Laurencich Minelli, and Antonio Sgamellotti. 2012. “Colouring Materials of Pre-Columbian Codices: Non-Invasive in Situ Spectroscopic Analysis of the Codex Cospi.” Journal of Archaeological Science 39 (3): 672–79. https://doi.org/10.1016/j.jas.2011.10.031.

Miliani, Costanza, Francesca Rosi, Brunetto Giovanni Brunetti, and Antonio Sgamellotti. 2010. “In Situ Noninvasive Study of Artworks: The MOLAB Multitechnique Approach.” Accounts of Chemical Research 43 (6): 728–38. https://doi.org/10.1021/ar100010t.

Miller, Mary Ellen. 2012. “Introduction: The Beinecke Map.” In Painting a Map of Sixteenth-Century Mexico City: Land, Writing, and Native Rule, edited by Mary Ellen Miller and Barbara E. Mundy, 1–8. New Haven: Yale University Press.

Newman, R., and M. Derrick. 2012. “Analytical Report of the Pigments and Binding Materials Used on the Beinecke Map.” In Painting a Map of Sixteenth-Century Mexico City: Land, Writing, and Native Rule, edited by Mary Ellen Miller and Barbara E. Mundy, 91–100. New Haven: Yale University Press.

Pottier, Fabien, Anne Michelin, Aurélie Tournié, Fabrice Goubard, Aymeric Histace, and Bertrand Lavédrine. 2019. “Preliminary Investigation on the Codex Borbonicus: Macroscopic Examination and Coloring Materials Characterization.” In Painting the Skin : Pigments on Bodies and Codices in Pre-Columbian Mesoamerica, edited by Élodie Dupey García and María Luisa Vázquez de Ágredos Pascual, 157–74. Tucson, AZ: The University of Arizona Press.

Robertson, Donald. (1959) 1994. Mexican Manuscript Painting of the Early Colonial Period: The Metropolitan Schools. New Haven: Yale University Press.

Russo, A. 2011. “Postface. Uncatchable Colors.” In Colors between Two Worlds: The “Florentine Codex” of Bernardino de Sahagún, edited by Gerhard Wolf and Joseph Connors, 389–410. Florence: Kunsthistorisches Institut in Florenz Max-Plank-Institut - Villa I Tatti The Harvard University Center for Italian Renaissance Studies.

Zetina, S., J. L. Ruvalcaba, M. Lopez Cáceres, T. Falcón, E. Hernández, C. González, and E. Arroyo. 2011. “Non Destructive In Situ Study of Mexican Codices: Methodology and First Results of Materials Analysis for the Colombino and Azoyu Codices.” In Proceedings of the 37th International Symposium on Archaeometry, 13th - 16th May 2008, Siena, Italy, edited by Isabella Turbanti-Memmi, 349–54. Berlin, Heidelberg: Springer. https://doi.org/10.1007/978-3-642-14678-7_50.

1 The analytical data regarding the five codices analyzed in the Bodleian Libraries were previously published in Domenici (2019; 2017). We provide herein a much more detailed discussion of the data regarding the Codex Mendoza.