¹ For recent statements and variations on this theme, see Linton, R., ‘The Tanala of Madagascar’, in Kardiner, A., The Individual and his Society (New York, 1939), 282–6Google Scholar; Isnard, H., ‘Les bases géographiques de la monarchie hova’, in Eventail de l'histoire vivante: Hommage à Lucien Febvre (Paris, 1953), 195–206Google Scholar; Cabanes, E., ‘Evolution des formes sociales de la production agricole dans la plaine de Tananarive’, Cahiers du Centre d'Etudes des Coutumes [Tananarive], x (1974), 47–60Google Scholar; Fauroux, E., ‘Dynamisme précoloniaux et transformations actuelles d'une communauté villageoise du Vakinankaratra’, Cahiers du Centre d'Etudes des Coutumes [Tananarive], x (1974), 61–91Google Scholar; Berg, G. M., ‘Historical Traditions and the Foundation of Monarchy in Imerina’ (Ph.D. University of California, Berkeley, 1975, University Microfilms International, no. 76–15, III), 92–134Google Scholar; Marks, Shula and Gray, Richard, ‘Southern Africa and Madagascar’, in Gray, R. (ed.), The Cambridge History of Africa, vol. 4 (1975), c. 1600 to c. 1790, 468Google Scholar; Bloch, M., ‘Property and the end of affinity’, in Bloch, , ed., Marxist Analysis and Social Anthropology (London, 1975), 206–12Google Scholar, and ‘The Disconnection between Power and Rank as a Process: an outline of the development of kingdoms in Central Madagascar’, Archives Européennes de Sociologie, xviii, i (1977), 109–18.Google Scholar

² Berg, , ‘Historical Traditions…’, 222–51.Google Scholar For a fresh approach to archaeological evidence, see the various works of Wright, Henry T. and Kus, S. to appear in Taloha and in Kent, R., ed. Madagascar in History (Albany, Calif., 1979), 1–31.Google Scholar

³ The most valuable approach to the analysis of oral traditions in Imerina is Delivré, A., L'histoire des rois d'Imerina. Interprétation d'une tradition orale (Paris, 1974)Google Scholar. See also Berg, G., ‘The Myth of Racial Strife and Merina Kinglists: The Transformation of Texts’, History in Africa, iv (1977), 1–30CrossRefGoogle Scholar, and ‘Some Words about Merina Historical Literature’, in Miller, Joseph C., ed., The African Past Speaks (Folkestone, Kent, 1980), 221–39.Google Scholar

⁴ On the coastal trade, see Filliot, J.-M., La traite des esclaves vers les Mascareignes au XVIII^e siècle (Paris, 1974).Google Scholar See also Thompson, A., ‘The Role of Firearms and the Development of Military Techniques in Merina Warfare, c. 1785–1828’, Rev. franç d'hist. d'Outre-Mer, lxi (1974), 417–35CrossRefGoogle Scholar, who argues against the idea that the introduction of firearms into Imerina was central to state formation (419–21).

⁵ Bloch, , ‘Property…’, 203–28Google Scholar, and ‘The disconnection’, 107–48; and Kottak, C., ‘The Process of State formation in Madagascar’, American Ethnologist (02. 1977), 136–55CrossRefGoogle Scholar, which concentrates on the Betsileo kingdoms to the south of Imerina.

⁶ The two pioneering efforts in this field are Raison, J.-P., ‘Utilisation du sol et organisation de l'espace en Imerina ancienne’, Tany Malagasy, xii (1972), 97–121Google Scholar; and Dez, J., ‘Elements pour une étude de l'économie agro-sylvo-pastoral de l'Imerina ancienne’, Tany Malagasy, vii (1970), 11–60.Google Scholar

⁷ Similarly, , a recent number of African Economic History (no. 7, 1979)Google Scholar, devoted entirely to the history of agriculture and fishing in Central Africa, emphasizes developments in technique rather than the relation of those developments to changes in social organization because the history of agrarian techniques has been largely neglected.

⁸ de Flacourt, E., Histoire de la Grande Ile Madagascar, 2nd ed. (Paris, 1661)Google Scholar, reprinted in A., and Grandidier, G., eds., Collection des ouvrages anciens concernant Madagascar [COACM], viii (Paris, 1913), 22–3, 36–7.Google Scholar He speaks of ‘Vohits anghombes’ (vohitsomby, cattle mountains) which refers to the highland area just south of Imerina in the Isandra and Illangina provinces of the Betsileo at 22^o 16' S. and 44^o 42' E. See Grandidier, A., Géographic de Madagascar (Paris, 1892), 183Google Scholar n. 1, 188 n. I. The earliest written reference to contact between Imerina and the coasts is Mariano, L., ‘Relation… 1613–1614’, in CO ACM, 11 (Paris, 1904), 14, 40Google Scholar, which reports that Merina merchants sold slaves at Boina and other West Coast ports. For Merina silk and iron sold on the southwest coast, see Drury, R., Madagascar or Robert Drury's Journal during Fifteen Years Captivity on that Island (1792), ed. Capt. Oliver, Pasfield (London, 1897), 277.Google Scholar The first definitive report of rice and cattle in Imerina proper is Parat, Mémoire à Ponchartrain (19 Sept. 1714), B.N., Nouv. acq. 9344, fol. 26 bis.

⁹ Mayeur, N., ‘Voyage au pays d'Ancove, autrement dit des Hovas ou Amboilambes, dans l'interieur des terres [1977]’, ed. Froberville, (1809)Google Scholar, British Museum Add. MS. 18128, published in Bulletin de l'Académie malgache [BAM], xii, i (1913), 158–60Google Scholar; Mayeur, , ‘Voyage au pays d'Ancove par le pays des Baizangouzangous [1785]’, ed. Dumaine, and Froberville, (1809)Google Scholar, B.M. Add. MS. 18128, published in BAM xii, ii (1913), 32–6Google Scholar; LaSalle, , ‘Notes sur Madagascar [1787–8], prises sous la dictée de M. LaSalle’ (ed. 1816), B.M. Add. MS 18135, fols. 127–8Google Scholar; B^myHugon, , ‘Apperçû de mon dernier voyage à Ancôva de l'an 1808’, B.M. Add. MS 18137, fols. 14–18Google Scholar; Dumaine, , ‘Voyage ay pays d'Ancaye autrement dit de Bezounzouns, 1790’, in Annates des Voyages, xi (1810), 179–80Google Scholar and B.M. Add. Ms 18128, fol. 135; Chapelier, L. A., ‘Lettres à Léger’, B.M. Add. MS 18133, no. 10 (15 Messidor 1804), fol. 29, no. 12 (7 Messidor, 1804)Google Scholar, fols. 34–8, no. 13 (22 Prairial 1804), fols. 39–49; LaSalle, , ‘Mémoire sur Madagascar en 1787’ (ed. 1797), Notes, Reconnaissances et Explorations, iii (1898), 175.Google Scholar

¹⁰ Mayeur, , ‘Voyage… 1777’, BAM, 158–60Google Scholar, and ‘Voyage… 1785’, BAM, 36; and Dumaine, , ‘Voyage… 1790’, 179Google Scholar; LaSalle, ‘Note… 1787’, fol. 127.

¹¹ M. de Faye, ‘Mémoire sur l'état présent de l'lle Dauphine (10 févr. 1668)’, Archives nationales de France [AN], Fonds des Colonies, C^5A, 1, no. 19, fols. 3–6. De Faye was the director of the East India Company at Fort Dauphin.

¹² Lack of knowledge about the relationships between local markets specializing in agrarian products and the export trade to the coasts remains the most important gap in eighteenth-century Merina historiography. Mayeur noted the pervasiveness of local markets and claimed that slaves were the major export (‘Voyage… 1777’, B.M. Add. MS 18128, fol. 89), and Dumaine reported that the Bezanozano and Betanimena acted as middlemen in the slave trade between Imerina and the coasts (Voyage… 1790, B.M. Add. MS 18128, fol 135v) and bought all their iron implements from the Merina (op. cit. fol. 136r). By 1800, it seems that rice as well was exported. See Lebel, ‘Exposé sur quelques parties de l'Isle de Madagascar (?1816)’, B.M. Add. MS 18135, fols. 204v–205r.

¹³ Tantara ny Andriana eto Madagascar [TA], ed. Callet, (Tananarive, 1873–1902)Google Scholar, reprinted and edited by the Académie malagache in 2 vols. (Tananarive, 1908), 1, 12, 14–15, 72, 147, 237. For a summary of ancient agrarian production see Raison, ‘Utilisation…’, Dez, ‘Elements…’, Kent, R., Early Kingdoms in Madagascar, 1500–1700 (N.Y. 1970), 213–19Google Scholar, and Herbert, C., ‘Rice and Rice Culture in Madagascar’, Antananarivo Annual [AA], iii (1888), 479–81.Google Scholar The history of manioc in Imerina remains a mystery, but it seems that manioc was not among the earliest food crops, and was introduced some time in the eighteenth century. Mayeur does not mention it in his 1777 list of Merina produce, but does in 1785 (‘Voyage… 1785’, 35). By 1808, Hugon reported that manioc cultivation was widespread (‘Apperçu… 1808’, fol. 14). For two conflicting views see Kent, R., ‘Note sur l'introduction et la propagation du manioc à Madagascar’, Tany malagasy [TM] v (1969), 177–83Google Scholar; and Raison, J.-P., ‘L'Introduction du manioc à Madagascar’, TM, xiii (1972), 223–8.Google Scholar

¹⁴ TA, 243, 580; Dez., ‘Elements…’, 21–2.Google Scholar

¹⁵ TA, 12: ‘…ala maizina avokoa izao tontolo izao’, and TA, 147: ‘Ala daholo ny iva, ny avo koa misy ala koa…’ (Forest covered the heights as well as the lowlands). For the possible symbolic significance of such statements about ancient forests, see Berg, , ‘Some Words…’, 7–8.Google Scholar

¹⁶ M. Bloch, personal communication, reports that these stories are prevalent in northern Imerina, though no tree trunks have actually been found.

¹⁷ Mayeur, , ‘Voyage… 1777’, 160 and ‘Voyage… 1785’, 34Google Scholar; LaSalle, ‘Notes…1787’, fol. 127.

¹⁸ Fressange, J. B., ‘Voyage à Madagascar en 1802–1803’, Ann. des Voyages, 11 (1808), 22–3.Google Scholar

¹⁹ Hugon, ‘Appençu…’, fol. 17; David Jones, Journal (Sept.–Oct. 1820), London Missionary Society – Journals [LMS-J], I/Ia and (1823), LMS-J. I/7, 33; J. Jeffreys, Journal (May–June 1822), LMS-J, I/5; Ellis, W., Three Visits to Madagascar (London, 1859), 306–7Google Scholar; Oliver, S.P., Madagascar (London, 1862), 39–43Google Scholar; ‘Manuscrits des Ombiasy’, ed. Vérin, , TM, v (1969), 98.Google Scholar For highland deforestation outside Imerina, see de la Bathie, H. Perrier, ‘Le Tsaratanana, l'Ankaratra, et l'Andringitra’, Mémoires de l'Academie malgache [MAM], iii (1927), 7–8, 16, 21, 26–7Google Scholar; and Marchal, J.-Y., La colonisation agricole au Moyen-Ouest malgache. La petite région d'Ambohimanambola (Souspréfecture de Betafo), (Paris, 1974), 16–17Google Scholar.

²⁰ Galtie, L., ‘Laculture du riza à Madagascar’, Bull. écon. Madagascar, xii (1912), 28–9Google Scholar; Humbert, H., ‘La destruction d'une flore insulaire par le feu’, MAM, v (1927), 9–11Google Scholar; Raison, , ‘Utilisation…’, 105–6Google Scholar; Dez, , ‘Elements…’, 22–4Google Scholar. On the other hand, Gautier claims that Imerina has always been bare of forests, though his only historical source is Mayeur's account of 1777. See Gautier, E., Madagascar, essai de géographie physique (Paris, 1902), 261.Google Scholar A recent discussion of the problem has concluded that deforestation was well under way before the ‘historical period’, though it may nevertheless have been caused by human action rather than geologic processes. See Battistini, R. and Verin, R., ‘Man and the Environment in Madagascar’, in Battistini, R. and Richard-Vindard, G., eds., Biogeography and Ecology in Madagascar (The Hague, 1972), 232–5.CrossRefGoogle Scholar

²¹ Oral tradition, exemplified by TA, p. 276, says as much and specifically mentions population growth as a spur to the development of irrigated riziculture: ‘Raha tsy misy hanina, dia tsy ho velona ny tany aloha, ka nisy hanina, k'izany no nahavelona azy, ary ny tanimbary sy ny asa no nahavelona azy…niaraka tamy ny olona ela ny nisehoany ny tanimbary sy ny asa-tanimbary.’ (Initially, when there was no need to eat, the land lay fallow, but when there was a need to eat, rice paddies and rice growing techniques arose… long ago rice paddies and rice growing techniques accompanied the people.’)

²² Geertz, C., Agricultural Involution (Berkeley, 1963), 15–37.Google Scholar

²³ Ibid. 25–8.

²⁴ Dez, , ‘Elements…’, 22.Google Scholar

²⁵ Dez, ibid. 23–5; Raison, ‘Utilisation…’, 106–8.

²⁶ TA recalls the building of dikes for rice paddies during the most ancient epoch (vasimba) preceding the foundation of monarchy. See TA, 15, 28, 275–6, 443–4. Since dike building represents a rather late stage in the development of riziculture technique, rice must have been known even earlier. Traditions about Andrianakotrina (see below) suggest that rice was initially broadcast into marshes, and Linton reports that marsh rice is known as vazimba-rice, that is, rice of the most ancient epoch. See Linton, R., ‘Rice, a Malagasy Tradition’, American Anthropologist, xxix (1927), 654.CrossRefGoogle Scholar Moreover, marsh and forest rice are well known in other areas of Madagascar and were grown in some parts of Imerina in the late nineteenth century. See Galtie, H., ‘La culture…’, 13–15, 27–9Google Scholar; Herbert, C., ‘Rice…’, 482–4; ‘Rapport du Chef de la station Nanisana’, Bull. écon. Mada. (1909), 191Google Scholar; and Dufournet, R., ‘Observations’, TM, vi (1969), 104Google Scholar, who notes that forest, marsh, and paddy rice are often not distinguishable biologically; only the technique of planting differs.

²⁷ Raison, , ‘Utilisation…’, 102.Google Scholar

²⁸ Nevertheless, three reservations to this conclusion should be noted. First, the archaeological record on this matter is ambiguous (see Berg, , ‘Historical Traditions’, 222–51Google Scholar). Second, the forest symbolism in oral literature may be ahistoric (ibid. 100, 104–5); and finally deforestation in many areas of the highlands may well be the work of nature rather than man. (Battistini, R. and Verin, R., ‘Man…’, 323–35.)Google Scholar

²⁹ Each version is followed by its probable provenance. Provenance has been determined by a method using locative prepositions first developed by Delivré, A., Histoire…, 47–8Google Scholar, and I have noted elsewhere ( Berg, , ‘Historical Traditions’, 296–7)Google Scholar the difficulties of applying this method to Merina historical literature. I, O., TA, 14–15.Google Scholar Prov.: (?) ‘aty Ankaratra’, ‘aty ambony’ [-plateau], (contd. p. 297).

O. II, TA, 15: ‘aty Alasora’.

O. III, TA, 15, n. 1: ‘Tatsy Ambariarivo’.

O. IV, TA, 632: (?) ‘Tatsy Ambariarivo’.

O. V, R. Linton, ‘Rice…’, 654: ‘Hova’.

O. VI, TA, n. 1, 20–21: ‘filaza ny Tatsimo’ [story of Southerners].

O. VII. TA, 401: Ambohinanjakana, East of Ambohimanga.

O. VIII, Dahle, Lars, ed., Specimens of Malagasy Folk-lore (Tananarive, 1877), 302Google Scholar: A variant of O. I and O. II is found in Razafimino, G., La Signification Religieuse du Fandroana (Tananarive, 1924), 50.Google Scholar

³⁰ TA, 632: ‘…nanaovana farihy kely… ’.

³¹ However, one version (O. VI) mentions the importance of well-watered soil for rice cultivation but indicates neither a northern nor southern origin for wet rice cultivation.

³² That a symbolic element is present in these stories should not invalidate them as historical sources. Though peaks are often said to be the point of origin of major innovations, it is historically significant that the mountains of these stories are in the south, though peaks abound in northern and western Imerina, and this can be taken as evidence of southern origin. To be sure, rice cannot be grown on the cold southern mountain peaks of the Ankaratra so I assume that traditions referring to these peaks reflect historical events which took place on the very fertile plains of the Vakinankaratra which border the Ankaratra mountains just to the east. In fact, Ambohitrakoholahy, one peak mentioned often as the origin of rice, is also the name of several villages in the Vakinankaratra plain.

³³ Version O. VII: ‘Ary teo ambany avaratr ‘Ambohinanjakana amy ny loharano eo ny nisehoany ny vary voalohany’.

³⁴ See Berg, , ‘Some words…’, pp. 224–5.Google Scholar

³⁵ Herbert, C., ‘Rice… ’, 479–86Google Scholar; ‘Ms. Ombiasy’, [c. 1853], in TM, vi (1969), 95–103Google Scholar; Baron, R., ‘Notes on the Economic Plants of Madagascar’, AA (1898), 218–23Google Scholar; Galtie, , ‘La culture du riz’, 16–26Google Scholar; Fauchère, A., ‘La culture du riz à Madagascar’, Riz et Riziculture [Paris], I, i (1925), 22–32Google Scholar; Razahamihanta, M., ‘Le riz à Madagascar’, Revue de Madagascar, nos. 39–40 (1967), 87–98.Google Scholar For general problems of rice growing see Grist, D. H., Rice (London, 1959), 28–49Google Scholar; Matsuo, T., Rice Culture in Japan (Tokyo, 1955), 109–12Google Scholar; Pelzer, K. J., Pioneer Settlement in the Asiatic Tropics (New York, 1945), 47–51Google Scholar; Schrieke, B., Indonesian Sociological Studies (The Hague, 1957), 11, 102–4, 288–301.Google Scholar

³⁶ Galtie, , ‘La culture du riz…’, 16–17, 19–21 and 24Google Scholar, and Raison, , ‘Utilisation…’ (1972), 110–11Google Scholar. For a general discussion of the labour involved in paddy rice cultivation see Geertz, , ‘Agricultural…’, 28–32.Google Scholar

³⁷ Galtie, , ‘La culture du riz…’, 25–6.Google Scholar Second-season rice produces 1,000–2,500 kilos/hectare while first-season rice produces 2,500–3,000 kilos/hectare.

³⁸ Versions of Rasoalao and Rapeto followed by provenance determined by locative prepositions: RR. I. TA, 15–17: near Ampanibe and Lohalambo, south of Ambohijanaka.

RR. II. TA, 16, n. 1: Vakinombifotsy deme on the river which bears its name between Tananarive and Lake Itasy during the reign of Ranavelona I.

RR. III. TA, n. 2, 16–17: Ambohimiangara, northeast of Lake Itasy.

RR. IV. TA, 239: Imerina.

RR. V. TA, 632: Ankaratra (?)

RR. VI. TA, 698–9: Imerina.

RR. VII. TA, 699, n. 1: at the tomb of Rasoalao. RR. I claims that the location of Rasoalao's tomb is unknown.

RR. VIII. Dahle, L., Specimens of Malagasy Folk-lore (1877), p. 302Google Scholar:?

RR. IX. Renel, C., ‘Anciennes religions de Madagascar. Ancêtres et Dieu X’, BAM (1920–1921), 49Google Scholar; Renel claims that this version is independent of TA.

³⁹ J.-P. Raison, personal communication.

⁴⁰ O. VIII is an unusual variant because it alone mentions specifically both rice origins and Rapeto. Therefore, it belongs to both sets and is labelled O. VIII/RR. VIII. It should not receive much weight as it is extremely brief and of unknown provenance though Imamo, in the west, seems likely.

⁴¹ RR. III is the only exception; it notes only that Rasoalao and Rapeto were of foreign origin.

⁴² Only one variant, RR. I from Ampanibe, unites northward with westward movement.

⁴³ So it is that traditions name Fenitra as the town where rice first flourished, and Andriampenitra (Andriana + fenitra) the town's ruler. Fenitra means figuratively ‘first beginnings’, ‘original’, ‘foundation’, and refers to the four strips of reed used to start to weaving of baskets. See ‘fenitra’ in Richardson, , New Malagasy English Dictionary (Antananarivo, 1885).Google Scholar

⁴⁴ For the movement west in the late seventeenth and early eighteenth centuries see TA, 238–239 (Vonizongo) and 571 (Imamo) and Dez's summary, ‘Elements…’, 12–14. RR. I is an exception since it covers not only Ankaratra origins but also the movement west. This combination may indicate that the tradition, from northwest Marovatana, originates from a group of recent immigrants from the south having knowledge of both southern and central Merina traditions. RR. V is the only RR. variant that connects Rasoalao with the descendants of Andriampenitra.

⁴⁵ J.-P. Raison, personal communication.

⁴⁶ Mayeur, ‘Voyage… 1777’, B.M. Add. MS 18128, fol. 89r; Flacourt, , ‘Histoire…’, COACM viii, 36–7, 151, 153, 159–60Google Scholar; Martin, F., ‘Memoire concernant l'île de Madagascar, 1665–1668’, in COACM, ix, 549Google Scholar for canal irrigation in the Lake Alaotra area. For horaka and paddy cultivation on the east coast see Chapelier, Lettre No. 40 (4 nivose 1804), B.M. Add. MS 18133, fols. 179–87; Fragments of a book of travels in Africa treating of the manners and habits of nations in the island of Madagascar (early eighteenth century?), B.M. Sloane MS 3392, fol. 886; ‘Mémoire sur l'état présent de l'Ile Dauphine (10 février 1668)’, AN, C^5A/I, no. 19, p. 4; Letellier, ‘Mémoire (5 juillet 1777)’, AN, C^5A/8 no. 52, fol. iv; de la Serre, Le Chevalier, ‘Journal du Voyage fait a Madagascar (1777)’Google Scholar, Service historique de l'Armeé, MR. 1676, 29, 76–7.

⁴⁷ For a corrective see Vansina, Jan, Children of Woot (Madison, 1978), 34–40.Google Scholar

⁴⁸ Rice origin traditions in the Vakinankaratra area prominently feature the founders of social groups: Andrianadranoala, Andrianjokotanora, and Faralahintaontany (O. I, O. II), and it was during the reign of Faralahintaontany that God's child is said to have descended from heaven with rice. However, the tradition relating Andrianakotrina's move to the north and his introduction of irrigation techniques places these events later in time than the founding of social groups in Vakinankaratra, namely during the reign of the Ralambo-figure, the second kind of Imerina (O. III).

⁴⁹ Berg, , ‘Historical Traditions…’, 117–18.Google Scholar

⁵⁰ For a discussion on ascending and descending anachronism in Merina traditions, see Delivré, , L'histoire…, 185–214.Google Scholar

⁵¹ TA, 238, 276. It is interesting to note that dike-building moves from Alasora in central Imerina to the west and corroborates the direction of movement culled from the stories of Rasoalao and Rapeto.

⁵² TA, 275–6. He not only built dikes (fefiloha) but also was the first to dredge rivers and reconstruct their banks to form shallow inlets (sakeli-drano) for use as rice paddies.

⁵³ See Joseph C. Miller, ‘Introduction’, in The African Past…, passim.

⁵⁴ Berg, , ‘Historical Traditions…’, 222–51.Google Scholar See also Wright, H. T. and Kus, S., ‘An Archaeological Reconnaissance of Ancient Imerina’, in Kent, ed., Madagascar….Google Scholar

⁵⁵ Malzac, , Histoire du Royaume Hova (Tananarive, 1912), 31Google Scholar; and Abinal, , Vingt ans a Madagascar (Paris, 1885), 54–5Google Scholar from which Delivré, , L'histoire…, 233–4Google Scholar derives his chronology. For a comparison of various chronologies see Kent, , Early Kingdoms…, 220Google Scholar, n. 115.

⁵⁶ Mille, A., ‘Anciens horizons d'Ankatso’, Taloha, iv (1971), 120.Google Scholar

⁵⁷ Battistini, and Vérin, , ‘Man…’, 311–76Google Scholar; Raison, , ‘Utilisation…’, 105–8; Dez, ‘Elements…’, 21–4.Google Scholar

⁵⁸ Berg, ‘Some words… ’, passim.

⁵⁹ TA, 16, 28, 238, 276, 443–4.

⁶⁰ TA, 275–6.

⁶¹ Dez., ‘Elements…’, 37–41.Google Scholar

⁶² TA, 276: ‘…niaraka tamy ny olona ela ny nisehoany ny tanimbary sy ny asa-tanimbary… see note 21.

⁶³ Dez, , ‘Elements… ’, 28–30.Google Scholar

⁶⁴ TA, 276: ‘…ary dia no zarai'ny any ny vahoaka hatao vary raha vita [ny fefiloha]: Fa amy ny andriana ny tany hatr'amy n'izay nisehoany ny andriana.’

⁶⁵ Berg, , ‘Royal Authority and the Protector System in Nineteenth-century Imerina’, in Kent, , ed., Madagascar…, 102–22.Google Scholar

⁶⁶ See Wittfogel, , Oriental Despotism (New Haven, 1957), 101–7, 140–54Google Scholar, for the characteristics he assigns to ‘despotic power’.

⁶⁷ On a theory of state formation among the Betsileo, see Kottak, , ‘The Process… ’, 146–7.Google Scholar It is often assumed that trade was the main source of arms for Imerina, though there is ample evidence to show that even as early as the late eighteenth century Imerina was almost militarily self-sufficient since it manufactured its own guns and gunpowder of a very high quality. See Hugon, B., ‘Lois’, 1818, AN-SOM, Mad. 7/15, pp. 30–1.Google Scholar

⁶⁸ Bloch, M., ‘Property…’, 206–12.Google Scholar A recent Wenner-Gren symposium at Wartenstein, Burg, ‘Human Adjustment in Time and Space in Madagascar’ (18–27 08 1979)Google Scholar emphasized the sacred character of land as one of the aspects of Merina society which distinguished it from most other regions in Madagascar.

⁶⁹ Berg, ‘Royal Authority…’, passim.

⁷⁰ Recent research undertaken by the author as part of a forthcoming history of Imerina raises an interesting paradox: the growth of labour-intensive riziculture at the end of the eighteenth century corresponds with the peak of slave exports from Imerina.