by Faun Rice
former kings. We have some good descriptions of what the process might look
like at its peak from the West African kingdoms of Asante, Benin, and Da-
homey (Law 1985; Rowlands 1993; Terray 1994), as well as Buganda (Ray
1991), which seem to roughly approximate, in scale and general tenor, what
has been documented archaeologically from Bronze Age China (R. Campbell
2014). As the circle of those slated for death expands from intimates, who one
can at least imagine gave up their lives voluntarily, to entire courts, to massacres
60. “Large-scale retainer sacrifices are typically witnessed when a state suddenly and
dramatically expands in geographic domain and coercive power. At such times the
conception of the ruler is ripe for reformulation” (E. F. Morris 2007: 17 n. 3).
444
ON KINGS
of hundreds or even thousands of prisoners or war, criminals, or simply random
subjects swept up on the roads, in displays of sheer arbitrary power.
The royal funerary process ended with an orgy of killing, called kiwendo, or “mass
execution,” which inaugurated the shrine of the deceased king. Such killings
occurred whenever a royal shrine was rebuilt or on the rare occasions when the
king visited such a shrine in person. In 1880, Mutesa reportedly had two thou-
sand people killed at [his father] Ssunna’s shrine after it was rebuilt. The victims,
who were peasants traveling to the capital and transporting goods in the service
of their chiefs, were captured by the royal police as they approached the narrow
bridges leading to the capital . . . . (Ray 1991: 169)
Here, sovereignty does break through its containing frames—as part of a curi-
ous double game whereby living kings, in an ostensible bid to honor their ances-
tors, actually vied to outdo them. But the fact of sovereignty exploding its limits
and challenging the dead in this way also tended to render it untenable. In many
African cases, we find kings remarking on the burden of constantly having to
display their vitality through the deaths of others.61 Mutesa, as it happened, had
no such compunctions, but as a result his memory became infamous and after
his reign the executions largely ceased.
This is what usually happens. In all the cases we know best, at least, once
matters reach such a pass, some kind of moral backlash eventually begins to set
in. Or popular unrest. Or both. This may lead to abandoning the custom entirely,
usually when converting to a world religion (e.g., Buddhism in Korea: Conte
and Kim 2016; Christianity in Buganda and Benin); it might lead to gradual
adoption of symbolic substitutes, such as the armies of terracotta soldiers in An-
yang (R. Campbell 2014); it might even lead to a kind of attenuated populariza-
tion, where broad sections of the elite adopt the practice but only in carefully
limited form. The latter appears to be the case with sáti in certain parts of India,
where the widows of high-caste men were expected to commit suicide at their
husbands’ funeral (E. J. Thompson 1928; Morris 2014: 84),62 but the fact that
61. So, for example, David Livingstone was told if the kabaka didn’t kill people, people
would think he was dead (Ray 1991: 179), and Benin’s oba told one foreign visitor
he was “sick of it all,” but felt he had no choice (Roth 1903: 66).
62. This statement is somewhat contentious. Retainer sacrifice used to be referred to in
the archaeological literature as “sáti-burials” until Trigger (1969: 257) pointed out
there was no evidence for retainer sacrifice in India. Still, the fact that the practice
NOTES ON THE POLITICS OF DIVINE KINGSHIP
445
the practice spread from rulers’ households via warrior castes is certainly sugges-
tive. It also brings home the deep logic of such sacrificial projects: to telegraph,
as clearly as possible, that certain lives have value only for the sake of others.
Foreign observers often expressed shock at the willingness of many wives, serv-
ants, and other intimates of dead kings and grandees to voluntarily follow them
to the grave (though, invariably, some were more enthusiastic than others). The
high-caste Hindu household where the wife is taught to address and treat her
husband like a god, and to throw herself on his pyre, is just a more explicit form
of the same logic of self-sacrifice that expects widows in Mediterranean coun-
tries to spend the rest of their days wearing black in mourning (or, in parts of
India where sáti is not practiced, white in mourning); but at the same it is also a
microcosmic version of the patrimonial kingdom, which, as described by Hocart
(1950), is itself seen as a giant household where each social group ultimate ex-
ists primarily for its allotted role in feeding, maintaining, and deifying the king.
One might go further. Are not these mass ritual killings—especially those
that ensue when the violence explodes the framework of the royal household
and does become veritable civicide (Feeley-Harnik 1985: 277)—moments
when the contradiction between two notions of the relations of king and peo-
ple, one, the notion of kingdom as household, the other, of the constitutive war,
is actually exposed? As always, the sheer arbitrariness, the lack of meaning63
in the selection of victims—who are often swept up entirely at random—is
itself a way of conveying the absolute nature of royal power. If one cannot kill
everyone, the closest one can come is to demonstrate one might kill any one.
This remains true when the massacres are specifically intended to appease royal
ancestors (as in Benin), when they are sacrifices to gods, but not to ancestors (as
among the Mexica), when they are directed against witches (as in Madagascar:
S. Ellis 2002), or, finally, when they serve no purpose other than to demonstrate
the king’s absolute “power of life and death over his subjects” (as in Ganda
coronation rituals: Mair 1934: 179). Whatever the excuse, the same logic of es-
calation seemed to apply—likely leading to the eventual abandonment of mass
began with rulers and then spread via the warrior castes, and foreign accounts of
hundreds of women sacrificing themselves at royal funerals (e.g., Barbosa [1518]
1918: 213–14), suggest something on the scale of the most dramatic examples of
retainer sacrifice elsewhere.
63. I am using “meaning” in the hermeneutic sense of intentionality behind a statement
or an act.
446
ON KINGS
killing for fear that the contradictions of sovereignty might eventually destroy
the kingdom itself.64
4. Reversing the direction of history
Just about all of these expedients, then, are fraught with difficulties, or likely to
self-destruct. There is one final approach. One can challenge the logic of sinking
status directly, by reframing history not as a story of inevitable decline, but as
one of incremental progress.
We are used to assuming that the idea of progress is a recent innovation,
and that all “traditional” societies (i.e., all societies up until, say, Renaissance or
even perhaps Enlightenment Europe) assumed, instead, that they descended
from gods rather than having evolved from savages. In fact, a significant nu
mber
of human societies seemed to have held both positions at the same time. As
Arthur Lovejoy exhaustively documented for Greco-Roman antiquity, it was
nearly universally assumed, by most ancient authors, that humans once lived
in caves and subsisted on nuts and berries, before the discovery of the arts and
sciences brought about urban civilization (Lovejoy and Boas 1935; cf. Adelstein
1967; Nisbet 1980). What they differed on—and here they often differed quite
sharply—was not the reality of progress, but its moral significance: whether
the earliest days of humanity should be considered a Golden Age, or a time of
benighted savagery. Advocates of “Primitivist” and “Anti-Primitivist” positions
continued to debate the question until Christianity succeeded in temporarily
settling matters in favor of Eden for roughly a millennium. (Then, of course, it
started up again much as before.)
These debates are relevant because very often the inventions or discoveries
that made civilized life possible were ascribed to kings. Or gods; but for this
very reason, the two categories, kings and gods, often came to overlap. Hecat-
aeus represented the Egyptian gods as great inventors, made kings for their
creations: of writing (Thoth), agriculture (Isis), viniculture (Osiris), and so on
(Diodorus Siculus 1.13.3, 1.14–16; 1.43. 6). Later Euhemerus was to turn his
account into a general theory, arguing that all stories about gods were really
memories of kings, queens, and other remarkable mortals, and it became regular
64. One might speculate that the common folktale motifs of wicked kings cursed with
insatiable demands for human flesh or blood, of whom the prototype perhaps is the
Persian Zahhak, might reflect to some degree on this structural condition.
NOTES ON THE POLITICS OF DIVINE KINGSHIP
447
practice, even among those who were not outright Euhemerists, to ascribe dis-
coveries to rulers who had gone on to become deities: Semiramis, for instance,
was credited by Pliny with the invention of both weaving (ibid.: 7.417) and
certain types of long-hulled ships (ibid.: 7.57).
China around the same time witnessed similar debates. There, too, was the
assumption of a gradual invention of arts and techniques, and the “sages” who
made these inventions and discoveries were frequently represented as monarchs.
(The Yellow Emperor, for example, was held to be have been personally respon-
sible for the invention of house-building and weaving, and his wife invented
silk.) There, too, a lively controversy about the moral status of technological
progress ensued, with Mohists seeing technological and social invention as a
rise from savagery, Taoists as a fall from a Golden Age, and Confucians taking
a variety of nuanced positions in between (Needham 1954: 51–54; J. Levi 1977;
Puett 2002).
If all one is doing here is taking stories of the creation of cultural institutions
by primordial gods, and transposing them to stories of their creation by primor-
dial kings, the ramifications might not be particularly profound. It still leaves
kings of the present day very much in the shadow of their ancestors. In fact, if
one represents the founder of one’s kingdom also as the inventor of farming, or
metallurgy, or music, competition would seem absolutely hopeless. Still, if one
sees history, instead, as a gradual and ongoing series of discoveries and inven-
tions—as some Hellenistic Greek and later Chinese writers did do—then this
at least allows for the possibility of competition, even the possibility of intro-
ducing revolutionary innovations in the present day. Perhaps the best way to see
it is this: a monarch who considers himself one of a long line of inventors can
treat the principle of sovereignty, which allows him to step outside traditional
structures and institutions, in the same way as that by which he can step outside
law and morality, to make himself a kind of internal stranger-king, capable of
injecting new infusions of creative power to disrupt existing traditions from
within. As a result—and this is what is crucial in this context—it allows a king
to both identify himself with an ancient tradition (of kings as innovators) and at
the same time to assert his ultimate superiority to them, by laying claim to the
cumulative legacy of all their innovations, and his own besides.
I’ve already noted, in chapter 5, that the history of the Merina kingdom was
conceived along such lines. The Tantara ny andriana, a twelve-hundred-page
history and ethnography of Imerina written in the 1860s or 1870s by Merina
authors and assembled by a Jesuit missionary named Callet (1908), represents
448
ON KINGS
the history of the kingdom as the progressive invention, discovery, or appear-
ance of key institutions by successive Merina kings (the terminology used seems
to be intentionally vague as to which):
R5 Andriamoramorana— division of game, basic principles of hierarchy.
R5 Andriandranolava— political oratory.
R5 Rafohy, Rangita— astrology, first-fruits ( santatra) rituals.
R5 Andriamanelo— iron weapons, metallurgy, pottery, canoes, circumcision ritu-
als, money and commerce.
R5 Ralambo— domestication of cattle, sheep, divination, medicine, protective
talismans ( sampy), marriage customs, New Year’s festival.
R5 Andrianjaka— customs of burial and mourning, royal ancestor cult.
R5 ( Andriantompokoindrindra— ancestor of noble order, did not reign: writing.)
R5 ( Andriandranandro— ancestor of noble order, did not reign: muskets.)
R5 Andriantsitakatrandriana— riziculture, irrigation, court etiquette.
R5 Andriamasinavalona— legal principles, poison oracle, slavery, additional
marriage customs (divorce, polygyny).
What’s crucial for our purposes is that where in most cases of inventor kings
(such as the Chinese, Vietnamese, Javanese, Persian, Inka, or Mexica tradi-
tions), it’s the earliest kings who discover the most important principles and
institutions, here it’s the three kings in the middle, Andriamanelo, Ralambo,
and Andrianjaka, who are the most creative. These three are in fact represented
as marking a dramatic break with those who came before, who are generi-
cal y referred to as “Vazimba,” fundamental y uncivilized beings familiar with
magic and certain elementary social forms, but innocent of metal urgy and
agriculture.
There has been a great deal of debate about the exact meaning of this word
“Vazimba,” which nowadays can be used to refer either to an ancestor whose
body was never properly buried or whose descendants have forgotten them,
and whose spirit thus lurks in wild and watery places, or, in many oral his-
tories, to an aboriginal population driven out by the country’s current inhab-
itants. When early missionaries heard stories about Vazimba, they inevitable
assumed they represented some kind of primitive “race,” possibly pygmies, who
had been driven out by the country’s current “hova” inhabitants, who they as-
sumed to be a wave of later immigrants from Malaysia. (“Hova” in fact just
means “commoner.”)
NOTES ON THE POL
ITICS OF DIVINE KINGSHIP
449
This was clearly not the case, and as a result, many contemporary historians
(e.g., Berg 1977, 1980; cf. Dez 1971a; Domenichini 2007) have come to dis-
count the stories entirely, suggesting, for instance, that when oral histories speak
of earlier rulers as “Vazimba,” they merely mean that they were buried in lakes
or their bodies were otherwise lost. But the matter is slightly more complicated.
On the one hand, numerous stories, often from quite early on, actually do speak
of wars between hova and Vazimba, with the former, led by King Andriamanelo,
taking advantage of their newly invented iron-tipped spears to put the latter to
rout and drive them fleeing to the west (W. Ellis 1838, II: Callet 1908; Savaron
1928, 1931; G. Ralaimihoatra 1973; Raombana 1980). On the other hand, they
also insist that Andriamanelo was himself the son one of the last two Vazimba
Queens—sources differ as to which—who bear the intentionally unappealing
names of Rafohy and Rangita (“Short” and “Frizzy”).65
There is no space here to go into the details—people have spent lifetimes
trying to figure these stories out, there are endless traditions with endless sub-
tleties of interpretation—but one thing seems abundantly clear: we are dealing
here with a classic stranger-king narrative that’s been rewritten. For instance,
the father of Andriamanelo the great inventor is either not mentioned at all,
or treated as completely insignificant.66 He is spoken of as if the only thing
important about him is that he was the child of Rafohy or Rangita (the sources
differ as to which). If we consult figure 3, however, it becomes clear what’s really
happening.67
65. In fact, our earliest printed source, William Ellis (1838, II: 117) even says
Andriamanelo himself was a Vazimba, presumably because his mother was. If
nothing else, this shows how fluid the categories were.
66. It is given either as the otherwise unknown Ramanahimanjaka (Callet 1908: 9) or
equally unknown Manelobe (Jully 1898), but the overwhelming majority of sources
are content simply not to bring up the question of the king’s father at all. Jully
attempted to link “Manelobe,” which seems a made-up name, to the Zafindraminia,
but these do not seem to be based on anything in the Malagasy accounts, just what
