by Faun Rice
colonial writers assumed ought to be the case.
67. The diagram is my own schematized synthesis; needless to say almost every
detail is contested by someone or other. Delivré (1974: 77–99) provides the most
comprehensive review of sources for royal genealogies, starting with the Rabetrano
manuscript of 1842; Berg (1977) and David Rasamuel (2007: 205–19) offer
important critical reviews; I have made extensive use of Gilbert Ralaimihoatra
(1974) here as well, which uses an alternate tradition based on manuscripts also
said to go back to the 1840s, as well as Savaron (1928, 1931).
450
ON KINGS
God
Andriampenitra
Rafohy
Rangita
Andriamanelo
Antehiroka
Ralambo
Zafimamy (full name
Zafin’Andriamamilaza)
Talasora (or Andriamitondra)
1. ZanadRalambo
Andrianjaka
2. Andriandranando
3. Zanak'Ambony
Andriamasinavalona
(Andriamboninolona)
4. Zana-Tompo
(Andriantompokoindrindra)
5. Andriamasinavalona
6. Zazamarolahy
Andrianampoinimerina
7. Havan’Andriana
KEY:
Radama I
Kings
Ranavalona
Andriana
Carriers of first-fruits
Radama II
Figure 3.
There are, in fact, two royal genealogies.68 The first, the one on the left, stretches
back through an unbroken line of fathers and sons to the very earliest days of
humanity. Once upon a time, the story goes, one of God’s children descended to
the earth to play with the Vazimba in the great eastern forests, but was trapped
there after being tricked into inadvertently eating some mutton (Callet 1908:
68. Actually considerably more than two—for instance, the Zanak’Andriamamilaza,
though represented here, have their own genealogy, which claims to stretch back
to a stranger-king founder Andriantomara, purportedly from Indonesia in the
thirteenth century ad (Ramilison 1952; Rakotomalala 2011). Any number of
others had to be integrated, or effaced, in order to patch together the existing one.
NOTES ON THE POLITICS OF DIVINE KINGSHIP
451
11 n. 1). God punished the Vazimba—who are represented as being primitives,
innocent of agriculture or herding—by leaving his son to rule them, and also
provided a daughter for him to marry. The fruit of that incestuous union became
the royal line. Their tombs (still remembered in the early nineteenth century
when all this was first written down) trace a steady movement westward from
the edge of the forests to the center of the Great Plateau, which was eventually
to become the heartland of the Merina kingdom.
The royal line on the right occupied that future heartland all along. So one
would imagine that they would normally be the “Vazimba” line. And indeed
some of their descendants (the Antehiroka and Talasora in particular) are still
referred to as “Vazimba” and occasionally as “owners of the land” (Fugelstad
1982; Domenichini 1982, 2004).
There are also hints this opposition, between forest and interior, must have
once been central to royal ritual. As explained in chapter 5, in this volume,
at least since the reign of Andrianampoinimerina (1787–1810), there was an
elite group with the somewhat daunting title of velond-rai-aman-dreny (“Those
Whose Mothers and Fathers Are Still Living”) drawn from certain privileged
ancestries,69 who were charged with maintaining royal tombs, conveying first-
fruits, and presiding over key events in the life of the royal household. While
some of these were ancestries rewarded for having performed some special ser-
vice for the monarchy, any team of Velondraiamandreny had to include at least
one group that lived on the edge of the Great Forest, and one Vazimba group
from the central highlands. In the diagram above, the Andriampenitra and
Zafimamy represent the foresters, and the above-mentioned Antehiroka and
Talasora are the indigenous “owners of the land.”70
69. Most highland descent groups are localized, and in principle endogamous; they focus
on tanin- drazana (ancestral lands) containing their founder’s tomb. Maurice Bloch
(1971) proposed to call these “demes,” but the term has not been largely taken up;
in the nineteenth century they are referred to as firenena (literally, “motherhoods,”
though they are cognatic with a patrilineal bias), or foko (just “group”). Nowadays,
there does not seem to be a regular term for them at all. I’m here adopting “ancestry”
as a generic usage from Feeley-Harnik (1991).
70. These were the only four for whom clear genealogical information was available.
Sometimes it’s ambiguous: for instance, the Tahiamanangaona claimed andriana
status on the basis of their having been “companions” ( namany) of Andriandranando,
which usually means collaterals of some kind but the actual link is not recorded
(Callet 1908: 1214–215). There were in fact two groups of velond-rai-aman-
dreny, one for Antananarivo, one for Ambohimanga, these being the twin capitals
452
ON KINGS
These velond-rai-amen-dreny must be carefully distinguished from the an-
driana, which means both “king” and “those descendants of royalty who still
partake in royal status.” Andriana are divided into seven orders, the oldest being,
as the principle of horizontal sinking status would dictate, the lowest in rank.
According to traditions the rank system was reorganized three times, by three
different kings (Ralambo, Andriamasinavalona, Andrianampoinimerina) to add
new orders and push the older ones back. Many of the velond-rai-aman-dreny
insisted they were really andriana as well, or had once been, but had been de-
moted in one of the previous reorganizations.
* * *
So, originally, there would appear to have been a fairly simple opposition be-
tween invading stranger-kings and indigenous “owners of the land,” with the
andriana being the most recent descendants of the first. But at some point,
however, something happened. The pieces were rearranged. The two Vazimba
Queens, Rafohy and Rangita—remembered as local ancestors in the old sa-
cred capital of Alasora—were inserted at the end of the first genealogy, and
the three generations after them were framed as a time of great inventions,
when the divine spark already manifest in that line from its origins was seen
as bursting out in great feats of—often violent—creativity. Stories of stranger-
king invaders became stories of technologically superior creator kings routing
their enemies. Most of the major institutions of Merina society were said to
of the Merina kingdom. We have much better information for the second. The
Antananarivo group include the Zanak’Andriampenitra (forest), Antehiroka/
Zanadahy (Vazimba), Zafintsoala (probably same as Trimofoloalina, benefactors),
and Tahiamanangaona (possibly Vazimba, former andriana); the Ambohimanga
group includes the Zanak’Andrianato and Andriamamilaza (forest), Talasora/
Andriamitondra (Vazimba), Tehitany and Zanamarofatsy (benefactors). Callet’s
sources go into elaborate detail about the different sorts of forest products (eels,
hedgehogs, honey, certain species of liana, etc.) brought by the different forest
groups, but are mostly vague about how these items were employed. Insofar as
there’s a clear division of labor, the Vazimba groups appear to carry out the key
ritual actions (particularly those involving aggression: killing the sacrificial oxen,
performing the circumcision on royal children, laying down the red earth in the
tomb and placing the royal bodies in it, etc.), whereas, as noted previously, andriana
groups do all the acts of creation, fashioning, or construction (ibid.: 15, 163–65, 254,
256–62, 306–9, 316, 390, 401, 407–11, 423–24, 435, 533–35, 589–90, 632, 812–13,
1136–137, 1211–214; Cousins [1876] 1963: 44–45; Domenichini 1978).
NOTES ON THE POLITICS OF DIVINE KINGSHIP
453
have been invented by, or “appeared” under, such kings and their companions
during the three generations immediately following Rafohy and Rangita, but
kings continued to be seen as at least potential inventors. Even those lines that
broke off the royal genealogy that continued to be recognized as andriana were
those who were identified with (and often continued to maintain monopolies
on) particular technological breakthroughs: writing, metallurgy, and so forth.
Crucially, this allowed all rulers before Andriamanelo, of any royal line, to be
uniformly written off as mere primitives from an earlier historical epoch “when
Vazimba ruled the land” (Savaron 1928: 68).
We don’t know precisely when this happened, but most likely it was around
the time of that Merina monarchs began reclaiming the vast Betsimitatra
marshes surrounding Antananarivo (Raison 1972; Cabanes 1974), an enor-
mous project which created thousands of hectares of new land, and, it seems,
a reversal of the older system where kings ruled over people, but the indig-
enous tompon’tany or “masters of the land” still owned the soil. Merina kings
began referring to themselves as tompon’tany, claiming possession of all land in
the kingdom, as an extension of the elementary principle of sovereignty,71 and
groups like the Talasora and Antehiroka were stripped of any ritual role in rela-
tion to the land they might once have had and reduced to “nurturing” the royal
household and maintaining the royal dead (Fugelstad 1982: 65–70).
Whereas in Bali, whose rulers also claim divine origins, kings continually
decline from grace, creating a contradiction of vertical and horizontal sinking
status, Merina kings in this new version of dynastic history advance—which, if
nothing else, means the two principles, ascending status for the kings, descend-
ing status for everyone who splits off the royal line, are brought in consonance
with one another.
The progressive ideology helps explain the Napoleonic ambitions of
nineteenth-century Merina kings, who all aspired to become Enlightenment
71. For instance, the Tantara observes that insofar as the king is “tompony’tany,” he has ultimate sovereignty; others may possess specific portions and even sell it, but
they may not sell it to anyone who is not the king’s subject as it would bring it
out from under the king’s ultimate control (Callet 1908:365). Sovereignty was
conceived not just as the power over life and death but as the power to appropriate
and dispose of land or possessions with impunity; thus, it is said that when King
Andriamasinavalona granted a benefactor’s family permanent immunity from all
accusations of crime against persons or property, his advisor Andriamampandry
quickly intervened to point out that anyone who had that right effectively would be
king, since that is what the essence of sovereignty consists of (Kingdon 1889: 5–6).
454
ON KINGS
philosopher-kings, and how they came to be seen by their subjects above all as
playful and obstreperous children. I think it also helps explain how the logic of
the royal ancestral cult came to be popularized, culminating in the development
of the kind of spectacular mortuary ritual—the landscape dotted everywhere
with stone tombs, the periodic festivals which draw even urban professionals
to flock to the countryside to exhume the bodies of their ancestors and rewrap
them with new silk shrouds—that the highlands are famous for to this day. The
problem the Merina sovereigns faced was above all how to hold their own in
the face of the ever-burgeoning, and ever more ancient, legions of the ancestral
dead. (“Even the dead,” one Malagasy proverb goes, “desire to be more nu-
merous.”) Even after the magnificent gesture of writing them all off as primi-
tive “Vazimba,” local people seem to have maintained the cults of their bygone
monarchs’ tombs, now mixed up with an older conception of Vazimba spirits
of lost spirits of the waters and the wild, as the true “owners of the land” (e.g.,
Callet 1908: 256). With the collapse of the monarchy after the French inva-
sion of 1895, every local descent group immediately claimed tompon’tany status,
insisting they were owners of their own ancestral territories. The overwhelming
majority also insisted that their founding ancestors, or razambe, were themselves
of some kind of royal descent.
Such claims are not necessarily fabrications. Since there were so many
kings and kingdoms over such a long period, and since cognatic descent al-
lows one to trace through either male or female lines, no doubt virtually any-
one in Imerina could make such a claim on some basis or another. What did
happen to al the col ateral lineages that spun off the ruling dynasty before
Andriamanelo (apart from those that were named velond-rai-aman-dreny)?
We cannot know. The sources are focused almost exclusively on those closest
to royalty. But occasional y more obscure groups do come into focus for one
reason or another.
In 1895, for instance, shortly after French forces had seized the Malagasy
capital, there was an insurrection in the lands surrounding Arivonimamo in
which a family of Quaker missionaries were kil ed. It was spearheaded by a
very large descent group cal ed the Zanak’Antitra, and for this reason that
group—which is not even mentioned in the 1,243 pages of the Tantara ny an-
driana— came under a degree of sustained attention (Clark 1896; Renel 1920:
39, 128–29; M. Rasamuel 1947, 1948; Peetz 1951a; Daniel i 1952). Thus we
have some idea what the story of a local ancestor might look like around the
time mortuary rituals, as they have come to be practiced today, were taking
NOTES ON THE POLITICS OF DIVINE KINGSHIP
455
shape (Larson 2001). The Zanak’Antitra claimed to be descended from a line
of andriana that split off from the Zanak’Andriampenitra, forest-dwel ers of
the Ankaratra mountains, far to the south of the Merina heartland—the lat-
ter, a group which themselves claimed to have split off from the royal dynasty
many generations before (as seen in figure 3 above).72 Sometime around 1790,
they say their ancestor Andriantsihianka and his family were fo
rced to flee their
ancestral lands during a disastrous war and asked for refuge from a certain Prin-
cess Ravao. She agreed to take them in on condition he renounce his andriana
status. Later, even when her husband, king Andrianampoinimerina, offered to
restore his rank, Andriantsihianka refused, insisting that he preferred not to
rule over others.
By the 1990s, this story appears to have become a template. The vast major-
ity of rural people insisted their ancestors were, at some point, andriana. Many
insisted they still were.73 In effect, the country is now populated by the descend-
ants of dozens of little stranger-kings—with the result that that same logic of
sinking status and the burden of rivalrous, oppressive, and ever-more-numerous
ancestors that once haunted the center of power has instead been pushed off
onto just about everybody else.74 We are left with a population struggling with
the memories of their own grandparents and great-grandparents through the
very tools (adverse sacralization, effacement, etc.) once deployed against kings,
in a kind of generalized war against the dead.
72. As with all such things, the derivation is contested: Grandidier (1914: 650) accepts
it; Dez (1971b: 104) is more skeptical.
73. Where I did my fieldwork around Arivonimamo, the largest descent groups were
the Andrianetivola and Zanakantitra, both of whom claimed to have been refugees
who gave up their andriana status, the Andriamasoandro, who claim descent from a
different line of kings, and the Andriatsimihenina, who alone insisted on commoner
status, but were otherwise vague about their origins. I myself worked in a community
of Andrianamboninolona or Zanak’Ambony, who traced back to the fifth officially
ranked order of andriana. They kept their rank as they had been placed as military
colonists in the region under Andrianampoinimerina (Graeber 2007a: 99–100).
Pier Larson has collected numerous local histories from the Vakinankaratra region
which begin with refugees willingly abandoning their andriana status.
74. I am, of course, simplifying massively. Most people in the very heartland of the old
kingdom identify with hova ancestries that were closely allied with royalty; matters
are also much complicated by the presence of a large population of descendants of
slaves who have simultaneously become de facto guardians of the royal ancestors
