onkings Page 41
recreated at the kingdom level in late pre-Conquest times in the relations be-
tween the Texcoco “Chichimecs” and the Mexica “Toltecs” (Ixtlilxóchitl 1840;
Townsend 1987; Douglas 2010; Duran 1994). The Triple Alliance was rather
more of a diarchy in which the Texcoco ruler–particularly the famous King
Nezhualcoyotl–bearing the inherited title of “Lord of the Chichimecs,” ap-
pears as second in authority to the Mexica “Lord of Culhua.” The Texcoco kings
adopted the Mexica tutelary god, Huitzilopochtli, and enshrined him alongside
Tlaloc in the central temple of their own city. Moreover, at least from the time
of the Moctezuma I, inaugurated in 1440, the Texcoco rulers were the king-
makers in Tenochtitlan. The Chichimec kings of Texcoco legitimated the Toltec
kings of the Mexica: Nezhualcoyotl and his successor Nezhualpilli were the
principal electors of the Mexica tlatoani, and the ones who actually crowned the
latter in the investiture ceremonies.
The Texcoco rulers’ claim as Lord of the Chichimecs represented their
descent from the legendary Chichimec conqueror Xotlotl, said to have cre-
ated a great “Empire” in the Valley of Mexico.17 True to classical stranger-king
traditions, the Texcoco people’s own story is that the “primitive” Chichimecs
17. The other title of the Texcoco king, “Lord of Alcohua,” as the heritage of the “empire”
of Xolotl, bore the same Chichimec connotation as “lord of the Chichimecs.”
THE STRANGER-KINGSHIP OF THE MEXICA
245
led by Xolotl predated the “civilized” Toltecs in the occupation of the Valley.
In Eduardo de J. Douglas’ reading of the old Texcoco documents such as the
Tlohtzin Map, the Chichimec ancestors are depicted as “self-generating and
autochthonous, like the Mixtec ancestors”; while in the Codex Xolotl and the
Quinitzin Map, “the Toltecs are the migrants, and the Chichimecs, the native
inhabitants . . .” (2010: 55, 58; cf. Bierhorst 1992: 5 for an analogous tradition).
In thus depicting their ancestors as the autochthons, the Texcoco traditions
differ from the common narratives of Chichimec migrations into the Valley fol-
lowing the collapse of Tollan of the Toltecs. But that is not the only anomaly of
Texcoco’s Chichimec identity.
The more interesting anomaly is that the Texcoco rulers, for all their Chi-
chimec identity, were affiliated genealogically and assimilated culturally to the
Toltecs in the same fundamental ways as the Mexica kings who claimed to
be the Lords of Culhua. Beginning with Xolotl’s immediate descendants, No-
paltzin and Tlohtzin, the founders of Texcoco dynasty repeatedly married and
fathered their successors by Toltec women: brides and mothers who, as Doug-
las notes, “transmit civilization, and the Toltec legacy, to their daughters and
eventually their male descendants” (Douglas 2010: 105). But only eventually to
their male descendants, who rather follow their fathers as Chichimecs for some
generations. Xolotl’s grandson Tholtzin was a Chichimec, although his mother,
Nopaltzin’s wife Azcaxochiti, was a Toltec, a daughter of the royal house of Cul-
hacan—the same kind of alliance that produced Acamapichtli and launched the
Toltec heritage of Mexica kings. In another example, two members of Xolotl’s
court move away and marry the daughters of the ruler of a Toltec city; each cou-
ple has a daughter and a son, and the daughters are born Toltecs and the sons
Chichimecs (ibid.: 223 n. 46). The implication of ethnic affiliation descending
in separate male and female lines, the former Chichimec and the latter Toltec, is
what actually is pictured diagrammatically in Texcoco documents. Whether this
was an expedient the illustrator devised for representing double descent—pat-
rilineal and matrilineal—is hard to say.18 But safe to say, the self-representation
of the Texcoco aristocracy as “Chichimec” is arbitrary: as arbitrary as it is politic.
18. In this connection, an emphasis on matrifiliation in an otherwise patrilineal order
is not necessarily an indication of a cognatic or a double-descent system (cf. Calnek
1974; Kellogg 1986), inasmuch as a kinship order of preferred patrilineal succession
and group affiliation commonly involves vital complementary relations to affinal/
maternal kin (Leach 1961; Sahlins 2013).
246
ON KINGS
“Do not forget that you are Chichimec,” the dying Texcoco king Ixlilxochiti
told his son Nezhualcoyotl (Ixtlilxóchitl 1840: 127).
It follows that the Toltec identity of Mexica kings is equally arbitrary, since
their maternal descent from Culhuacan kings does not differentiate them from
Chichimec rulers who could claim as much. Rather than some sort of pre-
scriptive identity, the process in play is a high-stakes mode of complementary
schismogenesis in which major kingdoms selectively position themselves vis-à-
vis each other by adopting contrasting values from a common stock of cultural
traditions. Texcoco was already involved with the Tepanecs of Azcapotzalco in a
conflict of the symmetrical schismogenesis variety before the rise of the Mexica
and the formation of the Triple Alliance. Not long after the Tepanec ruler Tezo-
zomoc proclaimed himself “Lord of the Chichimecs,” which was something of
a usurpation in the Alcohua view, the Texcoco king Ixlilxochiti went him one
or two better by having himself installed as “Universal Monarch” and “Lord
of All the Earth” (Davies 1980: 56). When the Mexica came along as Toltecs,
their ruler assuming the title of “Lord of Culhua,” it was not so much a play of
a symmetrical kind as what Bateson called “complementary schismogenesis,”
which works rather on claiming a status “different to and better than” the other
party.19 By virtue of such oppositions, a complex set of values may come into
operation: the way the Chichimecs are mere “barbarians” from the perspective
of the civilized Toltecs, but in their own view “hardy and great warriors” of the
hinterlands as opposed to the city-dwelling Toltec artists and craftsmen. On
the other hand, as Bateson (1935, 1958) pointed out, insofar as such contrasting
values are complementary and interdependent, the conflict between them may
reach a point of equilibrium and reciprocal exchange.20 It seems fair to say that
this is what happened in the case of the Texcoco “Chichimecs” and the Mexica
“Toltecs,” no doubt motivated by a political situation in which alliance was the
better part of valor.
I close, then, with one of the many anthropological lessons that could be
drawn from such histories of interacting and interdependent societies. Clearly,
these societies are formed in relation to one another: perhaps to the extent that
19. In other registers, the Mexica double claim of being Chichimecs and Toltecs—or
alternatively worshipers of Huitzilopochtli and Tlaloc—thus as Lords of Heaven
and Earth—would indeed symmetrically top even the Lord of All the Earth.
20. For an excellent example, see Lévi-Strauss’ (1971) study of the differentiation and
reconciliation of the Native American Mandan and Hidatsa peoples.
THE STRANGER-KINGSHIP OF THE MEXICA
<
br /> 247
the legitimacy of their own rulers depends on their foreign origins—stranger-
kingship—or that these rulers take on borrowed notions of authority that are
beyond their own political means—galactic mimesis. The lesson is that anthro-
pology has long been implicated in a major theoretical scandal, insofar as it has
been futilely engaged in various ways of explaining cultures from within, as if
they were self-fashioning, although even their differences are formed in relation-
ships to others—schismogenesis. “Human societies are never alone,” as Lévi-
Strauss (1952: 9) said, although he might have added that anthropologists—as
others in the human sciences—have generally acted as though they were. I won’t
go into the reasons, which range from the prevalence of nationalist concepts to
the limitations of ethnographic practice, hence the habitual epistemological in-
clination to know cultures as isolated monads. Suffice it to note that, with some
important exceptions such as globalization and world-systems studies, virtually
all our received paradigms of cultural order and development—functionalism,
structural-functionalism, cultural materialism, evolutionism, Marxism of base
and superstructure, the new ontology and the old ecology, patterns of culture,
even postmodern discourses, epistemes, and subjectivities—all these paradigms
suppose that the forms, relations, or configurations at issue are situated within
a more or less coherent cultural scheme, and that the articulations and dynam-
ics of that scheme are the theoretical matters at issue. French structuralism has
had an interesting history in this respect, likewise inner-directed so long it was
based on Saussurean notions of a systematic semiotic field in which “tout se
tient,” but from which it broke out in Lévi-Strauss’ intercultural permutations
of mythical structures. By and large, however, as Fredrik Barth succinctly put
it decades ago, “Practically all anthropological reasoning rests on the premise
that cultural variation is discontinuous: that there are aggregates of people who
essentially share a common culture, and the interconnected differences that dis-
tinguish such culture from others” (1969: 9).
The scandal is that this is empirically not so and evidently never has been
so. The great majority of societies known to ethnography and archaeology, as
remote in space and far back in time as we can get, are formed by their situation
within fields of cultural others. Even as Immanuel Wallerstein (1976) was de-
veloping the notion of a capitalist world-system, Stanley Tambiah had worked
out the galactic polity, and Kajsa Ekholm (1980) and Jonathan Friedman (1992)
had discovered the five-thousand-year-old history of regional core–periphery
configurations focused on dominant civilizations (cf. Ekholm and Friedman
1979). Not to forget Alfred Kroeber’s (1945) similar discussion of the Eurasian
248
ON KINGS
ecumene and its underdeveloped margins, Morton Fried’s (1975) arguments
about the relations between pristine states, secondary states, and dependent
“tribes,” and Aidan Southall’s (1988) depiction of “segmentary states.” Early on,
Kroeber (1947) and Clark Wissler (1938) were mapping Native American “cul-
ture areas” that were likewise centered on certain “dominant cultures” or marked
by “cultural climaxes.” More recently, Deborah Gewertz (1983), Anthony Forge
(1990), and Simon Harrison (1990), reporting on the peoples of the Middle
Sepik, New Guinea, and Dan Jorgensen (1990c, 1996), on the Sepik Headwa-
ters, have ethnographically recorded just such regional systems, focused on the
Iatmul and Telefol peoples, respectively. These studies show that by means of
the demonstration-effects of punctuated warfare and the demonstrated efficacy
of the spirits and rituals of the leading peoples, regional patterns of core–pe-
riphery relations analogous to the galactic polities of “high cultures” exist as
well in the tribal zone. They may even involve a degree of economic domina-
tion, insofar as the agricultural prosperity of the peripheral peoples depends
on rituals controlled by the central groups. In sum, this hierarchical system of
“interculturality,” as it may be called, is the normal state of cultural affairs in
many places. Perhaps everywhere. Certainly from my own limited knowledge
of the primary texts on Native Mesoamerica and the brilliant scholarly analyses
made of them, I suspect that such interculturality is equally a feature of Mexica
history. It seems appropriate, then, to leave the last word to the greatest student
of intercultural relations among Native American peoples, Lévi-Strauss:
It is high time that anthropology freed itself from the illusion gratuitously in-
vented by the functionalists, who mistake the practical limitations imposed on
them by the kind of studies they advocate for the absolute properties of the ob-
jects with which they are dealing. An anthropologist may confine himself for one
or more years within a small social unit, group or village, and endeavor to grasp it
as a totality, but this is no reason for imagining that the unit, at levels other than
the one at which convenience or necessity has placed him, does not merge in
varying degrees with larger entities, the existence of which remains, more often
than not, unsuspected. (1990: 609)
chapter 5
The people as nursemaids of the king
Notes on monarchs as children, women’s uprisings,
and the return of the ancestral dead in central
Madagascar
David Graeber
The institution of kingship embodies a kind of paradox. Kings are both omnipo-
tent and helpless. On the one hand, the essence of sovereignty is the sovereign’s
power to do as he likes with his subjects and their possessions, and the more ab-
solute the monarchy, the more absolute, arbitrary, and unaccountable this power
tends to be. On the other hand, kings are, in part for this very reason, dependent
on their subjects. They are fed, clothed, housed, and have their basic physical
needs attended to by those ostensibly under their power. And the more absolute
their power, the greater that dependency will also tend to be.
For those educated in European philosophy, this observation will imme-
diately call to mind Georg Wilhelm Friedrich Hegel’s master–slave dialectic,
where the conqueror, in reducing his rival to servitude, becomes dependent on
him for his means of livelihood, while the conquered at least achieves a kind of
paradoxical autonomy, and mastery, in work. But Hegel was by no means the
first to have noticed this dynamic. There have been times and places, far from
250
ON KINGS
Hegel’s Germany, where the paradox is seen as lying at the very center of the
idea of kingship; not just in the reflections of philosophers but in the ritual life
surrounding kings themselves.
In this essay I’d like to explore one of them. In the Merina kingdom of the
north central plateau of Madagascar, kings were quite often represented as in-
fants, toddlers, or petulant adolescents. They were assumed to be both willful,
diff
icult, and utterly dependent on their subjects. Framing institutions of gov-
ernment in this fashion created a peculiar moral alchemy whereby selfishness,
imperiousness, even occasional outbursts of vindictive violence, could actually
be seen as endearing, or, at the very least, as reinforcing the feeling that it was
the duty of commoners to attend to royal needs. Yet this way of imagining royal
power clearly cut two ways. For one thing, it was accompanied by a sense that,
while the living ruler could be seen as a kind of perpetual minor, it was dead
kings—the royal ancestors—who really represented mature authority. For an-
other, it gave subjects a language with which to chasten and admonish rulers,
whether in their own names or the names of those ancestors, whenever they
were seen to have gone too far.
THE PEOPLE AS NURSEMAIDS OF THE KING
251
This aspect of Merina kingship has largely been ignored in the existing
scholarly literature. But as soon as you start looking for it in the primary sources,
it’s everywhere.
It is one of the unique pleasures of studying the Merina kingdom that these
sources are so rich. After King Radama I invited foreign missionaries into his
kingdom to set up a system of primary education in 1820, literacy became wide-
spread, and the result was an unprecedented outpouring of texts, official and
unofficial—ranging from histories, compendia of customs, poetry, oratory and
folklore, including a 1,243-page collection called the Tantara ny andriana eto
Madagascar (“History of the kings of Madagascar”) which consists of a detailed
history and ethnography of nineteenth-century Merina1 society in pretty much
all of its aspects, drawn from a wide variety of authors. Infuriatingly, since most
of this material only reaches us through missionary sources, almost none of the
authors’ names have been preserved. (The Tantara, for instance, is remembered
as the work of one Père Callet, a Jesuit priest who assembled and edited the
manuscripts.) Still, combined with the detailed correspondence, legal records,
and official registers preserved in the National Archives, they provide an almost
unparalleled window on nineteenth-century Merina society.
What I’m going to do in the fol owing pages, then, is to examine how the