8.

THE BODY OF MEANING

IN CHAVÍN ART

Gary Urton

I n the study of pre-Columbian art in South

America, one particular artistic tradition, that
known (from the eponymous site of Chavín de
some of the standard formal elements and organi-
zational principles of Chavín art. The purpose of
this exercise is to develop a basis for beginning to
Huántar, Peru) as Chavín, has been the focus of a discuss the relationships between form and mean-
great deal of interest and attention over the ing that may have been important to Chavín artists
years.1 The interest in Chavín art is related par- as they went about their work, rendering subjects
tially to the antiquity of works produced in this in the style to which they had become accustomed.
style (ca. 850–200 B.C.), as well as to its impres-
sive range of distribution. Objects of stone, pot- A N E X P L O R AT I O N O F S T Y L E
tery, gold, shell, and other media rendered in the
I N C H AV Í N A RT
Chavín style have been found in archaeological
sites along the coast, in the Andean highlands, and To begin with, and following Boas’s dictum (1955:
at sites along tributaries of the Amazon River 9–10) to the effect that “without skill, there is no
within much of the territory of the present-day art,” I maintain that underlying all fascination
nation of Peru.2 Beyond its impressive distribu- with Chavín art is the perception that a consider-
tion in time and space, the central question that able degree of artistic skill is represented in the
has motivated so much research on this artistic composition, design, and execution of most works
tradition is, What is the meaning of Chavín rendered in this style. Therefore, at the most basic
iconography? This is the question that motivates level—that of execution—we are constantly reas-
the present study. sured when we view a range of works in this style
In more concrete terms, I address the follow- that the individual Chavín artists had mastered
ing questions: What attracts us when we view a their crafts.
work of art in the Chavín style? What were Chavín At the next level, I suggest that when we view
artists communicating about their understanding particularly complex examples of Chavín art
of their world in the iconography of this art? And, work, such as the Lintel of the Jaguars (Roe 1974:
what (if anything) do we know about the subject Fig. 9) or the Yauya stela (plate 8.1; see Roe 1974:
matter of Chavín iconography as a basis for un- Fig. 11), we are often simultaneously repelled and
derstanding and interpreting its meaning? To ad- attracted by the absence of a clearly identifiable
dress these questions, I begin with a rather fanci- focal subject. By this I mean that the eye imme-
ful characterization of the process whereby (at diately encounters a profusion of complex, inter-
least in my own experience) one gains a familiar- locking forms with no easily discernible central
ity with, and gradually becomes accustomed to, image and with no clearly defined figure/ground

215
216 C H AV Í N : A RT, A R C H I T E C T U R E , AND C U LT U R E

relation among the various parts of the clusters of portance of animals in metaphorical con-
images. It is as though one is viewing a very com- structions more broadly —“[t]his becoming
plex, highly stylized jigsaw puzzle, but as to an object, this taking the other, this prediction
whether or not the pieces are all in their proper upon the pronoun, is a process that has for
places, and if so what the image in the puzzle is millennia turned to the animal world”—the
supposed to be “about,” one cannot say with cer- Chavín artists achieved an even greater level
tainty. However, as one begins to sort out and re- of affective power of such representations
group individual elements in a work such as the through their emphasis on wild carnivores.
Yauya stela (figure 8.1) or the Tello obelisk (fig-
ure 8.2), the abundance of often grim-looking,
profusely tusked animals, such as felines, reptiles,
and raptorial birds, gradually resolve themselves
into a central image.
To summarize, and assuming skill as the fun-
damental requirement for the production of any
work of art, I argue that on the formal level, we
are often fascinated by works of art in the Chavín
style because:
‰ We encounter an immediate challenge and
dilemma in identifying the relationship be-
tween what appear to be innumerable parts
with a recognizable whole. That is, Chavín art
seems to tinker, in a way often uncomfortable
for us, with our part-whole system of classifi-
cation according to which we manage, on an
everyday basis, to recognize patterns among
disparate forms.
‰ The subject matter of Chavín iconography is
overwhelmingly concerned with life forms
that have held a fascination for humans in all
times and places—that is, animals. Perhaps
only dimly, through our remnant pre-indus-
trial senses, do we recognize that these ani-
mals are for the most part carnivores. The
Chavín artists consistently provide an index
of the dietary predilection of their subjects by
equipping them with oversized canines, claws,
and occasionally, fierce-looking agnathic
mouths. Miller and Burger (1995:453–454)
have noted that while the principal animals
consumed by residents of Chavín de Huantar
included llamas and (to a lesser extent) deer
and vicuña, the animals represented in the
iconography were in all cases the wild, carniv-
orous animals of the tropical forest. Thus,
while the subject matter of Chavín iconog-
raphy generally conforms to what Fernandez Figure 8.1. The Yauya stela (from
(1974:122) has noted with regard to the im- Roe 1974:47, Fig. 11)
 C H A P T E R 8: T H E B O DY OF MEANING IN C H AV Í N A RT 217

‰ And finally, despite the initially unfamiliar what the particular work of art is “about.” I
patterns of arrangement of the various ele- would argue that much of the affective power
ments of Chavín composite figures, there is of Chavín art is its focus on the body—the ob-
ultimately always the body of an animal and/or ject in the world with which we are most fa-
human that we can identify as constituting miliar and through which we experience the
world. But, what does this art do with, or say
about, the body? It is the Chavín artists’ ma-
nipulations of the body—the transformation
of joints and penises into fierce jaguars; of
hair into entwined serpents; and of vaginas
into the sharp-toothed mouths of piranhas—
that, as possessors of bodies, commands our
attention.
Following from the above observations, I will
argue herein that the body was a subject of great
interest to Chavín artists. Moreover, I argue that
one of the primary vehicles for representing
meaning in Chavín art is the play of transforma-
tions and substitutions of body parts and that the
bodies of Chavín composite creatures represent
structural “frameworks,” models, or paradigms for
organizing and classifying relations (for example,
of ancestry, filiation, and affinity) among a host of
phenomena and domains of meaning and experi-
ence in nature and culture. By analyzing these re-
lationships between bodily form and meaning in
art, we can hope to arrive at a point from which
to articulate some of the core classificatory prin-
ciples, values, and meanings in Chavín society and
culture more broadly.

THE PROBLEM OF “K E N N I N G S ”

Any new attempt to develop an understanding of

form and meaning in Chavín art requires a con-
sideration of the dominant paradigm in the study
of Chavín art today. I am referring to the ideas ar-
ticulated in John H. Rowe’s classic study, “Form
and Meaning in Chavín Art” (1967a). One of the
most influential suggestions made by Rowe in this
article was that Chavín iconography should be un-
derstood in terms of the construction of various
levels and forms of visual metaphors, beginning in
simple similes and metaphors and culminating in
the representation of metaphors of metaphors;
Rowe refers to the latter as “kennings.” Rowe de-
Figure 8.2. The Tello obelisk. Museo National, rived the idea of kennings from a common liter-
Pueblo Libre, Lima. Photo: Gary Urton ary convention, or trope, used in Norse sagas. It
218 C H AV Í N : A RT, A R C H I T E C T U R E , AND C U LT U R E

will be instructive to begin with a rather lengthy cause the [Chavín] literary tradition is lost.
quotation from Rowe’s study in which he outlined (Rowe 1967a/1977:313– 314; my emphases)
the main features of kennings and suggested how
such a device might have been employed in the Now, while Rowe’s introduction of the con-
production of Chavín iconography. cept of “kenning” has proven to be an important
stimulus for thinking and writing about Chavín
The type of figurative elaboration which is
iconography over the years (certainly it was the
characteristic of Chavín art is one with which
point of departure for the present work), there are
we are more familiar in literary contexts; it is a
a number of problems with the use of this trope
series of visual comparisons often suggested
in interpreting Chavín (or any other) art style.
by substitution. To give a literary example, if
The principal problem is, in fact, signaled by
we say of a woman that “her hair is like
Rowe’s continual switching between artistic and
snakes,” we are making a direct comparison
literary examples in explaining the meaning of
(simile). If we speak of “her snaky hair” we
kennings and other related tropes (such as simile
are making an implied comparison (meta-
and metaphor) in the above quotation, all the
phor). We can go even further, however, and
while claiming that the literary examples are valid
simply refer to “her nest of snakes,” without
for iconographic representation. However, even a
using the word hair at all, and in this case we
cursory look at the comparison between literature
are making a comparison by substitution. In
and iconography will convince us that something
order to understand our expression the
is amiss here. Literary or verbal statements that
hearer or reader must either share with us
make use of the devices of simile (“her hair is like
the knowledge that hair is commonly com-
snakes”), metaphor (“her snaky hair”), and ken-
pared to snakes or infer our meaning from
ning (“her nest of snakes”) produce sequential, or
the context. Comparison by substitution
layered, images in the mind. That is, saying “her
was an especially fashionable device in Old
hair was like twisted snakes” evokes, first, an image
Norse court poetry, and it was given the
of hair, and then, the transformation of “hair” into
name “kenning” by the thirteenth century
“snakes,” However, when represented visually,
scholar Snorri Sturluson (1178– 1241).
these three distinct linguistic tropes will all look ex-
In Old Norse court poetry kenning be- actly alike—that is, hair rendered as snakes. The
came the chief basis on which verse was reason for this, of course, is because a visual rep-
judged. The poets responded to this devel- resentation of, for instance, a metaphor cannot
opment in taste by devising ever more portray the simile of which that metaphor is a fur-
complex and far fetched kennings as well ther elaboration. In short, a visual representation
as increasing the frequency with which of any one of the three verbal statements can be
they used these figures. The elaboration of accomplished only by means of “comparison by
kennings was of two kinds, the kenning of substitution” (that is, Rowe’s “kenning”). Thus,
kennings and the introduction of kennings the distinctions Rowe makes among the various
which depended on a reference to a story literary devices, or tropes, break down when ap-
which the hearers were assumed to know. plied to the domain of visual art.
By appealing to the literary trope of “ken-
The same kind of development in the direc- nings,” Rowe was trying to develop a method-
tion of increasing figurative complexity ological approach whereby the symbols and
which we have described for Old Norse po- metaphors suggested by “substitutions” and
etry took place also in Chavín art. Kennings “comparisons” in Chavín iconography could be
became more numerous and more far used as a basis for the study of meaning in Chavín
fetched, and we can identify cases of the society and culture more generally. The most dar-
kenning of kennings. We cannot identify ken- ing suggestion made by Rowe for the potential
nings referring to stories in any specific way, be- value of this approach in his 1967(a) article was
 C H A P T E R 8: T H E B O DY OF MEANING IN C H AV Í N A RT 219

that the symbols and metaphors derived by means vín artists. While I would not suggest that we in-
of the analysis of artistic “kennings” might reveal terpret our bodily experiences in the same ways
elements that would have been explained by, and the people of Chavín did, nonetheless I maintain
therefore were representations of, Chavín myths that the structures, processes, and experiences of
and legends. However, in order to make use of the body are the most logical points of departure
kennings in art for interpreting a mythical state- for an informed analysis of form and meaning in
ment, we would have to be able to refer to the vi- Chavín art.
sual ancestry—that is, the prior chain of compar-
isons and substitutions—from which the final
image was derived. Only if we were to have access T H E PA I R E D A M A R U S O F THE
to Chavín narratives, which would allow us to TELLO OBELISK
move beyond, or behind, iconographic compar-
isons and substitutions could we follow the cumu- In order to develop the ideas outlined above, I ex-
lative chains of ever more complex and indirect amine closely a single, but quite complex piece of
comparisons that are the hallmark of “kennings” Chavín art, the so-called Tello obelisk (plate 8.1;
in literature. Thus, the argument becomes circu- figure 8.2). The Tello obelisk, which is currently
lar and falters once again on the same point noted housed in the Museo Nacional in Pueblo Libre,
earlier—that is, that similes, metaphors, and ken- Lima, was the focus of previous studies of Chavín
nings are visually indistinguishable. art and iconography undertaken by Tello (1961)
Therefore, since the route of analysis so cre- and Lathrap (1977b). The descriptive and analyt-
atively opened up by Rowe leads, in the end, to ical strategy to be used here, based as it is on one
any number interpretive quandaries, we must re- work of art, will obviously be insufficient to elab-
turn to Chavín art itself and look for some other orate fully the iconographic details and variations
route of analysis to follow. I would suggest that thereof common to the full corpus of Chavín art.
we go back to a body of information in Chavín art What we can hopefully accomplish in the space
that most students of Chavín iconography have available here is the articulation of the theme of
undoubtedly recognized as central to the style it- the “well-ordered body,” as defined in Chavinoid
self but which, perhaps because of its prominence terms, as well as the principles of organization and
and its familiarity to us, has been entirely neglect- classification that informed the construction and
ed as a focus of study; I am referring to the body. representation of meaning by Chavín artists, as
I argue herein that the structures and rela- indicated by the metaphorical comparisons and
tions organizing the bodies of humans and ani- metonymic connections which they customarily
mals in Chavín art represent models of and for mapped onto the well-ordered body.
structured relations among actors (or other ele- The Tello obelisk is a vertical, rectangular
ments), processes, and systems of classification in shaft of granite with a step-like notch at the top.
other domains of life (for example, kinship, hunt- The shaft is carved in relief on all four sides.
ing, curing, eating). The “mapping” of sets of non- When the four sides are depicted in a single, two-
corporeal objects and relations onto the body rep- dimensional image (see figure 8.3), we see that the
resented the strategy whereby Chavín artists statue consists of two representations of what is
constructed their iconographic conventions on the apparently a single type of creature. The head,
proper and “natural” order of things according to body, and tail of each creature occupy one or the
Chavín cosmology. The resulting frameworks and other of the broad sides of the stela (figure 8.3, A1
paradigms of the body constituted what I refer to and B1), while the legs and genitalia, as well as
here as the “well-ordered body.” Finally, it is im- other subsidiary elements, occupy the respective
portant to stress that the body, with its joints reg- narrow sides, to the right of the main body (fig-
ulating movement, its orifices regulating body–en- ure 8.3, A2 and B2).
vironment transactions, and so forth, is virtually What kind of “creatures” are depicted on the
the only thing we have in common with the Cha- Tello obelisk? Julio C. Tello, who was one of the
220 C H AV Í N : A RT, A R C H I T E C T U R E , AND C U LT U R E

first scholars to address seriously the questions of wings are depicted in the form of tail feathers (fig-
form and meaning in Chavín iconography, iden- ures 8.3, 8.4; elements A-36 and B-36). Lathrap
tified this pair of images as “cat-dragons” (Tello (1977b: 339) and Rowe (1962:19) concluded that
1961:183–185). Rowe (1962:18) and Lathrap these elements represent fishtails. However, when
(1977b:338) identify them as caymans (Melana- one compares the tails of the creatures on the
suchus niger), the large alligators which, until their Tello obelisk (as well as those on the Yauya stela;
virtual extinction during this century, commonly see figure 8.1) with the objects that are obviously
inhabited the middle and lower floodplains of the tail feathers on the harpy eagles shown in Rowe
Amazon River basin. On one level, I would agree 1962:Fig. 14 and Roe 1974:Fig. 1, I think a strong
with Rowe’s and Lathrap’s identification but case can be made that these creatures are “feath-
would insist that, by virtue of the transformations ered caymans.” Thus, I am in basic agreement
of the caymans’ body parts into other animals, with Burger’s earlier interpretation of these ele-
these creatures have become something more ments as the tail feathers of an eagle or hawk
than just caymans; to use what I think might be (1992a:151). For these several reasons, I think it is
their proper Andean designation, these are warranted to refer to the pair of creatures on the
amarus (“dragon, giant serpent”). As we will see Tello obelisk by the Quechua designation: amaru.
below, amarus incorporate elements of Rowe’s With regard to the representations of the two
and Lathrap’s caymans with Tello’s “cat-dragon.” amarus, we see that they are composed of body
We do not have space here to discuss exten- parts rendered as other animals (or other animals’
sively the concept of amaru as it is used in the body parts) such that, for example, a feline/knee
Andes. Briefly, amarus have been identified is attached to a reptile/wrist (or ankle), which ter-
throughout the ethnographic and ethnohistoric minates in the clawed foot of a cayman. Thus, the
literatures as several different kinds of animals, amarus of the Tello obelisk are composed of jux-
such as cats (Zuidema 1967); large aquatic con- taposed animals, or animal body parts, represent-
strictors, like anacondas (Garcilaso 1966:222– ed within the framework of the bodies of two cay-
223, 495–496; Guaman Poma 1980:50, 65; Pacha- mans. I argue that the structure and organization
cuti Yamqui 1950:242); black bulls (Ortíz 1973), of these compositions give us information about
“dragons” (González Holguín 1952 [1608]:24); Chavín structural relations and classificatory prin-
and rainbow-serpents (Urton 1981; Whitten ciples on two levels at once. First, certain animals
1979). Two commonly recurring characteristics are regularly related to each other through an as-
of amarus are especially appropriately men- sociation of juxtaposition, or contiguity (that is,
tioned in this context. First, amarus are general- metonymy). Second, through metaphorical com-
ly thought of as composite creatures. For in- parisons, these same animals are regularly com-
stance, Ortíz Rescaniere recorded one myth from pared to certain classes of body parts; for exam-
the Mantaro Valley of an amaru that had the ple, elbows and knees are commonly represented
body of a toad, the head of a huanuco, small by, or transformed into, similar animals (such as
wings, a tail like a serpent, and was white with jaguars), as are wrists and ankles (entwined ser-
age (Ortíz 1973:69–70). In a similar fashion, the pents). These particular body parts are members
Milky Way which, among other things, is consid- of the class of body-part connectors we term
ered to represent the body of an amaru, is com- “joints.” My presumption here is that such com-
posed of several different “dark cloud” animal parisons as that just outlined between a class of
constellations, including a snake, toad, tinamou, body parts and particular types, or classes, of an-
llama, and a fox (Urton 1981). Therefore, amarus imals provide us with important information
are composite creatures—chimeras (compare about Chavín ways of viewing, ordering, and clas-
Bompiani 1989). sifying the world.
Second, as suggested in the quote from Ortíz In summary, I propose that the bodies of the
cited above, amarus are often represented as pair of amarus on the Tello obelisk serve as struc-
winged creatures. I think it is arguable that the tural framing devices for classifying and comparing
creatures on the Tello obelisk are also winged.The certain animals in relation to particular body parts,
 C H A P T E R 8: T H E B O DY OF MEANING IN C H AV Í N A RT 221

Figure 8.3. Rollout of the reliefs on the Tello obelisk Figure 8.4. Reference key to design elements on the
(from Rowe 1967a/1977:328, Fig. 6) Tello obelisk (from Rowe 1967a/1977:328, Fig. 7)

and/or classes of body parts. This suggests that it T H E O R G A N I Z AT I O N A N D

may be fruitful to discuss some general principles M E A N I N G O F A N ATO M Y
of ethnoanatomy and body symbolism before pro-
ceeding with the analysis of the particular classifi- From a number of studies of ethnoanatomical
catory principles and structural relations encoded classification and symbolism within different non-
in the bodies of the amarus on the Tello obelisk. Western cultures (for example, Classen 1993;
222 C H AV Í N : A RT, A R C H I T E C T U R E , AND C U LT U R E

Feher 1989; Franklin 1963; Lenormand 1950; are certain characteristics of the structure and or-
López Austin 1988; Marsh and Laughlin 1956; ganization of the human body that are respected
Perey 1975; Stark 1969; and Swanson and Wit- in all systems of ethnoanatomical classifications.
kowski 1977), it has become clear that the terms These include, perhaps foremost, a universal un-
and principles employed in classifying the parts of willingness to violate the order in which the parts
the body are often the same as those used in clas- of the body are physically connected. For exam-
sifying other domains of nature and culture, such ple, no society has been found in its classification
as plants, animals, and social groups. In addition, of body parts to unite the foot and thigh as a cat-
in all languages, human body-part terms are used egory opposed to the lower leg (Swanson and
in the naming of animal body parts. In light of Witkowski 1977:328). In this sense, the body rep-
this, as well as the demonstrable fact that human resents a predetermined grid of connections and
anatomical classifications are more elaborate than relations whose basic structural features are al-
the anatomical classification of non-human ani- ways the same, regardless of the cultural setting.
mals, Ellen has argued that the human body is the Another characteristic element of the grid of body
primary model of classification, “in both an evo- classifications is its symmetry. There is no known
lutionary and logico-operational sense” (1977: example in which the two ears, eyes, arms, or legs
353). In addition, Douglas has argued persuasive- are given different primary lexemes, although
ly that in many cultures, the body serves as a these symmetrical right and left body parts are, of
model of and for society: “In its role as an image course, commonly accorded different—usually
of society, the body’s main scope is to express the opposed—symbolic values (see Needham 1973).
relation of the individual to the group (Douglas Second, certain parts or elements of the body
1975:87; see also Ellen 1977:360). One context in are accorded special significance. These include
which this theme has been developed in Andean especially the joints and orifices. The principal
studies is in Zuidema’s analysis of the body of the named joints, the body “dividers,” include the
puma used as a metaphor for the organization of shoulder, elbow, hip, and knee. Interestingly,
the Inca capital city of Cusco (1985; see also Swanson and Witkowski found in a survey of the
Classen’s study of body symbolism in Inca cos- ethnoanatomical classifications of seven lan-
mology [1993:96ff.]). guages that wrists and ankles are not widely
There is still considerable speculation about named (1977:331). They note that “it is what we
whether body-part terms used in other domains might call the ‘dividers’ or ‘general markers of
are egressive (that is, extended from the body to boundaries’ that most closely approach what we
those other domains) or ingressive (that is, project- might refer to as named semantic universal con-
ed from other domains to the body). However, the cepts or primes” (1977:331). As for the orifices,
preponderance of the data suggests that in most these include especially the mouth, nostrils, eyes,
cultures, the body is primary; it provides a model ears, genitalia, and the anus. In an intriguing study
of and the terminology for classification within of body–environment transactions, Watson and
other domains. This extends to the use of relations Nelson (1967) developed a “paradigm of orifices,”
among body parts as a way of describing relations which recognizes the centrality of the orifices as
among different elements within another domain, the loci of the major exchanges between an or-
or between two similar elements belonging to two ganism and its environment:
different domains (see Ellen 1977: 357–358). All three [mouth, anus, genitalia] function to
Apart from these general observations on the relate the organism to its environment
use of the body as a model and source of classifi- through the ingestion of sustaining sub-
cations, symbols, and metaphors, there are a few stances or the expulsion of wastes and other
related points that should be stressed because of secretions. For this reason, all can easily
their relevance for interpreting certain body symbolize the exchange of gifts and dona-
parts, or features, that are emphasized in the com- tions with the rest of nature. (Watson and
posite bodies (amarus) in Chavín art. First, there Nelson 1967:296)
 C H A P T E R 8: T H E B O DY OF MEANING IN C H AV Í N A RT 223

These observations are interesting in relation While Swanson and Witkowski argue, on the
to Lathrap’s analysis of the plants represented on basis of their study of Hopi ethnoanatomy, that
the Tello obelisk. Lathrap argued that, in its to- inalienable possession is the most salient of the
tality, the Tello obelisk represents a “huge, three classificatory modes (1977:322, 325), none-
granitic doxology” in which the two creatures ap- theless, as this classificatory principle depends for
pear to be delivering the gift of cultivated plants its realization upon verbal statements, we cannot
to mankind (Lathrap 1977b:347–348). The expect to get much purchase in analyzing this
plants often extrude from the orifices of animals classificatory principle from the study of iconog-
appended to the amarus. For instance, both Tello raphy. Thus, we will focus here on the first two
and Lathrap interpreted elements A23 and A24 classificatory principles (a and b, above) in this
(figures 8.3, 8.4) as, respectively, a manioc discussion.
(Manihot utilisima) plant extruding from the Part-whole and kind-of classifications coin-
mouth of a jaguar, the latter of which is in the po- cide, respectively, with the rhetorical strategies, or
sition of the penis of amaru A (Lathrap 1977b tropes, of metonymy and metaphor. Furthermore,
:344–346; Tello 1961:184). Lathrap further sug- metonymy and the part-whole mode of classifica-
gested that this particular iconographic compo- tion are comparable to what is termed “syntag-
sition represented something on the order of a matic” relations, while metaphor and kind-of clas-
“credo” of tropical forest, horticultural societies, sifications are conceptually and in principle linked
projecting the message: “Manioc is the semen of to “paradigmatic” relations (see Leach 1979: 25–
the Great Cayman” (1977b:348). It is timely to 27). Turner (1985) has employed the contrast, and
take note in this regard of Gregor’s comment that relationship, between syntagmatic and paradig-
among the Mehinaku, who live on the upper matic elements in his highly insightful analysis of
Xingu River, in central Brazil, manioc tubers are the Kayapo myth of the “bird-nester and the fire
commonly compared to phalluses. Furthermore, of the jaguar.” As we will see later, Turner’s study
what the Mehinaku refer to as “women’s food,” of the ways these classificatory modes structure
the principal example of which is manioc, is con- form and meaning in tropical forest myths pro-
sidered to become transformed into semen in its vides a useful model for analyzing syntagmatic and
passage through the body (Gregor 1985:81–86). paradigmatic elements in art as well. I now sum-
Thus, it appears that Watson and Nelson’s “par- marize the various methodological approaches
adigm of orifices,” with its emphasis on bodily and theoretical orientations for our analysis of
openings as sites of body–environment transac- Chavín art discussed up to this point.
tions, may have considerable relevance in analyz-
ing Chavín body-part classification and symbol- M E TO N Y M Y, M E TA P H O R , A N D
ism—especially in relation to the symbolism of A N ATO M I CA L C L A S S I F I CAT I O N S
body fluids and boundaries.
Third, an important theme in ethnoanatom- Having digressed in several different directions in
ical studies has concerned the principles that un- the discussion of how we might approach an
derlie different ways of classifying body parts. analysis of Chavín body metaphors and symbols,
These principles have been defined as: I provide in figure 8.5 a diagram of the artistic
forms, tropic principles, ethnoanatomical con-
(a) part-whole (part-of, partiality; analytic)—for cepts, and the classificatory and cosmological
example, “my nose is part of my face” principles that I propose to use in analyzing the
paradigm of the well-ordered body in Chavín art
(b) kind-of (class inclusion; synthetic)—for exam- and iconography.
ple, “my index finger is a kind of finger” Figure 8.5 begins with a restatement of the
central elements and compositional forms of
(c) inalienable possession—for example, “this is my Chavín art—that is, animals, plants, and compos-
ear” ite creatures (for example, amarus). The chart is
224 C H AV Í N : A RT, A R C H I T E C T U R E , AND C U LT U R E

Bodily Chavín
Artistic Chavín Tropic Structure & Comparison Classificatory Cosmological
Form Iconography Principle Organization to Anatomy Practice Paradigm

Animals Part-whole
Elements Metonymy Connections Joints = Syntagmatic
& plants (analytic)

Composite Symmetry & Kind-of

Composition Metaphor Orifices = Paradigmatic
creatures translation (synthetic)

Figure 8.5. Elements of Chavín iconography, body symbolism, and classification

intended to be read both horizontally, along the side of the body may be likened or compared to
rows labeled “elements” and “composition,” and its mirror image on the opposite side of the body
vertically, between items in columns falling under (the left eye, and the left hand/wrist). As for the
the same heading. The horizontal reading of the comparison between upper and lower, the hand/
“elements” row in the chart traces the manner in wrist combination of the upper body may be
which discrete elements, such as identifiable likened to that of the foot/ankle connection of the
plants and animals, represented in Chavín art may lower body. The classificatory mode of such
be understood to form connected series of ele- iconographic expressions appears to rest on the
ments that reflect Chavín part-whole (analytic) principle of class inclusion (that is, “kind-of”). For
classificatory practices. In the paradigm of the instance, the right and left eyes are each a “kind-
well-ordered body, this dimension refers to the of” eye, just as the hand/wrist and the foot/ankle
representation of discrete parts connected in syn- are “kinds-of” terminus/joint combinations. The
tagmatic chains, the crucial anatomical expres- latter comparison is, of course, especially com-
sions of which are joints. pelling and generalizable when the subject in
The reading of the “composition” row in question is a quadruped, as is the case with the
Figure 8.5 points to the synthetic, paradigmatic two amarus in the Tello obelisk.
dimension of Chavín art and iconography. In this It is argued here that a full reading of figure
dimension, metaphor guides the comparison of 8.5, like a full reading of a Chavín artistic compo-
elements in compositional form. This includes sition, is realized in the combination of the
such expressions as, for instance, the comparison metonymic, syntagmatic chains with the meta-
of body parts on the right side of the body with phorical, paradigmatic transformations to pro-
those on the left, as well as substances inside the duce the well-ordered body of Chavín cosmolo-
body with metaphorically comparable (according gy. This will serve as the model or paradigm for
to Chavín ideology) substances outside the body our analysis and interpretation of Chavín iconog-
(for example, as with semen and manioc). In the raphy and classificatory practices as represented
well-ordered body, such expressions, or transla- on the Tello obelisk. As a prelude to undertaking
tions, occur through the orifices of the body. that analysis and interpretation, it will be useful
Concerning the notion (suggested in figure 8.5) to take account of the semantic strategies and
that symmetry is a kind-of, or synthetic classifica- classificatory principles of the naming of body
tory form, we see expressions of this in Chavín art parts in an indigenous Andean language. Now, we
in two dimensions: right/left and upper/lower. As do not know what language was spoken by the
to the former, whatever body part (for example, people who built, occupied, and regularly visited
the right eye), or set of interconnected body parts the site of Chavín de Huántar. In my discussion
(for example, the right hand/wrist), exists on one below, I make use of material drawn from one of
 C H A P T E R 8: T H E B O DY OF MEANING IN C H AV Í N A RT 225

the varieties of Quechua (that is, Southern Peru- data that appear to be most directly relevant for
vian), which was a widespread language of the understanding body symbolism and body-part
coast, highlands, and parts of the tropical forest classifications as represented in Chavín iconog-
of Peru at the time of the Spanish conquest.3 raphy; that is, I focus here on terminology relat-
Before turning to this discussion, I want to ing to joints, orifices, teeth, extremities (hands
make clear that my purpose in presenting and feet), and what I refer to as the “landscape,”
Quechua ethnoanatomical material is not to sug- or geography of the body.
gest that these data are directly relevant to the
interpretation of Chavín iconography. In gener- Joints
al, my hope is that the Quechua anatomical terms One of the body parts that we find to be of cen-
and concepts that we encounter in this discussion tral importance in the anatomical vocabulary in
may provide us with useful conceptual and clas- the Tello obelisk is the joints. In Quechua, there
sificatory tools with which to talk about the or- are distinct primary lexemes for the elbow
ganization of form and meaning in this ancient (kukuchu, “corner; something bent over”), knee
Andean iconographic tradition. Certainly, we (muqu, “hill, mound”), pelvis (chaka teknin, “hip
will be better off in our attempt to develop a cross[-ing]”; or cintura pata, “waist ledge”; Stark
meaningful heuristic device and an analytical vo- 1969:10) and the neck (kunka). However, the
cabulary with which to talk about Chavín wrists (maki muñica, “arm/hand wrist”) are simi-
iconography by exploring terms and concepts lar to the ankles (chaki muñica, “leg/foot wrist”).
that derive from “well-ordered bodies” in any As for the body parts connected by joints, Ellen
one of the varieties of Quechua, rather than if we (1977:366) has noted that joints often represent
rely on English body-part terms and classes, or, links between areas of relative undifferentiation,
worse yet, if we refer to works of Chavín art by or of classificatory uncertainty. The common
such potentially misleading characterizations as term for such long, undifferentiated segments of
“the Smiling God,” “the Staff God,” or “guardian the extremities in Quechua is llañu (“a long, thin
angels” (Rowe 1962). cylindrical thing”). For example, the buttocks and
knee stand at the terminus points of the under-
Q U E C H UA E T H N OA N ATO M Y side of the “thigh” (llañu chaka); the elbow and
shoulder terminate the long stretch, called llañu
The data on Quechua ethnoanatomy discussed rikra, from the shoulder down to the elbow; the
below are drawn from several different sources. portion of the arms between the elbow and the
One particularly valuable source is a study by wrist is called llañu maki.
Louisa Stark, entitled “The Lexical Structure of
Quechua Body Parts” (1969). I also refer to data
Orifices
on contemporary Southern Peruvian Quechua
ethnoanatomy that I collected in the community As for the orifices, the “eye” (ñawi), “ear” (ninri),
of Pacariqtambo (Prov. of Paruro, Dept. of and “nose” (sinqa) openings are all conceived of as
Cusco) in two sessions of fieldwork, in 1981–82 “openings, splits, or windows” (t’uqu). However,
and 1987–88.4 Finally, wherever appropriate, I unlike these other orifices of the head, the mouth
have also drawn on anatomical terms and con- is not conceptualized as a t’uqu; rather, it is com-
cepts provided in the early seventeenth century posed of a mouth [opening] (simi) and an interi-
Quechua dictionary of González Holguín (1952 or mouth cavity (simi uxu). The lower body ori-
[1608]; this is the late pre-hispanic, Incaic variety fices are also (like the mouth) lexically distinct:
of Quechua, probably ancestral to contemporary anus (uhete; or sip’uti; Stark 1969:10); vagina
Southern Peruvian Quechua; see Mannheim (raka); and the urinary opening of the penis (his-
1991).5 As there is not space here to give a com- p’añin; Stark 1969:10).
plete accounting of Quechua body-part termi- The terms outlined above suggest that there
nology, I confine my discussion below to those exists in Quechua thought a dual classification of
226 C H AV Í N : A RT, A R C H I T E C T U R E , AND C U LT U R E

orifices. This involves a distinction between the tokma (“canine; fang”), and chocta quiru (chocta =
three orifices involved primarily in the intake, “long pointed [thing]”). Therefore, the canine
through t’uqus, of sensory data (that is, eye, ear, teeth define the boundary of the “doorway” into
nose) as opposed to the orifices through which and out of the body. Marked canines indicate, of
material transactions take place between the body course, a particular dietary predilection shared by
and the environment (mouth, anus, vagina, and humans and certain animals, most notably felines.
urinary opening). Therefore, the possession of canines represents a
It is no doubt significant that body–environ- point of comparison linking humans and jaguars.
ment transactions in the Tello obelisk occur pri- As a marker of boundaries and site of transforma-
marily through the mouth, the penis, and the tion, canine teeth can be compared to joints and
nose. The Quechua category of orifices through orifices.
which body–environment transactions occur, and
therefore those through which material transfor- Hands and Feet
mations take place between the inside and the out- Although in contemporary Quechua the hands
side, is similar but not identical to what we see de- and feet are given different primary lexemes
picted on the Tello obelisk. The addition of the (hand = maki; foot = chaki), nonetheless their con-
nose as an orifice of body–environment transac- stituent parts are identical—that is, finger/toe =
tions in Chavín art may be related to the practice riru (Sp. dedo; Q. rucana; González Holguín 1952
of ingesting powdered hallucinogenic snuffs [1608]:319); palm/sole = pampa (“flat plain”); hol-
through the nose. We have a substantial amount low of hand/arch = puxyu (“spring”); and finger-
of archaeological evidence for this practice at nail/toenail = sillu (Stark 1969:11). Stark’s com-
Chavín sites, including mortars, bone trays, spat- ments on this point are relevant here:
ulas, and snuffing tubes (see Burger 1992a:
157–159). This evidence is complemented by de- This use of identical lexemes may indicate
tailed ethnographic descriptions from present-day that the Quechua Speaker conceptualizes
societies in the lowlands of South America (see the hand and foot as being similar, if not
Reichel-Dolmatoff 1975; Dobkin de Rios 1984 identical, entities. This interpretation is
[1972]), as well as by historical sources. Among the supported somewhat by the fact that in the
latter is the following description of drug use by pre-Conquest (Inca) art of Quechua speak-
the Muisca of Colombia provided by the seven- ers, the hands and feet of the human figure
teenth-century friar Pedro Simon: are generally almost identical in represen-
tation. (Stark 1969:8)
[T]hey take these powders and put them in
their noses and which, because they are These observations are important for our
pungent, make the mucus flow until it analysis of the Tello obelisk, because the hands
hangs down to the mouth, which they ob- and feet—or the front and hind legs—of the
serve in the mirror, and when it runs amarus are rendered identically (see figures 8.2–
straight down it is a good sign. (cited in 8.4, A- and B-11, and A- and B-34).
Burger 1992a:157)
The Landscape of the Body
Teeth An important phenomenon in Quechua anatomi-
The importance of the mouth as a portal of entry cal classification is a lexical overlap in body-part
and exit in Chavín iconography may be reflected terms with terms and concepts referring to topo-
in the strong emphasis on the canine teeth. In graphical features of the landscape. Stark (1969:
Quechua, the teeth are categorized into two prin- 8–9) has provided a good discussion of these terms,
cipal groups: (1) the incisors and canines = punku and I take one example here in order to suggest
kiru (“entry, or door teeth”), and (2) all the teeth the possible significance of geographical concep-
behind the canines = waqo kiru (“cheek teeth”). tions and classifications for the construction and
González Holguín refers to the canine teeth as representation of meaning in the Tello obelisk.
 C H A P T E R 8: T H E B O DY OF MEANING IN C H AV Í N A RT 227

The Quechua term for knee is chaki muqu, from the point of view of 1) the horizontal
and the term for shoulder blade is wasa muqu. contrasts it makes on the level on which it
Muqu is also the word for “hill.” These various occurs, and 2) its vertical relationship to the
uses of the term muqu indicate the existence of a lexeme of which it is a constituent. (Stark
synthetic principle of classification in which 1969:3)
“hills” can be said to stand—and function—in the We will return to comment on this important ob-
geographical domain in a manner comparable to servation in our discussion of the vertical and hor-
knees and shoulder blades in the anatomical do- izontal layout of iconographic elements in the
main. That is, hills provide links between river Tello obelisk (below).
valleys, just as knees link thighs and shins (that is, Having outlined a number of Quechua ethno-
two stretches identified as llañu) and as the shoul- anatomical terms and principles, we can now turn
der blades connect shoulders (= rixra, “shoulder our attention to the Tello obelisk to see how the
of a mountain”) with the depression between the bodies of the two amarus on this stela are organ-
shoulder blades (= wasa wayq’u, “back ravine”). ized. Following this overview, we then return to
Furthermore, I would argue that the anatom- compare the features of the Chavín “well-ordered
ical correlates of “hill” (muqu), representing as they body” to the organizational concepts and princi-
do points of connection and mediation, allow us to ples of classification in Quechua. Our objective in
extend these metaphorical connections to the an- this comparison is to investigate whether or not
imal domain (as is clearly done iconographically in there are any commonalities between the two sys-
the Tello obelisk) by linking, for example, elbows, tems that might provide us with a grounded, con-
hills, and jaguars. These elements and characters textual approach—that of Quechua ethnoanato-
perform similar mediating functions in their re- my—for interpreting form and meaning in
spective domains; therefore, they may be meta- Chavín iconography.
phorically compared to each other in iconograph-
ic expressions. This is the conceptual basis for
making “comparisons by substitution” (kennings),
ANIMAL SYMBOLS AND
as discussed earlier. However, while the literary A N ATO M I CA L C L A S S I F I CAT I O N S
trope of kenning—as the end point of a progres- IN THE TELLO OBELISK
sion of tropes beginning with simile and passing
through metaphor—has no natural grounding, I begin with an outline of what I think are some
nor can it be expressed, in iconography, the inter- of the principal diagnostic features and structur-
pretive paradigm of the “well-ordered body” pro- al relations of the two amarus on the Tello obelisk.
vides the motivation, rules, and logic for compar- Several general observations should be made ini-
isons by substitution in Chavín art. tially:
(a) For the most part, the two amarus are shown
Forms of Classification in Quechua in right profile; thus, the representational
Ethnoanatomy statement made here seems to be that these are
Although her argument has been subject to crit- two different creatures. However, I say this is
icism (see Swanson and Witkowski 1977:324), true “for the most part” because elements A-
Stark maintains that Quechua body parts are con- and B-36 (figure 8.3)—the heads from which
ceived of in horizontal and vertical hierarchical the tail feathers protrude at the bottom of the
levels; these, she argues, reflect the overarching stela—are shown (respectively) in right and
operation of a principle of partiality (that is, the left profile. It is as though the two amarus
relations of part-to-whole) in Quechua anatomi- begin as a single entity, with complementary
cal classifications and naming. As stated by Stark: right and left profiles, at the base but then be-
[T]he semantic dimension of [an anatomi- come differentiated, moving up the stela as the
cal] . . . lexeme may depend in part upon its single, lower amaru is transformed into two
position within an overall hierarchy, both parallel versions of the original image. This
228 C H AV Í N : A RT, A R C H I T E C T U R E , AND C U LT U R E

may derive from a principle of the unity of like (e) There is an overwhelming iconographic in-
objects; that is, the understanding that differ- terest in the canine teeth, which are often de-
ent forms of a type (in this case, amarus) are at picted on felines located at the joints, the
some level alike and share a common origin. “body dividers” (that is, at the elbows, necks,
knees, pelves, and the point of connection be-
(b) As noted earlier, the bodies of the amarus are tween the pelvis and tail).
divided vertically so that the head, trunk, and
tail of each occupy one broad side of the slab (f) Joints are often associated with orifices, espe-
while the extremities and genitalia occupy cially with the mouths of jaguars.
the narrow side (to the viewer’s right). This In addition to these general observations on
establishes a hierarchy of body parts with the structure and organization of the body parts
those elements on the broad side(s) as most and composite creatures on the Tello obelisk,
inside and those on the narrow side(s) as out- there are a number of more specific observations
side; the two classes—inside and outside— to be made:
are connected at body joints. (g) Elbows and knees are represented by similar
(c) In keeping with the Chavín design convention but not the same kind of felines.
of dividing figures into a number of horizontal (h) Wrists and ankles both incorporate reptilian
“modular bands” (Rowe 1962:14), the bodies forms.
of the amarus appear to be divisible into five
modular bands, organized as shown in table 8.1. (i) Hands and feet are depicted in almost the
same way.
(d) There is a general emphasis on heads; for ex-
ample, the various parts of the bodies of the (j) There is an equal number of fingers and toes
amarus are made up of the heads of other an- — four of each; these are represented as one
imals, especially felines (for instance, elbows, thumb/big toe together with three fingers/
knees, tails, and penis = heads of jaguars). The toes.
message seems to be that identity is formed I interpret the significance of the above obser-
and expressed by, or in, the head of an animal. vations in the following ways, in reference to the

Table 8.1. The organization of modular bands and body parts on the Tello obelisk
Modular band Body parts Design elements
(see figs. 8.3, 8.4)
I Head A- and B-5, -6
Wrist A- and B-12
Hand A- and B-11
II Neck A- and B-7, -8, -9, -15
Elbow & Forearm A- and B-14
III Trunk A- and B-16
Genitalia A- and B-24
IV Pelvis A- and B-25, -26, -27, -28, -29
Knee A- and B-31
Ankle A- and B-33
V Tail A- and B-36
Foot A- and B-34
 C H A P T E R 8: T H E B O DY OF MEANING IN C H AV Í N A RT 229

interpretive diagram shown in figure 8.6. This dia- section from the top down to the center (band I
gram illustrates the organization and classification to III) is complemented by the section from the
of body parts according to the classificatory “grid” bottom up to the center (band V to III). In view
provided by (a) the vertical division of the two bod- of the emphasis on symmetry and complementar-
ies into “core” (that is, trunk) and “periphery” ity in other expressions of the organization of
(namely, extremities), and (b) the horizontal mod- iconographic elements in the Tello obelisk, I ar-
ular bands coordinating relations among upper/ gue for the latter of the two alternatives outlined
central/lower and inside/outside body parts. above.
Under the category of metonymic relations, I Combining this observation with that made
reemphasize the point that the head, trunk, and above, in point (b), I would suggest that there is a
tail on each of the broad sides are juxtaposed to strong design convention in the Tello obelisk em-
the extremities and genitalia along the respective phasizing vertical and horizontal complementar-
narrow side. The connections between the cen- ity. This represents a form of “parallelism”—that
tral shaft of the body and the appendages occur is, the iconographic means for positing a structur-
at “joints”—the elbows, knees, and the crooked al metaphor—between the elements from the
penis—marked by jaguar heads. The element in head and tail inward to the trunk with that from
the position of the genitalia of amaru B (element the extremities on the narrow panels to their at-
B-24) is crooked but is not depicted as a jaguar tachments with the trunk on the broad panel (that
head. Thus, except for element B-24, we learn is, from modular bands I to II and V to IV). Thus,
that joints in the anatomical domain are likened to the head downward to the neck is likened to the
jaguars in the animal domain. hand inward to the elbow; and the tail upward to
The second observation to stress concerning the pelvis is likened to the relationship from the
metonymic relations in the layout of the Tello foot inward to the knee.There is, therefore, a pow-
obelisk is the juxtaposition of the five modular erful proposition represented in the modular band
bands in the vertical dimension, from the top to organization of the anatomy of the two amarus,
the bottom of the stela. This arrangement could which takes the form of a complementary relation-
be interpreted as establishing either a hierarchical ship between upper and lower with outside and in-
organization of elements along a continuum from side. The tropic principles directing these expres-
the top to the bottom or from the bottom to the sions of parallelism are metonymy and metaphor,
top, or a symmetrical organization in which the acting simultaneously and in concert.

Figure 8.6. Schematic represen-

tation of the well-ordered body on
the Tello obelisk
230 C H AV Í N : A RT, A R C H I T E C T U R E , AND C U LT U R E

Finally, the modular band division of body and outside of a body. The Tello obelisk glosses
parts along the central axis of the amarus incor- this relationship between classes of anatomical
porates an iconographic theme noted earlier in parts and processes through a particular kind of
the discussion of the points of connection be- animal: the jaguar. This key iconographic propo-
tween the extremities and the trunk; that is, joints sition requires further contextualization and
are compared to jaguars. We note that along the commentary.
central axes, jaguars are found at the neck (ele-
ments A- and B-7) and at the point of connection THE ICONOGRAPHY OF
between the hips and the tail (elements A- and B-
T R A N S I T I O N A N D M E D I AT I O N :
25, -27, and -36). Therefore, the points articulat-
ing the extremities with the trunk are compared J AG UA R S , J O I N T S , A N D U N C L E S
to the points of articulation between the upper Reichel-Dolmatoff’s masterful study (1975) of the
and lower parts of the body inward to the trunk. ideology and symbolism of human/jaguar trans-
It is instructive to note that the overall struc- formations among Tucanoan-speaking peoples of
ture of the bodies of the amarus arrived at above is southeastern Columbia provides ample evidence
strikingly similar to the divisions and organization that, in many parts of the lowlands of South Amer-
of body segments by means of body decorations ica, the jaguar was (and still is) thought of as the
found among the Yekuana of the Upper Orinoco principal animal capable of transforming into a
River basin, as described by Guss (1989; see also human, specifically, a shaman; in this role, the
Seeger’s 1975 discussion of body ornaments jaguar articulates, or mediates, between humans
among the Suya). The Yekuana body decorations and animals and between humans and spirits
include arm bands on the upper arms (between (Reichel-Dolmatoff 1975:130–132; on the role of
biceps and shoulder); strands of beads wrapped jaguars as mediators in myth, see Turner 1985:
tightly around the calves, just below the knees; 63–64). Furthermore, Whitten has noted (1976)
white beads wrapped around the wrists; neck- that among the Sacha Runa of eastern Ecuador,
laces; and a loincloth passing just below the stom- jaguars are conceptually and terminologically re-
ach (Guss 1989:41–42). As Guss notes, lated to the kinship category of “uncle,” especial-
Yekuana dress is a deliberate charting of the ly mother’s brother. It is interesting to note in this
human space, with the trunk of the body fas- regard that in the invocation to keep jaguars away
tidiously circumscribed from the outer limbs from villages among the Guaraní-speaking Paí-
and head. . . . [E]ach Yekuana body is inter- Cayúa, the jaguar is referred to as che tuty, “my
sected by two imaginary concentric circles, uncle” (Cadogan 1973:98).
the outer running through the wrist and To expand what appears, upon comparative
ankle bands and the inner through those of study, to be an interconnected set of symbolic as-
the biceps and calves. Thus. . . the outer ring sociations, including jaguars, shamans, and medi-
of the body—between knee and ankle, bicep ating classes or categories of kin (such as uncles),
and wrist—is also a world of differentiation with other “mediating” classes or categories of in-
and division. The inner circle . . . is a world dividuals, Kensinger reports that most of his male
of wholeness and union. (Guss 1989:42) Cashinahua informants said that when aroused,
women “become sexually aggressive and insa-
To return to the Tello obelisk, the compari- tiable because they have an erect and hot penis
son between joints and orifices in the iconogra- (clitoris) which causes them to be like jaguars,
phy suggests an important conceptual relation- both dangerous and exciting” (Kensinger 1995:
ship in Chavín thought between articulation 81). On a related theme, Riviére has noted that
(joints) and transformation (orifices). That is, a among the Trio Indians of Surinam, shamans are
point at which a bend, break, or articulation oc- compared to menstruating women (1969:268),
curs in a connected sequence is similar to a place and Hugh-Jones has remarked on the saliency of
where transactions are made between the inside such a connection for the Barasana, among whom
 C H A P T E R 8: T H E B O DY OF MEANING IN C H AV Í N A RT 231

shamans play a central role in regulating menstru- view of any particular ego, one may reach an in-
ation through their control and manipulation of dividual—such as an “uncle”—who articulates
sacred paraphernalia (1988:125). This last com- ego’s kin with other groups. Such a linking posi-
parison links the transactional symbolism of ori- tion becomes a point of “articulation” and “trans-
fices—or of that which, like menstruating women, action” between ego and those outside his/ her
is considered to be excessively “opened up” (see group, or of transformations from one group or
Hugh-Jones 1988:125–126)—with joints through status and another. Such would be the case, for ex-
the comparison among menstruating women, ample, in the numerous lowland South American
shamans, and jaguars, all of whom are considered societies which have (or had) prescriptive cross-
to be in positions or states of mediation, transfor- cousin marriage and an accompanying termino-
mation, and transaction with the environment. logical equation between, say, wife’s father and
Finally, it is relevant to note here Burger’s in- mother’s brother.6
terpretation of a veritable iconographic “pro- In terms of the social and reproductive sig-
gram” at the site of Chavín de Huántar focusing nificance of the iconic set jaguar/joint/uncle, I
on jaguar-shaman transformations based on the would note that on the Tello obelisk, the penis of
medium of hallucinogenic drugs. The iconogra- amaru A (figure 8.3, A-24), which is situated on
phy in question is that seen on several sculpted the narrow panel A1, is represented with a jaguar
stone tenon heads, which adorned the walls of one head. In this position, the jaguar-penis medi-
of the temples at Chavín de Huántar. The tenon ates—that is, is a “joint” between—the two com-
heads can be grouped into three sets, each repre- posite creatures (see Lathrap 1977b, for more
senting a stage in the progressive and reciprocal discussion on the jaguar as a mediator in Chavín
transformation between shaman and jaguar. iconography).
Burger’s analysis of these sets of sculptures is that
“they represent different stages in the drug-in- THE ICONOGRAPHY OF
duced metamorphosis of the religious leaders (or R U P T U R E , T R A N S I T I O N , A N D B I RT H :
their mythical prototypes) into their jaguar or
CANINES AND EGG TEETH
crested-eagle alter egos” (Burger 1992a:157). One
of the iconographic markers of the transition be- Finally, in terms of ideas and iconographic expres-
tween shaman and jaguar is the representation of sions relating to transition, mediation, and the
the snuff-induced flow of mucus from the nostrils like, we should take note of the interesting but
of these transforming beings. This reaffirms our puzzling Chavín convention of the “agnathic”
earlier suggestion identifying the nose as a major tooth row of caymans. This convention involves
orifice of body–environment transactions in the representation of a sharply pointed tooth pro-
Chavín iconography. truding from the center of the upper jaw, between
To the degree that comparisons between the incisors (see, for example, figure 8.1). Lathrap
anatomical and kinship classifications may be rep- (1977b:339) suggests that this trait may have
resented in the Tello obelisk by such strategies as started out as a depiction of the constantly visible
comparing articulation and transformation, this upper tooth row of the cayman. Concerning the
may have enormous significance for our study of agnathic element, Rowe stated:
the social implications of Chavín iconography.
As early as Phase AB we find front view ag-
For such comparisons as those mentioned here
nathic faces provided with a pointed tooth
may be interpreted as motivated by Chavín ideo-
in the center as well as the usual canines on
logical principles, which may point to what were
each side. The central tooth is a pure prod-
common social practices in Chavín society. For in-
uct of the imagination which can be based
stance, a “confusion” between articulation (joints)
on no observation of nature. (1962:17)
and transaction (orifices) may suggest a wider cul-
tural principle such as the notion that, in reckon- I suggest that Lathrap and Rowe may both
ing along a collateral line of kin from the point of have been wrong, and that the agnathic trait was,
232 C H AV Í N : A RT, A R C H I T E C T U R E , AND C U LT U R E

indeed, based on an observation of nature. The tion constructed of such heterogeneous elements
central tooth probably represents the “egg tooth” may, in fact, be its methodological and genealog-
of a baby cayman. As the naturalist Cutright noted ical similarity to the processes of selection and
long ago: composition by which the Chavín artists con-
structed the two images on the Tello obelisk. This
Often the young [caymans] are unable to es-
similarity may, however, provide just enough of an
cape from the eggs without maternal assis-
opening to allow a new perspective on form and
tance, even though each one is generally
meaning in Chavín iconography. By this I mean
equipped with an egg tooth. This is an ex-
the following.
iguous structure sticking up from the ante-
As we have seen here, the bodies of the ama-
rior end of the upper jaw like the sight on
rus provide the framework of a universal logic—
the end of a rifle barrel. (Cutright 1943:233)
based as it is on the human body (see Brown
The egg tooth is the instrument for ruptur- 1991)—for the study and interpretation of Cha-
ing the container (the egg), which allows the cay- vín iconography. The elements, or terms, of that
man to make the transition from inside the egg logical framework are formed by the segments,
out into the world. It is of great interest to note nodes, and relations of what I have called here the
as well that Chavín artists often combined the egg well-ordered body. The unity and the sense of
tooth with well-developed canines—the marker such composite wholes are represented in, and are
of “felineness” in Chavín iconography. This com- to be understood through, the language and logic
bination appears to link the jaguar, the animal of based on metonymic (syntagmatic) and meta-
transition, mediation, and transformation par ex- phoric (paradigmatic) structures and relations of
cellence, with the cayman at the moment of the the human body. These devices and strategies are,
transition of the latter from inside (the egg) to in fact, similar to those used by Turner in his in-
outside (see Roe 1982a, who has discussed these terpretation of the Kayapó myth of “the bird-
and related symbolic features with extraordinary nester and the origin of cooking fire” (1985). For
insight). example, Turner concluded from his study of this
In the variety of ways outlined in this section, myth that:
I argue that the well-ordered bodies of the amarus
on the Tello obelisk represent sites for portray- symbols have an internal structure, not
ing, and working out, the various terms and ex- only of static oppositions but of coordinat-
pressions of a few key symbols and organizing ed transformations of the relations among
principles in Chavín cosmology. their constituent meaningful features. This
structure is homologous with the relations
between the symbols in question and the
SYMBOLS AND PRINCIPLES OF
other symbolic elements of the composi-
C H AV Í N C O S M O L O G Y tions to which they belong. The meaning
and structure of a symbol is . . . radically in-
In the previous three sections, I have attempted
separable from the structure of the com-
to pull together in a preliminary way data from
posite form in which it is embedded.
different ethnographic and natural historical
(Turner 1985:52– 53)
sources in order to interpret the significance of
animals and body parts in Chavín iconography. Turner’s analysis (which he argues convinc-
The product of the combination of these sections ingly is fundamentally different from a structural-
itself has the feel of a composite “creature”—a ist analysis; 1985:53) of the form and meaning of
chimera—like the amarus on the Tello obelisk. a multi-episodic myth—that is, a “composite
That is, both these interpretation and the amarus creature”—from the corpus of Kayapó myths
are composites of elements drawn from diverse provides us with a useful perspective from which
sources from around the South American tropi- to think about both the relationship between the
cal forest. The principal virtue of an interpreta- individual parts and the composite whole figures
 C H A P T E R 8: T H E B O DY OF MEANING IN C H AV Í N A RT 233

composing the two amarus on the Tello obelisk, the two amarus carved in opposition to each other
as well as the possible relationship between these across a slab of granite display a complex, redun-
(part and whole) images and the myths inform- dant “message” built up around the topic of the
ing the iconography.7 These comments lead us body and organized according to the themes of
back to the topic of the relationship between mediation, transition, and transaction; of bound-
myths and kennings with which we began this aries and the rupture of boundaries; of fertility, re-
study. production, and birth; and the complexities of in-
As I pointed out earlier, the original idea be- dividual and social identity and alterity. As to the
hind Rowe’s (1962) introduction into Chavín identities of the characters in this drama and the
studies of the heuristic device of the kenning nature of the individual and collective actions that
seems to have been the desire to find a way to in- led them to become the subjects of representation
duce the iconography to “speak” to us about the in the Tello obelisk, I suggest there is only one
myths that lay behind, organized, and gave mean- moderately reliable source of information that we
ing to the imagery. We have been straining to can turn to—this is to the ethnographic materials
hear some murmur of these myths resonating containing accounts of myths, rituals and artistic
from the sounding board of the kennings we have practices, as well as classifications of animals,
identified in Chavín art over the past three de- plants, social groups, and body parts, that have
cades, without notable success. As I have tried to been collected over the years among native peo-
show in this article, I don’t think the device of the ples of the tropical forest of South America. I have
kenning is well suited to the task we have assigned undertaken only a very preliminary review of
it. That is, to say that an icon is an example of a these materials in the present paper; much more
“comparison by substitution” (that is, a kenning) work remains to be done.
does not move us very far along the path of in-
terpretation. However, saying that that same icon N OT E S
is located at a determined position within a body
immediately places that icon in a logical (in bod- 1. This chapter is reprinted in revised form from a
ily terms), well-ordered framework of surfaces, previous article, “The body of meaning in Chavín
Art,” Res 29/30, Spring/ Autumn 1996: The Pre-
joints, orifices, and—based on our knowledge of
Columbian, pp. 237–255. Copyright 1996 by the
how bodies work—internal processes, fluids, and President and Fellows of Harvard College. I
sensations. Here, we are in a world about which would like to express my appreciation to the fol-
we have some good intuitions—intuitions that lowing people for reading and commenting on
may provide us with ways of articulating some of earlier versions of this paper: Richard Burger,
the principles, structures, and values by which Billie Jean Isbell, Bruce Mannheim, Julia Meyer-
metaphorical and metonymic relations among son, Ann Peters, Johannes Wilbert, Tom Zui-
body parts, animals, and plants are organized in dema, and an anonymous reader selected by the
editor of Res. The various comments and sugges-
Chavín iconography, such as that on the Tello
tions for improvements made by all of these peo-
obelisk. But the implications of this interpretive ple have been enormously helpful to me in writ-
approach do not stop at the corporeal level; this ing this paper. I alone, of course, am responsible
is both because society is consistently imagined, for the opinions expressed, and any errors that re-
or seen metaphorically, as a body, and because main, in the paper.
while bodies are individually experienced, they are 2. For information on the history of research at the
collectively socialized. Thus, whatever we learn site of Chavín de Huantar, and of studies of
about bodies and their states of transition from Chavín culture more broadly, see Benson 1971;
and Burger 1984, 1992a.
the study of Chavín icongraphy ought to be a
3. For studies of the differentiation among and the
path of entry into the study of Chavín society. chronology of the various dialects of Quechua, see
From our preliminary examination here of Parker 1963; Torero 1964; and Mannheim 1991.
some of the images mapped onto the well-or- 4. Unless otherwise indicated, the contemporary
dered bodies of the Tello obelisk, I conclude that Quechua anatomical terminology given below
234 C H AV Í N : A RT, A R C H I T E C T U R E , AND C U LT U R E

derives from my own fieldwork. As the site of ing Quechua conceptions of the senses, rather
Chavín de Huantar is located in central Peru, it than (as is the case with the present study) concep-
would no doubt be more directly relevant to our tions and classifications of bodies and body parts.
study to work with one of the Central Quechua 6. See, for example, the series of articles describing
varieties of this language. I have used the such relationships and terminology in lowland
Southern Peruvian variety (Cusihuamán 1976) South American societies in Kensinger 1985. I
here because that is the variety with which I have would note here the interesting comparative per-
considerable experience and a reasonable level of spective on such relationships that we gain from
fluency in speaking. I invite my colleagues who Bulmer’s study (1967) of ideas about cassowaries
specialize in any one of a number of central —which are considered to be like sisters and cross-
Peruvian languages spoken in the central high- cousins to men—among the Karam of the New
lands or the tropical forest to take up this study. Guinea highlands.
5. Classen has recently (1993) provided a valuable 7. Although not incorporating ethnographic mate-
summary and analysis of Inca anatomical terms rials, one of the best examples of the general type
and concepts, primarily as recorded in the Que- of iconographic analysis that I have proposed
chua dictionary of González Holguín. Classen’s herein is Ann Peters’s study of animal and plant
study is concerned principally with understand- imagery in Paracas embroidered textiles (1991).