landscape politics: the serpent ditch and the rainbow in west africa

13
NEIL L. NORMAN KENNETH G. KELLY Landscape Politics: The Serpent Ditch and the Rainbow in West Africa ABSTRACT Anthropologists and archaeologists have long been interested in the intersection of social, political, and religious insti- tutions and landscape features. Recent efforts have been aimed at elucidating the tensions between the perception and description of such features among Western and non-Western groups. This article seeks to contribute to this project through an analysis of a series of massive ditches (c. 17th–19th centuries A.D.) in southern Bénin, West Africa. In their accounts of the region, European travelers de- scribed these features through tropes and terminology that ascribe Western military designs and exploits. With insights drawn from archaeological and anthropological data, we argue a different perspective: that groups from the West African kingdoms of Hueda and Dahomey used the built landscape to reference cosmological factors, in attempts to negotiate and shape the political landscape of the region. [Keywords: West African Archaeology, landscape politics, Vodun, Dangbe, earthenworks] W HILE MAPPING 1 A PORTION of the massive ditch complex that surrounds the archaeological site of Savi 2 in coastal Bénin, 3 West Africa, one of us (Neil Nor- man) asked a Dahomean collaborator to run a transect through a ditch segment to plumb its depth. He looked at the ditch with trepidation and suggested it might be more appropriate if Norman took the measurement, since for him the ditch evoked the presence of Dangbe, the python deity worshiped by Vodun practitioners in the area. 4 This statement was striking, since virtually all of the ancient ditches were incorporated into a modern productive agri- cultural landscape, and other local collaborators had been involved with identification, mapping, and archaeological testing of these very earthworks. When Norman asked if it would be appropriate for him to enter the ditch himself, our collaborator suggested that it should be fine as long as Norman proceeded respectfully and realized the potency and potential danger of the space. This moment demon- strated clearly that the earthworks embodied differing cul- tural significance for specific individuals, compelling the research that follows. We consider this study as part of the growing body of comparative data elucidating the interrelations of political, social, and religious institutions with the built landscape (e.g., Bauer 1998; Bender 1993; Fox 1996; Holley et al. 1993; Kolb 1994; Koontz et al. 2001). In settings involved in the European colonial expansion following the late 15th cen- tury, contextualizing landscape features within such insti- tutions is often aided by research designs that conjoin his- torical and anthropological methodologies (cf. Hantman 1990; Sahlins 1985). Researchers investigating West Afri- can history from the 16th century to the present often ground their interpretations in European observations of the social and political interactions of the region (e.g., Curtin 1975; DeCorse 1998, 2001; Wallerstein 1986). Pri- mary documents have aided greatly in the understanding of African/European interactions in the area. However, others argue that greater efforts should be made to enu- merate the tensions, dissonances, and conflicts between European accounts of contact and interaction and other lines of evidence (Fleisher in press; Kelly 1997b; LaViolette in press; Miller 1980; Stahl 2001). Ann Stahl (2001:1–40) suggests that archaeological data are in a prime position to help us address not only con- flicts between primary sources but also the “silences” asso- ciated with the selective and subjective process of record- ing historical accounts (cf. Trouillot 1995). We suggest that common themes in attempts by West African groups to manipulate the built landscape and shape the ways it is presented and perceived offer an avenue to explore such historical processes at the local and regional levels. This methodology borrows from Martin Hall (1987:3–4), who suggests evaluating elements of the built landscape based on their relation to the goals and strategies used to natu- ralize and reinforce power relations within and between groups. Hall (2000:44) notes that these strategies are often AMERICAN ANTHROPOLOGIST 106(1):98–110. COPYRIGHT © 2004, AMERICAN ANTHROPOLOGICAL ASSOCIATION

Upload: mail2agastaya7024

Post on 01-Jan-2016

17 views

Category:

Documents


0 download

DESCRIPTION

NEIL L. NORMANKENNETH G. KELLY

TRANSCRIPT

N E I L L . N O R M A NK E N N E T H G . K E L L Y

Landscape Politics: The Serpent Ditch and theRainbow in West Africa

ABSTRACT Anthropologists and archaeologists have long been interested in the intersection of social, political, and religious insti-

tutions and landscape features. Recent efforts have been aimed at elucidating the tensions between the perception and description of

such features among Western and non-Western groups. This article seeks to contribute to this project through an analysis of a series

of massive ditches (c. 17th–19th centuries A.D.) in southern Bénin, West Africa. In their accounts of the region, European travelers de-

scribed these features through tropes and terminology that ascribe Western military designs and exploits. With insights drawn from

archaeological and anthropological data, we argue a different perspective: that groups from the West African kingdoms of Hueda and

Dahomey used the built landscape to reference cosmological factors, in attempts to negotiate and shape the political landscape of the

region. [Keywords: West African Archaeology, landscape politics, Vodun, Dangbe, earthenworks]

WHILE MAPPING1 A PORTION of the massive ditchcomplex that surrounds the archaeological site

of Savi2 in coastal Bénin,3 West Africa, one of us (Neil Nor-man) asked a Dahomean collaborator to run a transectthrough a ditch segment to plumb its depth. He looked atthe ditch with trepidation and suggested it might be moreappropriate if Norman took the measurement, since forhim the ditch evoked the presence of Dangbe, the pythondeity worshiped by Vodun practitioners in the area.4 Thisstatement was striking, since virtually all of the ancientditches were incorporated into a modern productive agri-cultural landscape, and other local collaborators had beeninvolved with identification, mapping, and archaeologicaltesting of these very earthworks. When Norman asked if itwould be appropriate for him to enter the ditch himself,our collaborator suggested that it should be fine as long asNorman proceeded respectfully and realized the potencyand potential danger of the space. This moment demon-strated clearly that the earthworks embodied differing cul-tural significance for specific individuals, compelling theresearch that follows.

We consider this study as part of the growing body ofcomparative data elucidating the interrelations of political,social, and religious institutions with the built landscape(e.g., Bauer 1998; Bender 1993; Fox 1996; Holley et al. 1993;Kolb 1994; Koontz et al. 2001). In settings involved in theEuropean colonial expansion following the late 15th cen-tury, contextualizing landscape features within such insti-

tutions is often aided by research designs that conjoin his-torical and anthropological methodologies (cf. Hantman1990; Sahlins 1985). Researchers investigating West Afri-can history from the 16th century to the present oftenground their interpretations in European observations ofthe social and political interactions of the region (e.g.,Curtin 1975; DeCorse 1998, 2001; Wallerstein 1986). Pri-mary documents have aided greatly in the understandingof African/European interactions in the area. However,others argue that greater efforts should be made to enu-merate the tensions, dissonances, and conflicts betweenEuropean accounts of contact and interaction and otherlines of evidence (Fleisher in press; Kelly 1997b; LaViolettein press; Miller 1980; Stahl 2001).

Ann Stahl (2001:1–40) suggests that archaeological dataare in a prime position to help us address not only con-flicts between primary sources but also the “silences” asso-ciated with the selective and subjective process of record-ing historical accounts (cf. Trouillot 1995). We suggest thatcommon themes in attempts by West African groups tomanipulate the built landscape and shape the ways it ispresented and perceived offer an avenue to explore suchhistorical processes at the local and regional levels. Thismethodology borrows from Martin Hall (1987:3–4), whosuggests evaluating elements of the built landscape basedon their relation to the goals and strategies used to natu-ralize and reinforce power relations within and betweengroups. Hall (2000:44) notes that these strategies are often

AMERICAN ANTHROPOLOGIST 106(1):98–110. COPYRIGHT © 2004, AMERICAN ANTHROPOLOGICAL ASSOCIATION

met with ambiguous and unexpected responses, and, thus,we fully acknowledge that the experiences and perceptionsof such ditches throughout West Africa were, as they aretoday, multifarious. One of the strengths of Hall’s researchin colonial and modern South Africa is to explore the em-bedded meaning(s) and political significance of landscapefrom the perspective of African groups involved in theirconstruction and use. Thus, necessary tensions often de-velop between European accounts of the African land-scape and the accounts of those African groups that createand maintain the features.

In the discussion that follows, we seek to demonstratesuch tensions between European and African accounts ofthe ways that ditches were used to signify and implementpower relations in social, political, and religious spheres insouthern Bénin. Particular attention will be focused on:(1) the Huedan5 capital of Savi (early 17th century to1727) and (2) the Dahomean capital of Abomey where, inthe early 18th century (see Figures 1 and 2), Dangbe wasincorporated into the Dahomean pantheon, as ditcheswere adopted as architectural features around that palacecomplex. European traveler’s accounts relate these ditches

to European design and inventiveness; however, we arguethat they more closely relate to efforts by West Africangroups to implement monumental architecture for thecreation of social distinction (Kelly 1995, 1997a), as wellas attempts to provide symbolic and physical protectionthrough evoking cosmological elements. We hope thatthis study will add to the voices of researchers calling fortheoretical models and heuristic devices that are sensitiveto African perspectives and cosmologies (McIntosh 1999;Schmidt 1996), that give time depth to historically specificmoments, and that situate these moments within the po-litical dynamics of the region (Piot 1999:1–26).

HISTORICAL AND REGIONAL BACKGROUND

Researchers commonly characterize the 14th and 15thcenturies as a period of intensification and consolidationin the coastal forest zone of modern-day Togo, Bénin, andNigeria, as kings in the region extended their spheres ofcontrol over areas of several hundred kilometers. In turn,these kings were reliant on regional governors who ad-ministered portions of their kingdom, collected tribute,mediated disputes, organized religious ceremonies, and

FIGURE 1. Project Area.

Norman and Kelly • Landscape Politics in West Africa 99

coordinated corvée labor for public works and agriculturaloperations. Attached artisans produced items from gold,brass, clay, ivory, and cloth, which were traded by coastalgroups to larger northern polities that controlled tradenetworks throughout the region and beyond (Law 1991;Polanyi 1966).

The arrival of European traders on the West Africancoast in the late 15th century, and their sustained effortsfrom the 16th–19th centuries to trade captive Africansacross the Atlantic, caused a radical reorganization of thisnetwork. Polities in the interior traded captives to coastalgroups, who used their position at the nexus of trade toexact tribute and taxes from both European and Africangroups. Coastal polities used this tribute to acquire Euro-pean armaments, further aiding the expansion of politicalcontrol and adding to the zeal to capture neighboringgroups. Regional conflict ensued as politically centralizedpolities scrambled to annex surrounding areas. During theheight of this conflict in the early 18th century, settle-ment patterns became tightly nucleated, as those livingnear territorial boundaries relocated to the centers of thekingdom to seek protection from raiding/slaving parties(Law 1990, 1991).

After visiting the region, the early-18th-century Euro-pean traveler Etiénne Des Marchais noted that the area inand around the Huedan capital of Savi was “so populated

with dwellings and inhabitants that it can be said to forma single town” (in Law 1991:59). Law estimates the popu-lation of the Hueda kingdom at approximately 100,000for the late 17th century; from this population of both freeand enslaved the king could field a military force of be-tween 20,000 and 40,000 for defensive and offensive op-erations (Law 1991:58–59). This military force was neces-sary to defend Savi and the coastal port and transatlanticentrepôt of Glewhe (later Ouidah) from the regional con-flict between the kingdoms of Oyo, Allada, and Dahomey.Dahomey posed the most immediate threat to Hueda, asDahomean kings sought to use military might to expandtheir political control to the Atlantic coast. A key featurein the military posturing and political hegemony involvedAfrican elite groups marshalling the symbols of their re-spective pantheons. In the section that follows, we outlinethe theoretic perspectives that guide this study of ditchsystems as representations of members of these pantheonsin southern Bénin, and the connection of these deities tothe political maneuverings of the area.

LABORING FOR DITCHES AND THE DEFINITION OFSOCIAL SPACE

Many of the older, large-scale urban settlements through-out the forest zone in Togo, Bénin, Ghana, and Nigeriashare the common element of incorporating ditches intosocial systems that attempt to construct physical barriersand social boundaries: zones of protection and inclusion.Patrick Darling’s (1984) research in and around Benin Cityin southern Nigeria includes the most complete discussionof the massive ditch features from this period. He notesthat the concentric bank-and-ditch systems in the areawere created not only to distinguish the royal palace com-plex from nonelite areas but also to encircle and protect amajority of the settlement. He estimates the total lengthfor the southern Nigerian earthenworks at 16,000 km withan average depth of three meters. Such a system wouldhave required the removal of 37 million cubic meters offill, and an estimated 150 million hours of labor to com-plete. This calculation led Graham Connah (2000) to writethat such structures speak as much to the ability of theKing of Benin and his religious administrators to organizeexcess, or corvée, labor as they do to the king’s ability toorganize defensive posturing of the kingdom. The ditchesappear to be a major outlet of the compulsory labor thatthe King of Benin commanded through powers of taxa-tion.

Connah (2000) notes a florescence of ditch systems inLater Iron Age (c. A.D. 1300) settlements throughout thesavanna and coastal forest areas of West Africa, where elitegroups coordinated enslaved and corvée labor for the con-struction of immense edifices and public works. RobinLaw (1991:93–94) correlates the political prominence ofkings in the area to their ability to keep these laborers inpublic displays of service and commerce. Africanist an-thropologists have recognized a close relationship between

FIGURE 2. Map of the Central Site Area of Savi.

100 American Anthropologist • Vol. 106, No. 1 • March 2004

the ability of an African political leader to command sur-plus labor and the extent of his or her administrative-juris-dictional authority (Guyer 1995; Webster 1990).

It is telling that in the coastal forest zone of West Af-rica, this labor was organized to create ditches rather thanwalls or other devices for separating social space. PaulaBen-Amos (1980:78–82) suggests the interpretations ofWest African ditches should not focus solely on their useas barriers but also take into account their use as symbolicstructures that demarcate the sacred from the profane. Aftercomparing historical accounts with contemporary inter-views, she suggests that in Edo cosmology the ditches andgates found throughout the Kingdom of Benin (14th–19thcenturies) served as a crossroads where aspects of the spiri-tual and natural worlds collided. Numerous offerings werethus required in these areas to placate the various entitiespassing between realms. Not surprisingly, the ditches asso-ciated with the most powerful crossroads between spiri-tual and earthly planes were those in closest proximity tothe palace complexes (e.g., Benin City), adding to theprestige of elite groups residing within palace complexes.Furthermore, elite groups were protected from outsiders andinternal political dissent by locating themselves withinthe physical and symbolic boundary provided by the ditchsystem.

In a similar fashion, Merrick Posnansky (1981) drawsfrom oral histories and archaeological data to interpret aseries of ditches at Notsé (occupied from mid–15th centuryto the present) in the coastal forest zone of modern Togo,as defining social boundaries.6 He recorded a four-sectionedwall-and-ditch system that enclosed most of the settle-ment. Subsequent work there by Dola Aguigah (1986, 1992)suggests the complex was not used for defensive purposesbut to demarcate the territory under Notsé control. Pos-nansky and Aguigah base their interpretation on the oc-currence of a ditch running inside the walls, where it addsno defensive value to the structure (Posnansky 1985). Thisinterpretation is countered by Nii Quarcoopome (1993),who draws from oral traditions to suggest the wall wasconstructed during the 17th century as protection againstraiding groups from the east and north. He also notes thata smaller wall (agbobovi), which circumscribes a muchsmaller area, was constructed during the reign of the in-itial king of Notsé to “isolate the royal domain and to pro-tect the notables” (Quarcoopome 1993:114–115, emphasisadded). Thus, it is conceivable that Notsé walls and ditchesserved multiple defensive and symbolic functions, includ-ing division of the site between elite and nonelite zones.Drawing from ethnohistoric accounts, Sandra Greene ech-oes this interpretation, claiming that portions of the wall-and-ditch system served to “symbolize the town’s status asa major economic power and ritual center within the re-gion” (Greene 2002:15). In the section that follows, wediscuss evidence for ditches in Southern Bénin being usedto define social space and represent political prominencein a similar fashion to those described above.

DITCH SYSTEMS IN SOUTHERN BÉNIN

The centrality of Savi to the trade of captive peoples drewFrench, English, Portuguese, and Dutch traders and emis-saries to reside in the area (Polanyi 1966). The variouskings of Hueda were adept at manipulating the competi-tive nature of these European groups to benefit their owncoffers. Huedan kings resisted exclusive trading statuswith any single European group, thus allowing maximumcompetition between European traders and Huedan con-trol of the last stage on the African side of the Atlantictrade. In turn, the authority of the kings of Hueda was reli-ant on their ability to monopolize these trade networksand to control related political relationships (Kelly 1997a;Law 1991).

As Europeans established ephemeral settlements insouthern Bénin in the late 17th and early 18th centuries,they became entangled in the politics of the region. Huedanelites restricted the movement and relative freedom ofEuropean traders and diplomats, by curtailing their oppor-tunities to move throughout the kingdom and their settle-ment opportunities (Kelly 1997a). Huedan kings orderedthe construction of European “trading lodges” within theseries of ditches that marked the boundary of their palacecomplexes, thus placing European activities under theconstant gaze of the king and his representatives.

This circumscribed settlement isolated the Europeantraders by separating them from their naval reinforce-ments at the coastal port of Ouidah. Archaeological evi-dence7 (Kelly 1997a, 2001) from Savi suggests that build-ing materials such as European bricks, originally destinedfor these European lodges, were diverted for use within thepalace complex. This requisition of material exemplifiesthe dominant position that Hueda kings held in tradingrelationships and alludes to the strategies used by elites tocreate differentiation through use of the built landscape.Namely, elites used architectural features to signify anddistinguish their privileged positions in social, political,and religious spheres. In addition to materializing thisseparation, the activities conducted within and outsidethe palace zone further point to the differentiation of in-teraction spheres.

The elite zone contained within the palace complexbecame the economic hub for the kingdom as it providedspace for negotiations between European and African trad-ers. At the same time, the palace complex was also thecenter for the creation of symbolic capital (Bourdieu 1990)through the manipulation of European traders (Kelly1997a), and it was the locus for religious ceremonies con-ducted by the king and other religious officiates. The ditchsystem that divided the palace complex from the remain-der of the settlement separated these spheres of interac-tion and placed symbolic and physical space between eliteand nonelite activities.

European visitors, traders, and emissaries residedwithin the palace complex at Savi more or less continu-ously from 1670 until its destruction in 1727. Yet these

Norman and Kelly • Landscape Politics in West Africa 101

European eyewitnesses, many of whom were trained ob-servers, made no mention of the earthworks that enclosedthe elite district (Kelly 1997b). These observers, nonethe-less, described in copious detail the features of the builtand social landscape that they considered relevant tomaximizing trade and advantage. Over a century after Savifell to Dahomey, Sir Richard Burton reported after a visitto the area that:

Nothing now remains of the ancient glories of Savi; evenin A.D., 1772, we are told, only the moats of the manyEuropean forts could be traced. A long trench, with a tallgrowth of trees, was the sole remnant of the palace occu-pied by the Whydah kings. [Newbury 1966:94]

It is not surprising that Burton attributes these structuresto European innovation and action, as such accounts andtropes were common among European travelers from the19th century (Hall 2000). Undoubtedly, Burton’s account,as well as modern historic accounts, were influenced byearlier European maps and documentary sources from the17th to 18th centuries that often exaggerated the size andposition of European trade lodges in relation to the Huedanpalace complex (Kelly 1995:277–279). Archaeological evi-dence suggests strongly that these ditches were the resultof Huedan efforts and designs.

Kelly (1997a) draws from the archaeology of Posnan-sky (1981) and Aguigah (1986, 1992) at Notsé when inter-preting the ditch system at Savi as an attempt by Huedansto define internal social relations. Kelly (1997b:361–362)downplays the physical defensive significance of the Saviditches as they were breached by repeated causeways thatapparently allowed access to the palace complex withoutcheckpoints or other means of controlling movement intothe elite zone. Furthermore, the ditches do not encompassthe entire town—only the elite section (Kelly 1997a:360).

Kelly concludes that the ditches served as a material,and monumental, representation of the authority of theking of Hueda, and a reminder of the political prominencenecessary to organize the labor involved. Archaeologicalevidence supports the assumption that ditches stood be-tween elites and nonelites, as survey (Kelly 1995) and ex-cavation (Brunache 2001; Kelly 1995, 1997a; Kelly et al.1999) data reveal a higher proportion of non-African tradegoods concentrated within the palace complex comparedwith those areas outside.

In a similar fashion to artifact variance in elite andnonelite areas, the massiveness of the architectural fea-tures located inside the palace complex served to signifythe relative status differences of those living inside. Struc-tures inside were up to 115 meters in maximum dimen-sion, while initial investigation of the structures outsidethe palace complex indicate that nonelite house structuresaveraged approximately two-and-a-half meters in length(Kelly 1995). The late-17th-century observer William Bos-man recorded a similar massiveness of architecture associ-ated with the palace, describing the Savi palace complexwith the king “who lives in . . . majesty at Sabbee [Savi].[The] palace is a . . . large bamboo building of a mile or

two round, where he keeps near a thousand women”(1721:366). This discrepancy suggests the amount of labororganized to construct and maintain the structures withinthe elite zone, as well as the ditch system that stands be-tween elite and nonelite areas of the site. Archaeologicalinvestigations indicate that the Savi ditches enclose anarea of approximately six-and-a-half hectares through asegmented system, with individual segments varying be-tween ten and 70 meters in width and up to 220 meters inlength. The maximum depth of the segments was re-corded at eight meters below the surface of the edge of theexcavation. Occasional low mounds, possible vantagepoints two to three meters in height, are located adjacentto the ditches, but they do not appear to be related towalls. The ditches do not retain, or drain, large quantitiesof water; however, standing water may collect in lower ar-eas of the ditches during the rainy season (Kelly 1997a;Kelly et al. 1999). They meander in noncontiguous seg-ments, that when viewed from the ground level, trace aserpentine pattern between the elite and nonelite sectionsof the site.

Kelly suggests that if the Hueda did not regard theditches as defensive, then it is possible that European ob-servers understood this, but saw no alternative role theditches could have played. He posits that European visi-tors and residents did not comment on the ditches be-cause they were not identifiable in terms of their own pre-conceptions of the size, shape, style, and placement ofmilitary fortifications (Kelly 1995:303). We propose thatwhen the historical silences associated with these ditchesare considered alongside the accounts of political and re-ligious posturing in and around the palace complex, theHuedan strategies associated with defining social spacecan be understood in more nuanced fashion (Stahl 2001).For Hall (2000), there is a close relationship between theability to make claims to authority and the ability to con-trol material representations (i.e., the built landscape),and products, of such relations of power. In his concep-tion of political power, a key strategy becomes attemptingto control the presentation and perception of these mate-rial representations at times when a group’s, or individ-ual’s, prominence is on display. Therefore, the ditchesmay also be interpreted in relation to attempts throughoutthe region to control the presentation and use of religioussymbols and political iconography associated with the ser-pent deity Dangbe. The section that follows provides anoverview of Dangbe worship in southern Bénin, and therelationship between those religious practices and theditches at Savi and Dahomey.

DANGBE WORSHIP IN SOUTHERN BÉNIN

Ian Hodder (1995) suggests issues of performance and dis-play should be considered in the interpretation of archae-ological assemblages, such that archaeologists should at-tempt to link material culture with the pageantry of pastevents. Melville Herskovits (1938), Herskovits and Frances

102 American Anthropologist • Vol. 106, No. 1 • March 2004

Herskovits (1933), and Law (1991) note that Europeantravelers to the region were fascinated with Huedan wor-ship of the python deity Dangbe and recorded numerousaccounts of Dangbe being called upon on occasions ofeconomic, political, and social significance. In the early18th century, Des Marchais recounted that Dangbe was in-voked for political purposes “on all occasions relating totheir government” (in Law 1990:209). In the mid–19thcentury, Richard Burton described the sacred python con-sidered the earthly manifestations of Dangbe as “a brownyellow-and-white-streaked python of moderate dimen-sions; and none appear to exceed five feet” (in Newbury1966:74). This characterization corresponds with the col-oration and physiology of one type of python indigenousto the area today: the Royal/Ball Python, Python regius,usually less than six feet (Villiers 1975). Similar to the ac-counts of modern Dangbe practitioners (Norman 2000),historic travelers noted that Dangbe was considered toregulate weather and agricultural fertility.

An early-18th-century traveler to the area recordedthat “the great men paid their respects to the king of Why-dah as the head of the descent group which had as itstohwiyo [symbolic head] the revered serpent Dan[gbe], andrecognized him as their leader in negotiations with theEuropeans” (Argyle 1966:16). However, the sociopoliticaland religious authority associated with Dangbe worshipboth legitimated and circumscribed the actions of theking, as he was obliged to provide gifts to governors andother royal retainers during the annual ceremonies associ-ated with Dangbe.

The most sacred ceremony on the Huedan calendarwas the procession of the King of Hueda to the main tem-ple of Dangbe located immediately outside the palacewalls. Des Marchais noted such a ceremony associatedwith the coronation of Huffon (c. 1713–15; see Figures 3and 4). He describes a formal procession to the temple ofDangbe by the wives of the king after the coronation cere-mony held within the palace complex, followed by a pro-cession to the shrine by the king himself three monthslater. Law suggests that this action secured Huffon’s en-dorsement for his position “from the kingdom’s nationaldeity as well as from its traditional overlord” (Law1990:225). Before this ceremony, there was civil unrest as-sociated with the legitimacy of Huffon, concerning thepolitical jockeying between two of his chief governors. Inattempting to mount a coup against Huffon, one governorrequired his supporters to swear allegiance through takingoaths to Dangbe (Law 1990:221).

The public ceremonies associated with the annualprocession were normally held immediately outside thepalace complex, with major rites thus adding to the sa-credness within, and further linking the authority of theking with the supernatural (Blier 1995; Law 1991). Ritualrestrictions reinforced the mystique of the palace and the“exalted status” of the king, who left the palace com-pound “once or twice” a year (Law 1991:77–78). The kingremained inside the palace complex except for the annual

procession to the temple of Dangbe, where his wealth wasplaced on display as his court paraded items from the pal-ace compound around the palace walls and to the templeof Dangbe (Law 1990:226). A large portion of the Huedanpopulation was mobilized for the procession by ritual spe-cialists, who led the drumming, dancing, and singing (Law1991:95–96). The king of Hueda served as the head offici-ate, while his wives demonstrated royal largesse by throw-ing cowries and other items of value to the assembledcrowd. The procession and related ceremonies resembleAchille Mbembe’s (1992) description of the use of specta-cle as a strategy for reinforcing political structures and ex-isting relations of power, as such events lend credibility tothe existing power structures (Orr 2001; Turner 1974, 1995).Following Mbembe, such structures would allow elite groupsto conduct ceremonies, which translated into politicalprominence in the more mundane aspects of life. Like-wise, the bodies (Figure 3) of those processing and observ-ing (including Europeans and those animals and personsoccasionally sacrificed to Dangbe) added to the aura of theHuedan elite groups and reinforced their position as themediator between the sacred and profane, arbiter of lifeand death. It is clear from documentary sources and sub-sequent archaeological investigations that the processionwould have coincided with the path of the ditch system,as the party moved outside the palace and proceeded tothe temple of Dangbe.

The ditches surrounding the palace at Savi circum-scribed the path followed by the procession and forcedparticipants to retrace the symbolic boundary of the elitedistrict. In one instance when the king of Hueda left thepalace to invoke Dangbe and reestablish the relationshipof exchange between Hueda and the Huedan pantheon, itis telling that the structure of the ritual focused on move-ment in and out of the palace complex, with the encir-cling ditches marking the point of transition between thetwo spheres of Huedan society. Dangbe is considered tocontrol movement and the transition between social cate-gories (Blier 1995:201). As the physical representation ofDangbe, the ditches would have literally shaped themovement of the procession into a form consistent withthe aesthetic principles associated with Dangbe, discussedbelow.

In the section that follows, we highlight the incorpo-ration of palace structures into the political machinery ofthe kingdoms of southern Bénin.

PALACES, POLITICS, AND THE BUILT LANDSCAPE

Numerous archaeologists (e.g., DeMarrais et al. 1996; Joyceand Winter 1996; Thomas 1993) have commented on therelationship between elements of the built landscape andpolitical activities. Recent attempts have been made toplace such landscape elements within their social context(e.g., Bender 1993, 2001; Kelly 1997a). Edna Bay (1998: 11–12)notes that in the 18th–19th century the palaces in thecoastal forest zone were active elements in the political

Norman and Kelly • Landscape Politics in West Africa 103

interactions of the area. The Dahomean palace at Abomeywas considered by individuals in the area to be the centerof the universe, and as such it also exemplified the funda-mental ideological and organizational principle in Da-homean cosmology: Any whole is the sum of complemen-tary forces.

The Dahomean palace itself was organized to exem-plify the oppositions in these complementary parts of in-side–outside, right–left, royal–commoner, male–female, andto invert the relationships experienced outside the com-plex (Bay 1998). By inverting traditional relationships andsocial categories, the palace presented Dahomeans a sa-cred space where many rules and norms upheld outsidethe palace did not apply. For example, gender roles werereversed within the palace, as the only way to talk to theking was through courtesans acting as intermediaries anddiplomats. Similarly, female warriors served as the king’spersonal guard, eunuchs appeared in palace audiences inwomen’s dress, and courtesans appeared in male dress ofYoruba style. Other norms observed outside the palacewere upheld, as right superseded left and elite supersededcommoner, and all who entered the palace were obliged toshow deference to the authority of the king (Bay 1998).

The ditches at Savi served to mark the boundarywhere outside social relations and practices did not apply.Inverting the typical relationship on the West Africancoast, and highlighting the tensions inherent in their rela-

tionship, European traders, the major source of Huedawealth, were compelled to live inside the palace. Europeantravelers and diplomats who held high offices outside thepalace were forced to genuflect on arriving before the kingand remain so until their audience ended. These inver-sions also affected the daily practices of the king; thoughmost individuals could move freely throughout the king-dom, the king’s movement was restricted to life within theboundary ditches of the palace (Law 1991:75–78).

It is not surprising, then, that a deity who permeatedthe religious and economic activities of the Huedan peo-ple would also be incorporated into the kingdom’s politi-cal iconography. Ben-Amos (1999) suggests that religiousiconography was incorporated into art created in the 18th-century Kingdom of Benin, and that elite groups used suchart to shape the interpretation of events and as mnemonicdevices used to spur the recollection of oral histories. Thebuilt landscape and monumental architecture of southernBénin was used to signify similar relationships betweenelites and the divine, and to naturalize their position atopHuedan and Dahomean society. For Hall (2000:25–27), ar-chitectural features effectively communicate politicalauthority when they are linked to elite groups throughperformative events such as the parade described above,and the action of corvée and enslaved laborers toiling inthe ditches in compliance to the king and his administra-tors. In such a social environment, the built landscape can

FIGURE 3. Procession to the Temple of Dangbe, from Des Marchais (1731: Plate 7).

104 American Anthropologist • Vol. 106, No. 1 • March 2004

be interpreted as an expression of social and politicalstructure that communicates these relative differences instructural positions, through symbolic associations thatdenote affiliation with, and differentiation between, cer-tain groups (cf. Lawrence and Low 1990; Kelly 1997a). Thefollowing section describes attempts to designate or ap-propriate symbols of Dangbe for such political purposes.

DANGBE AND THE DITCHES

Beyond the proximate relationship between ceremoniesfor Dangbe and the ditches that surrounded the palace,modern residents living near the site of Savi associate rem-nant ditches with the aesthetic elements used to describeserpents in general, and Dangbe in particular. DanielMiller and Christopher Tilley (1984) and Tilley (1994) sug-gest that landscape features should be evaluated with anapproach empathetic to the possible experiential qualitiesof certain settings. They suggest that adequate inferencesregarding the political significance of archaeological fea-tures must include analogies related to emotions evokedby such features. Our collaborators considered the dank-ness of the ditches and the serpentine patterns they traceon the landscape analogous to similar qualities of serpents.Melville Herskovits (1938:245–255), while comparing andcontrasting the serpent deities worshiped throughout

southern Bénin in the mid–20th century, noted that thesedeities are associated with things in the world that areflexible, moist, and that fold, refold, and coil, as well asthings that are considered to connect the earth and sky.These associations parallel the physical characteristics ofpythons and the ecological areas that they inhabit. Al-though pythons spend most of their lives in terrestrial set-tings where they brood in abandoned burrows, they alsohunt from trees and return to branches for protection afterengorging on the ground (Villiers 1975). The associationwith dankness, and moisture might relate to the python’sadept ability to negotiate aquatic environments, as well asthe slight oily feel of their skin. Pythons are excellentswimmers and occasionally hunt from rivers and streams.John Murphy and Robert Henderson (1997:17–18) notemodern accounts of pythons swimming several kilometersin and around the Victoria Nyanza islands of eastern Africa,and those of Ghanaian fishermen who reported catchingpythons in fish traps set in the sea, some of which werealive after being submerged for several hours. The impor-tance of the Savi ditch system referencing qualities associ-ated with pythons in general and Dangbe in particular canbe better understood as strategies for reinforcing politicaland religious hegemony of those groups residing withinthe ditch system, when it is considered in relation to ac-counts describing the manner in which pythons shaped

FIGURE 4. Huedan coronation ceremony. Note the conical python shrine in the center of the palace courtyard, from Des Marchais (1731:plate 4).

Norman and Kelly • Landscape Politics in West Africa 105

the daily practices and bodily comportment of peoplefrom the area.

In the mid–19th century Forbes recorded that aroundthe Dangbe temple at Ouidah there were:

many snakes of the Boa [python] species. These are al-lowed to roam about at pleasure; but if found in a houseor at a distance, a fetish man or woman is sought, whoseduty it is to induce the reptile to return, and to reconductit to its sacred abode, whilst all that meet it must bowdown and kiss the dust. Morning and evening, many areseen prostrated before the door [of the temple of Dangbe].[1966:109]

The treatment of pythons in this account (see also Figure5) is similar to the description the king received while hewas holding court. As mentioned above, courtesans as wellas European visitors were compelled to genuflect through-out their audience with the king (Law 1991:78). In a simi-lar sense, the king and pythons normally resided in sancti-fied spaces out of view and shielded from the profaneelements of daily life. John Duncan noted that in the mid-19th-century homage was paid to pythons:

When one of [the pythons] is picked up by any one, oth-ers will prostrate themselves as it is carried past, throwing

dust on their heads, and begging to be rubbed over thebody with the reptile. After taking the snake up, a veryheavy penalty is incurred by laying it down, before it isplaced in the fetish-house. [1847:126–127]

When the reverence shown to these earthly representa-tions of Dangbe is considered, it is not difficult to imaginethe symbolic importance of placing monumental architec-ture that references the qualities associated with the deity,in close proximity to the symbolic center of the kingdom.The symbolic protection that the ditches afforded the resi-dents of the palace complex at Savi should be viewed inrelation to Huedan conceptions that Dangbe will protectthose groups contained within, or beyond, physical struc-tures that evoke qualities associated with serpents. Baynotes that an early-18th-century European observer of thebattle for Savi suggested that the Huedans

only went every Morning and Evening to the River side,to offer to their principle God, which was a particularharmless Snake they adored, and prayed to on this occa-sion, to keep their Enemies from coming over the river . . .there is a constant Tradition amongst them whenever anyCalamity threatens their Country, by imploring theSnake’s Assistance, they are always delivered from it . . .the Pass of the River being . . . wholly left to the Care ofthe Snakes, whom the Enemy little feared; and they hav-ing observed for several Days, that the Whidaws kept noset Guard there, it encouraged the King of Dahomè’s Gen-eral to send two hundred of his Soldiers to ford the River.[Bay 1998:60]

Through activation by sacrifice and ceremonies, theriver became a symbolic boundary defended by Dangbeagainst attack from invading forces. In light of these ac-tions taken by Huedan devotees to engage Dangbe in thedefense of Savi, it becomes difficult to support conten-tions in European travelers’ accounts and, later, interpre-tations that draw on them (Akinjogbin 1967:69–72; Argyle1966:19; Law 1991:284–285), that little was done in de-fense of the Huedan Kingdom. European observers over-looked the nuances of warfare being waged by Dahomeyagainst Hueda, using the application of flintlocks, bayo-nets, castles, and cannon as their referent of military op-erations. They furthermore underestimated the impor-tance, in terms of Huedan cosmology, of the defense beingmounted at the river. In the stand at the river boundingHuedan and Dahomean territories, Huedans appear tohave used one of the most potent forces available in theirarsenal—an appeal to the serpent deity to control move-ment at the frontier of their kingdom.

A similar strategy can be seen in relation to elite groupsat Savi, who used the annual procession to the shrine ofDangbe to activate the ditch system at Savi as the last lineof defense against enemies who would target the symboliccenter of the kingdom. The center of the Savi was strategi-cally important not only as the ritual center but also as aphysical sanctuary for sacred pythons and the head offici-ates for Dangbe worship. Travelers’ accounts suggest thatHuedans treated symbols of Dangbe worship with the ut-most respect, which made their avatars the targets for

FIGURE 5. Temple of Dangbe in Ouidah, from Chaudoin (1891:343).

106 American Anthropologist • Vol. 106, No. 1 • March 2004

outside groups. Law (1991:28) notes that recent oral histo-ries record that the key to military victories throughoutthe region involved capturing and retaining the royal py-thons, the incarnation of Dangbe.

In such a social environment in which individualsassociated military victories and political and religiousauthority with serving as the steward for the royal serpents,it is not difficult to comprehend the importance of pro-tecting the serpents from outside groups. Likewise, if out-siders understood the power fields that could be mobilizedby Dangbe, the importance of monumental architecturethat references the aesthetic qualities associated with thedeity can be more clearly associated with the strategies as-sociated with Huedan defense.

The relationship between Dangbe worship and earth-works becomes even more striking when viewed in light ofDahomean efforts to incorporate elements of the panthe-ons, cosmologies, and material culture associated withDangbe into Dahomean religious and political practices.

THE APPROPRIATION OF DANGBE AND DITCHES

Dahomean soldiers targeted not only the Huedan palacecomplex and its inhabitants but also the royal serpents as-sociated with Dangbe. The serpents were incorporated lit-erally and symbolically into the Dahomean kingdom bybeing consumed by Dahomean troops at the time of Savi’sconquest in March of 1727 (Law 1991). In a similar fash-ion, the Huedan populace was incorporated into the Da-homean fold. William Snelgrave witnessed the decapita-tion of 400 war captives in 1727 and noted a collection of4,000 skulls from sacrifices offered after the conquest ofHueda earlier the same year (in Law 1986:249). WilliamSmith recounts that the general who put the torch to theEuropean lodges at Savi ordered “all the boys in the Camp;some of which were not above Seven or Eight Years of Age,to cut off the Heads of All the Aged and Wounded amongthe Captives that were unmerchantable” (Smith 1744:192).The collection of skulls is a key strategy, because the Da-homean king considered the heads of his subjects hisproperty, with skulls the symbolic seat of knowledge andreligious essence (Law 1989). Through incorporating headsof the conquered Huedans into the royal treasury, Agaja,king of Dahomey at Savi’s conquest, was symbolically in-creasing the size of his kingdom. In doing so, and annexingHuedan lands, Agaja also incorporated aspects of Huedancosmology into the Dahomean pantheon, and Huedanmaterial culture into the repertoire of forms created by ar-tisans housed within the palace compound at Abomey.

European observers recognized that Dahomey’s strategyof permitting continued worship of Dangbe was instru-mental to assimilating the conquered Hueda. After Agaja’sconquest of the Hueda, he negotiated with Huedan ritualspecialists for Dahomean acquisition of Dangbe and otherlocal deities. Dahomean administrators shifted the wor-ship of the python deity from the former capital of Savi tothe coastal port of Ouidah where Dahomean bureaucrats

could monitor its activities (Law 1991). The compound ofthe Yovogan, the Dahomean chief of Europeans and stateofficial governing the affairs of Ouidah for the king of Da-homey, was located between the two largest threats to theking’s influence over the area; the Temple of the Pythonand the European trading quarters. Later, the Catholic Ca-thedral entered the ideological battle after being estab-lished directly adjacent to the Temple of the Python (Kelly2002).

Although Dangbe worship never became as prominentas other major gods worshiped on the Abomey plateau,Agaja attempted to incorporate Dangbe into the liturgy ofDahomean Vodun. While Dangbe was being installed inthe Dahomean pantheon, Law notes that Etienne Gallot, aFrench officer serving in Ouidah, was engaged by Agaja toteach the Dahomeans “how to dig trenches and raise crudefortifications, which was unknown among these people”(Law 1992:110). The statement that these techniques wereunknown is not credible, because Dahomey had political,economic, and social relations with both the Kingdoms ofHueda and Benin, both of which employed ditches aroundtheir palace complexes since at least the 17th century (Law1991). Furthermore, architectural features found aestheti-cally pleasing or militarily advantageous readily spreadthroughout the southern forest zone (Shinnie 1971:22–24).It seems more likely that the ditches represent an effort toincrease the defensive qualities of the Abomey palace com-plex, while at the same time incorporating monumental ar-chitecture associated with Dangbe into the material cultureof the capital. The fact that the creation of the ditches co-incided with a time when thousands of war captives wereremoved from Savi to Abomey and most likely partici-pated in the construction of the ditches as enslaved laborsuggests that Agaja used Huedan labor to construct thephysical symbol of the merger of the two kingdoms. Thissymbol was a ditch system that evoked the political, relig-ious, and economic prominence of the Huedan palacecomplex so closely associated with Dangbe.

Such earthworks were followed by other attempts toincorporate serpent imagery into the Dahomean palace atAbomey and associate elite activities with its prominence.In the 1850s, Guezo, King of Dahomey, commissioned thesculpture of bas-reliefs representing Dan Ayido Houédo, aserpent biting its own tail, as one of the symbols of hisreign. Dan Ayido Houédo, known also as the “rainbow ser-pent,” is considered by our collaborators in the Savi areaas a derivation of Dangbe and serves in a similar fashion inthe Dahomean cosmology and mythology, as it did earlierin Huedan cosmology. Namely, Dan Ayido Houédo wasconsidered to control the eternal movement between theearthly plane and the hereafter, as well being the source ofmaterial wealth for those closely associated with his wor-ship (Blier 1995). By comparing and relating himself toDan Ayido Houédo, Guezo was able to project an “aura oftimeless power” (Piqué and Rainer 1999:75).

Beyond referencing the element of timelessness asso-ciated with Dangbe, it was strategically important for Guezo

Norman and Kelly • Landscape Politics in West Africa 107

to associate himself with the earlier military exploits ofAgaja. The bas-reliefs located throughout the palace com-plex were central elements to the transition and mainte-nance of political power in the kingdom, because royalstorytellers used them as mnemonic devices that illus-trated and evoked the oral histories of the kingdom.Through retelling and shaping of the histories both ver-bally and through the built landscape, the legacy of theDangbe became inextricably linked to that of the rainbowserpent Dan Ayido Houédo, as the history of the Daho-mean palace became linked to the Huedan palace.

CONCLUSION

In numerous societies in the West African coastal forestzone, groups created ditch systems as means of protectionby placing both literal and symbolic space between zonesof their settlement. As symbolic barriers, the voids of theditches created earthly manifestations of cosmological fea-tures, patrolled by members of their respective pantheons.Following Hall’s (1987) discussion of the way elements ofthe built landscapes are used as political features to distin-guish relationships and boundaries between African andEuropean groups, we have described the series of massiveditches at the urban centers of Savi and Abomey in rela-tion to the political strategies of the Huedans and Da-homeans. That such earthworks were the product of large-scale mobilizations of corvée and enslaved labor points tothe prominent role they played in West African social rela-tions, as well as physical reminders of the ability of elitegroups to organize such labor. European observers relatedthe construction of the ditch systems at Savi and Abomeyto European design and deployment. However, archaeo-logical and ethnographic data show that ditch and otherearthenworks were used throughout the region prior toEuropean incursion (Connah 2000; Darling 1984; Posnansky1981). Thus, earthworks at Savi and Abomey most prob-ably relate to attempts to mitigate West African interpolityconflicts through incorporating the physical and symbolicdefensive value of ditches related to cosmological bounda-ries into the material culture of the kingdoms.

In these ways, the symbols of Dangbe inspired by pre-vious practices at Savi were incorporated into the pan-theon and material repertoire of the Dahomean kingdom,evidenced today by the ditches still visible near the palaces atAbomey and the place that the rainbow serpent holds inthe Dahomean cosmology. The importance of Dangbe’srelationship to ditches begs comparison to earthen struc-tures far from West Africa, in places where individualsfrom the area came to live. Two examples are the use ofditches in modern Haiti associated with Vodun rites forthe “Rainbow Serpent” (de Heusch 1989:295–296), and de-fensive ditches associated with the 17th-century MaroonKingdom of Palmares in Brazil (Diggs 1953:66). This tre-mendous expenditure of labor on ditch construction byWest Africans and their descendant groups, on both sides

of the Atlantic, speaks to the prominence of such struc-tures in the negotiation of power relations.

By focusing on strategies using elements of the builtlandscape to create social boundaries, this case study at-tempts to revisit moments of European interaction throughthe perspective of nonwestern groups. Such critical reex-aminations are sorely needed for periods and areas whereEuropean travelers, though a demographic and politicalminority, form the recorded and vocal majority throughtheir primary accounts. By highlighting these alternative,and often contradictory, versions of the significance ofelements of the landscape in nonwestern settings, otherlines of evidence can be brought to the interpretation offeatures that, since the moment of “contact,” have lan-guished under theories drawn primarily from European ac-counts. We hope to add to the scholarship of anthropo-logical investigation of articulation of such socioculturalfactors and landscape features, by situating this specificcase study from southern Bénin within a broader regionalphenomenon of West African groups’ incorporation ofditches and other barriers into social, political, and cosmo-logical institutions.

NEIL L. NORMAN Department of Anthropology, University ofVirginia, Charlottesville, VA 22904KENNETH G. KELLY Department of Anthropology, Universityof South Carolina, Columbia, SC 29208

NOTESAcknowledgments. This research was made possible through thesupport and encouragement of many people. We would particu-larly like to recognize Professor Merrick Posnansky, UCLA; Pro-fesseurs Alexis Adandé, Elisée Soumoni, and Joseph Adandé, De-partment d’histoire et d’archéologie, Université Nationale duBénin d’Abomey-Calavi; Mme. Rachida de Souza, SecrétaireGénérale de le Ministère de la Culture et de la Communication;M. Eric F. Totah, Directeur du Patrimoine Culturel; Mme.Micheline Egounlety, Conservateur, Musée d’Histoire de Ouidah;the Agomadje family, Savi; M. Boniface Bossoukpè. The Universityof South Carolina College of Liberal Arts Scholarship SupportGrant provided funding for the 1999 research of Kelly, underwhose guidance and auspices Norman worked. The University ofVirginia, Graduate School of Arts and Sciences and Department ofAnthropology provided funding for Norman’s summer 2003 re-search project. At the University of Virginia, this article benefitedfrom comments of Carolyn Heitman, Dennis Blanton, SeoyeonChoi, Yadira Perez, Abigail Holeman, Alex Caton, Matthew Meyer,David Sapir, and Dell Upton. Concepts for the article were devel-oped through coursework with Hanan Sabea and Patricia Watten-maker. Jeff Hantman was a stabilizing force and an endless sourceof theoretical information through the article’s construction andrevision. Norman owes a deep debt of gratitude to Adria LaVio-lette, whose previous research, guidance, and counsel fundamen-tally informed this article and who clarified tangled prose throughreading and commenting on numerous drafts. We appreciate thereview and critical commentary of Ann Stahl, Barbara Bender, anda third anonymous AA reviewer.

1. This article originated as Neil Norman’s M.A. thesis in anthro-pology at the University of Virginia, which explored the cosmol-ogy of the ditches at Savi and Abomey. It builds on published andunpublished work, intellectual guidance, and fieldwork opportuni-ties provided by Kenneth Kelly since 1998.

2. Kelly directed six field seasons of archaeological survey, testing,and data recovery in the region from 1991–99. Norman conducted

108 American Anthropologist • Vol. 106, No. 1 • March 2004

two ethnoarchaeological field seasons of interviews and guidedsurveys of sacred sites in the region in 1999 and 2003.

3. By convention and for clarity, the modern political state ofBénin will be differentiated from the historical Kingdom of Beninby the accented “é.”

4. For a detailed discussion see Jakob Spieth 1911, Melville Hersko-vits 1938, Melville Herskovits and Frances Herskovits 1933, andSusan Blier 1995.

5. The historic polity of “Hueda” is transliterated using the Eng-lish form to distinguish it from its coastal port with the same name“Ouidah” (French transliteration), as well as historic sources thatuse Whydah, Whidaw, Fida, etc. (see Figure 1).

6. See also Jack Goody’s 1957:83–87 description of the Lo Da-gaba’s use of ditches as social boundaries.

7. For a detailed discussion of the archaeological work at Savi, seeKelly 1995 and 2001.

REFERENCES CITEDAguigah, Dola A.

1986 Le Site de Notsé: Contribution à l’ Archaéologiques du Togo.Ph.D dissertation, Université du Paris I.

1992 LesPavements en Tessons de Poterie, Organization del’Espace dans leRégion du Golfe du Bénin: leCas du Togo. WestAfrican Journal of Archaeology 22:133–144.

Akinjogbin, I. Adeagbo1967 Dahomey and Its Neighbors 1708–1818. Cambridge: Cam-

bridge University Press.Argyle, William. J.

1966 The Fon of Dahomey: A History and Ethnography of the OldKingdom. Oxford: Clarendon Press.

Bay, Edna G.1998 Wives of the Leopard: Gender, Politics, and Culture in the

Kingdom of Dahomey. Charlottesville: University of VirginiaPress.

Bauer, Brian S.1998 The Sacred Landscape of the Inca:The Cuzco Ceque System.

Austin: University of TexasPress.Ben-Amos, Paula

1980 The Art of Benin. London: Thames and Hudson.1999 Art, Innovation, and Politics in Eighteenth-Century Benin.

Bloomington: Indiana University Press.Bender, Barbara

1993 Introduction: Landscape-Meaning and Action. In Land-scape: Politics and Perspectives. Barbara Bender, ed. Pp. 1–17.Providence, RI: Berg.

2001 Contested Landscapes: Movement, Exile, and Place. In Con-tested Landscapes: Movement, Exile and Place. Barbara Benderand Margot Winer, eds. Pp. 1–18. Oxford: Berg.

Blier, Susan Preston1995 AfricanVodun: Art, Psychology, and Power. Chicago: Uni-

versity of Chicago Press.Bosman, William

1721 A New and Accurate Description of the Coast of Guinea, Di-vided into the Gold, the Slave, and the Ivory Coasts. London: J.Knapton.

Bourdieu, Pierre1990 The Logicof Practice. Stanford: Stanford University Press.

Brunache, Peggy L.2001 West AfricanLandscapesand MaterialCulture: An Archae-

ological Investigation of Intrasite Variation at Savi,Bénin. M.A.thesis, Department of Anthropology, University of South Caro-lina.

Chaudoin, E.1891 Trois Mois de Captivité auDahomey. Paris: Hachette.

Connah, Graham2000 AfricanCity Walls: A Neglected Source? In Africa’s Urban

Past. DavidM. Anderson and Richard Rathbone, eds. Pp. 36–51.Oxford: James Curry Press.

Curtin, Philip1975 Economic Change in PrecolonialAfrica: Senegambia in the

Era of the Slave Trade. Madison: University of WisconsinPress.

Darling, Patrick1984 Archaeology and History in Southern Nigeria: The Ancient

Linear Earthworks of Beninand Ishan, 2 vols. Cambridge: BAR In-ternational Series 216.

de Heusch, Luc1989 Kongo inHaiti:A New Approach to Religious Syncretism.

Man, n.s. 24(2):290–303.DeCorse, Christopher

1998 Culture Contact and Change in West Africa. In Studies inCulture Contact: Interaction, Culture Change, and Archaeology.James G. Cusick, ed. Pp. 358–377. Carbondale: Southern IllinoisUniversity.

2001 An Archaeology of Elmina: Africans and Europeans on theGold Coast, 1400–1900. Washington, DC: Smithsonian Institu-tion Press.

DeMarrais, Elizabeth, Luis Jamie Castillo, and Timothy Earle1996 Ideology, Materialization, and Power Strategies. Current An-

thropology 37(1):15–31.Des Marchais, Etiénne

1731 Voyage du Chevalier Des Marchais en Guineé. Amsterdam:Aux Dépens de la Compagnie.

Diggs, Irene1953 Zumbi and the Republic of OsPalmares. Phylon 14(1):62–70.

Duncan, John1847 Travels in Western Africa in1845 and 1846: Comprising a

Journey from Whydah, through the Kingdom of Dahomey, toAdofoodia, in the Interior, vol. 1. London: Richard Bentley.

Fleisher, JeffIn press Behind the Sultan of Kilwa’s “Rebellious Conduct”: Local

Perspectiveson an InternationalEast African Town. In AfricanHistorical Archaeologies. Paul Lane and Andrew Reid, eds. Klu-wer Academic/Plenum.

Forbes, Frederick E.1966 Dahomey and the Dahomans, vol. 1. London: Frank Cass

and Company.Fox, John G.

1996 Playing with Power: Ballcourts and Political Ritual inSouth-ern Mesoamerica. Current Anthropology 37(3):483–509.

Goody, Jack1957 Fields of Social Control among the Lo Dagaba. Journal of the

Royal Anthropological Institute of Great Britain and Ireland87(1):75–104.

Greene, Sandra E.2002 Sacred Sites and the Colonial Encounter: A History of Mean-

ing and Memory and Meaning inGhana. Bloomington: IndianaUniversity Press.

Guyer, Jane I.1995 Wealth in People, Wealth in Things: Introduction. Journal of

AfricanHistory 36(1):83–90.Hall, Martin

1987 Archaeology and Modes of Production inPre-ColonialSouthern Africa. Journal of Southern AfricanStudies14(1):1–17.

2000 Archaeology and the Modern World: Colonial Transcripts inSouth Africa and the Chesapeake. London: Routledge Press.

Hantman, Jeffrey L.1990 Between Powhatan and Quirank: Reconstructing Monacan

Culture and History in the Context of Jamestown. American An-thropologist 92(3):676–690.

Herskovits, Melville J.1938 Dahomey: An Ancient West AfricanKingdom, volume 2.

NewYork: J. J. Augustin.Herskovits, Melville J., and Frances S. Herskovits

1933 An Outline of Dahomean Religious Belief. Vol. 41. Menasha,WI: American AnthropologicalAssociation.

Hodder, Ian1995 Theory and Practice inArchaeology. London: Routledge

Press.Holley, George R., Rinita A. Dalan, and Philip A. Smith

1993 Investigations in the Cahokia Site Grand Plaza. AmericanAn-tiquity 58(2):306–319.

Joyce, Arthur A., and Marcus Winter1996 Ideology, Power, and Urban Society in Pre-HispanicOaxaca.

Current Anthropology 37(1):33–86.Kelly, Kenneth G.

1995 Transformation and Continuity inSavi, a West African TradeTown: An Archaeological Investigation of Cultural Change on

Norman and Kelly • Landscape Politics in West Africa 109

the Coast of BéninDuring the 17th and 18th Centuries. Ph.D. dis-sertation, Department of Anthropology, University of California,Los Angeles.

1997a The Archaeology of African-European Interactions: Investi-gating the Social Roles of Trade, Traders, and the Useof Space inthe Seventeenth- and Eighteenth-Century Hueda Kingdom, Re-publicof Bénin. WorldArchaeology 28(3):351–369.

1997b Using Historically Informed Archaeology: Seventeenth andEighteenth Century Hueda/ European Interaction on the Coastof Bénin. Journal of ArchaeologicalMethod and Theory4(3–4):353–366.

2001 Change and Continuity in CoastalBénin. In West Africa Dur-ing the AtlanticSlave Trade: Archaeological Perspectives. Christo-pher R. DeCorse, ed. Pp. 81–100. London: Leicester UniversityPress.

2002 Indigenous Responses to ColonialEncounters on the WestAfricanCoast: Hueda and Dahomey from the Seventeenththrough Nineteenth Century. In The Archaeology of Colonial-ism. Claire L. Lyons and John K. Papadopoulos, eds. Pp. 96–120.Los Angeles: Getty Research Institute.

Kelly, Kenneth G., Peggy L. Brunache, and Neil L. Norman1999 Preliminary Report on 1999 FieldSeason atSavi,Republicof

Bénin. Nyame Akuma 52:2–10.Kolb, Michael J.

1994 Monumentality and the Riseof Religious Authority in Pre-contact Hawai’i.Current Anthropology 34(5):521–547.

Koontz, Rex, Kathryn Reese-Taylor, and Annabeth Headrick, eds.2001 Landscape and Power inAncient Mesoamerica. Boulder:

Westview Press.LaViolette, Adria

In press Swahili Archaeology and History on Pemba Island, Tanza-nia: A Critique and Case Study of the Useof Written and OralSources inArchaeology. In AfricanHistorical Archaeologies. PaulLane and Andrew Reid, eds. Kluwer Academic/Plenum.

Law, Robin1986 Dahomey and the Slave Trade: Reflections on the Historiog-

raphy of the Rise of Dahomey. Journal of African History27(2):237–267.

1989 “My Head Belongs to the King”: On the Political and RitualSignificance of Decapitation inPre-Colonial Dahomey. Journalof African History 30(3):399–415.

1990 “The Common People Were Divided”: Monarchy, Aristoc-racy, and Political Factionalisminthe Kingdom of Whydah,1671–1727. International Journalof African Historical Studies23(2):201–229.

1991 The Slave Coast of West Africa,1550–1750. Oxford: OxfordUniversity Press.

1992 Warfare on the West AfricanCoast, 1650–1850. In War inthe Tribal Zone: Expanding States and Indigenous Warfare. R.Brian Ferguson and Neil L. Whitehead, eds. Pp. 103–126. SantaFe, NM: School of American Research Press.

Lawrence, Denise L., and Setha M. Low1990 The Built Environment and Spatial Form. Annual Review of

Anthropology 19:453–505.McIntosh, Susan K.

1999 Pathways to Complexity: An AfricanPerspective. In BeyondChieftans: Pathways to Complexity inAfrica. Susan K. McIntosh,ed. Pp. 1–30. Cambridge: Cambridge University Press.

Mbembe, Achille1992 The Banality of Power and the Aesthetics of Vulgarity in the

Postcolony. PublicCulture 4(2):1–30.Miller, David, and Christopher Tilley, eds.

1984 Ideology, Power, and Prehistory. London: Cambridge Uni-versity Press.

Miller, Joseph C.1980 Introduction: Listening for the African Past. In The African

Past Speaks: Essays on Oral Tradition and History. Joseph C.Miller, ed. Pp. 1–59. Hamden, CT: Archon.

Murphy, John C., and Robert W. Henderson1997 Tales of Giant Snakes: A Historical Natural History of Ana-

condasand Pythons. Malabar, FL: Krieger Publishing Company.

Newbury, C. W., ed.1966 A Mission to GeleleKing of Dahome by SirRichard Burton.

London: Routledge.Norman, Neil

2000 Through the Medium of the Vessel: An EthnoarchaeologicalInvestigation of Ritual Earthenware in Southern Bénin, West Af-rica. M.A. thesis, Department of Anthropology, University ofSouth Carolina.

Orr, Heather S.2001 Procession Rituals and Shrine Sites: The Politics of Sacred

Space in the Late Formative Valley of Oaxaca. In Landscape andPower in Ancient Mesoamerica. Rex Koontz, Kathryn Reese-Taylor, Annabeth Headrick, eds. Pp. 55–79. Boulder: WestviewPress.

Piot, Charles1999 RemotelyGlobal:Village Modernity in West Africa. Chicago:

University of Chicago Press.Piqué, Francesca, and Leslie H. Rainer

1999 Palace Sculpturesof Abomey: A History Told on Walls. LosAngeles: Getty Conservation Institution.

Polanyi, Karl1966 Dahomey and the Slave Trade: An Analysisof an Archaic

Economy. Seattle: University of Washington Press.Posnansky, Merrick

1981 Notsé Town Wall Survey. Nyame Akuma 18:56–57.1985 Togo and New Directions in the Study of WestAfrica’s Past.

Actes du Séminaire UCLA-UB sur les Sciences Sociales, Universitédu Bénin, Lomé.

Quarcoopome, Nii. O.1993 Notse’s Ancient Kingships: Some Archaeological and Art-

Historical Considerations. African Archaeological Review11:109–128.

Sahlins, Marshall1985 Islands of History. Chicago: University of Chicago Press.

Schmidt, Peter R.1996 Using Archaeology to Remake History inAfrica. In Making

Alternative Histories: The Practice of Archaeology inNon-Western Settings. Peter R. Schmidt and ThomasC. Patterson, eds.Pp. 119–148. Santa Fe, NM: SAR Press.

Shinnie, Peter L.1971 The African Iron Age. Oxford: Clarendon Press.

Smith, William1744[1967] A NewVoyage to Guinea. London: Frank Cass and Co.

Spieth, Jakob1911 Die Religion der Eweer in Süd-Togo. Leipzig: Dieterich’sche

Verlagsbuchhandlung.Stahl, Ann

2001 Making History in Banda: AnthropologicalVisions of Af-rica’s Past. Cambridge: Cambridge University Press.

Thomas, Julian1993 The Politics of Vision and the Archaeologies of Landscape. In

Landscape: Politics and Perspectives. Barbara Bender, ed. Pp.19–49. Providence, RI: Berg.

Tilley, Christopher1994 A Phenomenology of Landscape: Places, Paths, and Monu-

ments. Oxford: Berg.Trouillot, Michel-Rolph

1995 Silencing the Past: Power and the Production of History. Bos-ton: Beacon Press.

Turner, Victor1974 Drama, Fields, and Metaphors: SymbolicAction inHuman

Society. Ithaca, NY: CornellUniversity Press.1995 The Ritual Process: Structure and Anti-Structure. New York:

Aldine de Gruyter.Villiers, André

1975 LesSerpents de L’ouestAfricain. Dakar: Université de Dakar.Wallerstein, Immanuel

1986 Africa and the Modern World. Trenton: Africa WorldPress.Webster, Gary S.

1990 Labor Control and Emergent Stratification in PrehistoricEurope. Current Anthropology 31(4):337–366.

110 American Anthropologist • Vol. 106, No. 1 • March 2004