The French word récade is a neologism formed from the Portuguese word recados, meaning “message” or “messenger. The word récade specifically refers to a type of object used in the Fon kingdom of Dahomey (seventeenth to nineteenth centuries ce, now part of the Republic of Benin). Récades often are axe- or crook-shaped staffs. Their handle is often made of wood and is roughly 50 cm high. The blade of an axe-shaped récade is often sculpted in the shape of a king or of a battalion's emblems, figures that often refer to quotes or proverbs (Fig. 1).When used by royal messengers, the récade authenticated the provenance of their message. The word récade also refers to formally similar staffs used by priests of the Fon deity of thunder, So, also known by the names Hevios(s)o, Hevies(s)o, Hebios(s)o, Hebies(s)o, or Djis(s)o. These two kinds of object were also used in the context of ritual dances.In Dahomey art historical studies, the dominant view of the origin of the récade is the one first recorded by Adande (1962; cf. recently Beaujean 2015). According to the tradition recorded in this work, the récade originated during the reign of Wegbaja (ca. 1645-1685), the first king of Dahomey. The ancestors of the Fon people, taken by surprise by their enemies during their agricultural work, used their hoe handles as ad hoc weapons to fight off their assailants. This tool was later used in parades as a symbol of Fon bravery, a reminder of their victory, and then as a symbol of royal authority and messengers.In this paper, I challenge this widely accepted hypothesis. I suggest that while derivation of some récades from battle weapons is not unlikely, there are certainly other sources for this very diverse type of object. My position is that the récade was functionally inspired by at least two different kinds of staff. One is a staff called opa ase. It is widely used in Yorubaland and neighboring areas by royal messengers. The other one is the ose Sango, the staff used by priests of the thunder cult in the Yoruba empire of Oyo. Dahomey was tributary to Oyo between the first part of the eighteenth century and the early nineteenth century. It was deeply influenced by Oyo in many respects, including arts, religion, and political structure. One such influence dealt with the thunder deity Sango, who was associated with Oyo's expansionism. I suggest that as Sango was considered the ancestor of the Oyo rulers, who were antagonistic to Dahomey kings, the latter took the association of the thunder deity with expansionism and applied it to So, the Fon deity of thunder.It is known that Sango priests were part of Oyo diplomatic and tribute collecting missions through the empire and beyond. It was customary for Sango priests to take with them their ose Sango, a kind of staff associated with the cult of Sango as their badge of office during their travels. I suggest that a similar practice was followed in Dahomey, where I claim priests of the local thunder deity traveled with the staff of the deity, the sokpo, along with ambassadors, who traveled with the wensagunkpo, the local royal messenger staff. This diplomatic context resulted an influence of the sokpo over the wensagunkpo, hence the formal similarity between both kinds of staff and their collective designation as récades.I will first explain why the widespread hypothesis of récades originating from hoe handles used as weapons is questionable. I note that Oyo diplomatic missions were composed of Sango priests holding ose Sango and ambassadors holding opa ase. I conclude by showing how this pattern was followed in the kingdom of Dahomey and provided a context for the influence of the sokpo over the wensagunkpo, resulting in their formal similarity, hence their reference as récades despite their functional dissimilarities.In his influential work on récades, Alexandre Adande (1962: 14) explains the origins of récades as follows (my translation from the French):I do not deny that battle weapons may have influenced the shape of some récades. This could explain the name makpo (literally “staff of anger”) sometimes given to récades. However, I claim that this explanation is insufficient. A first problem with Adande's explanation is the association of the first appearance of the récade with the reign of King Wegbaja. As shown by Law (1988), Wegbaja, who is traditionally considered the founder of the Dahomey kingdom, became a kind of cultural hero to whom many innovations from before or after his reign are abusively attributed.A more crucial issue with Adande's explanation is the fact that it fails to explain the formal similarities between the récade as a messenger staff and as a staff of the priesthood of So, of all deities. I suggest that in order to properly understand the formal sources of the récade, one has to rely on the indigenous designations of the artifacts this word refers to. Rather than using the French word récade for both messenger staffs and staffs of So priests, I will use the respective Fon words for them, namely wensagunkpo and sokpo.In a work hardly cited in Dahomey art historical studies, Widstrand (1958: 120) has provided an explanation for the formal similarity between both types of récades. According to him, the thunder deity staff formally influenced the messenger staff. I agree with this statement. However, Widstrand does not provide arguments supporting his claim. He also mentions the elevation of the cult of So as a “vodun officiel” under the rule of King Glele (1858— 1889) and his introduction in the kingdom during the rule of Tegbesu (1740-1772). He seems to connect those to the influence of the sokpo over the wensagunkpo. Again, I agree with the latter, but I consider this explanation and context to be insufficient.As we have seen, Adande claims that the tradition of sculpting the blade of axes in the form of an emblem originated in the kingdom of Dahomey during the early seventeenth century. This is all the more problematic as the sokpo, a staff whose blade is sculpted in the form of the emblem of the thunder deity, is found in a much wider area than the kingdom of Dahomey and seems to predate its foundation. The cult of So is indeed common to Gbe-speaking people, among whom are the Fon people of Dahomey. Gbe- speaking people largely share a pantheon of deities that probably predates their separation as speakers of the same language.1Widstrand (1958: 119) shows several sokpo from Togo, that were presumably from outside the scope of the Dahomey kingdom. The latter was indeed mostly located in what is now the south of the Benin Republic. Widstrand also noted a peculiar feature that distinguishes some sokpo from the kingdom of Dahomey from all other sokpo. Many sokpo are wooden curved staffs with a blade sculpted as the emblem of So, that is, a horizontal stem at the end of which is found a crescent whose points go backwards. In some Fon sokpo, the curved part of the staff has been reinterpreted as the head of an animal called agbo in the Fon language. This word can refer to either a ram, a buffalo, or an antelope. It is is thought by the Fon and other Gbe-speaking people to be the incarnation of the deity So. The symbol of So looks like it is is inserted into the animal's mouth. The root of the stem is sometimes linked to the points of the crescent by two zigzag lines that seem to spread from the animal's mouth. In Dahomey, this type of sokpo was only used by high-ranking So priests (Adande 1962: 19).If the sokpo was originally from Dahomey and then spread among western Gbe speakers, one could wonder why the elaborated Fon version of the staff has not reached the western Gbe area, like the simpler version has. It seems more probable that this type of artifact was originally shared by many Gbe-speaking people and that the Fon innovated with the more elaborated form afterwards. Another fact suggesting the antiquity of this axe-shaped sokpo is the widespread association, in West Africa, between thunder cults and the simple axe.2 Since this sokpo resembles a simple axe, one can assume that its shape predated the existence of the similar wensagunkpo.The Yoruba state of Oyo was founded around the tenth century in what is now the western part of central Nigeria. It progressively rose to prominence thanks to its strategic position in the trade route linking the savannah and forest states, becoming the most powerful state of the Yoruba-Edo area with the kingdom of Benin by the sixteenth century. From this period onward, Oyo became an imperial power, progressively forcing a large number of states from the Yoruba- and Gbe-speaking areas to pay tribute.The imperial cult of Sango was key to this aspect of Oyo political organization. Oyo ambassadors, messengers, or tribute collectors sent in missions outside or inside the Oyo empire were royal slaves called ilari. This name, which means “scarified head” in Yoruba, is a reference to the fact that half of their head was completely shaved. The ilari were either Sango priests or were accompanied by them in those duties (Morton Williams 1964: 255). One reason for this is that Sango priests were able to manipulate thunder and lightning. Using this threat, they were able to coerce people into paying expensive tributes in a region particulary stricken by lightning (Losi 1923: 4, Schütz 2009: 81).During their official duties, Sango priests brought their own ose Sango, the staff identifying them as members of Sango's priesthood (Lawal 1970: 57, Willis 2017: 51) (Fig. 3). The staff of the thunder deity priests was thus used in diplomatic contexts in the Yoruba empire of Oyo.However, the ose Sango was not the only staff in those diplomatic contexts among the Yoruba. A staff called by various names, such as opa ase (“staff of authority”) or opa iranse (“staff of the messenger”), was used in different areas of Yorubaland. Similarly to the wensagunkpo, it was used by messengers conveying official messages or performing duties ordered by their rulers. Despite the ruler being absent, the presence of opa ase symbolized the presence of the king. The staff was thus owed the same ritual deference one would show the ruler. The form of the opa ase seems to have been very variable following context and areas of use (Fig. 4). For example, Oroge (1971), referring to the emese (the equivalents of the ilari in the Yoruba state of Ife), noted thatSince they were used in diplomatic contexts, one could expect the ose Sango to have been used as the local form of the opa ase. With the ruler of Oyo considered as an incarnation of Sango, the ose Sango could conceivably have symbolized his presence as well as that of the deity. However, it definitely does not seem to have always been the case. Lawal (2000: 106) explains that the opa ase used in the Oyo empire was “usually a beaded staff surmounted by a bird or equestrian figure motif, ” a description that points to a staff formally different from an ose Sango.Further evidence suggests that staffs distinct from ose Sango were used by the ilari in Oyo. It can be found in the testimony of Richard Henry Stone, an American Baptist missionary who visited the Yoruba state of Ijaiye in the late 1850s. Ijaiye was, along with Ibadan, one of the successor states of the Oyo Empire, whose capital city, Oyo-Ile, was razed to the ground in 1835 and its power considerably reduced. In his published memoirs, Stone has noted the use of opa ase, apparently of different forms depending on the ruler who used it. According to Stone (2010: 55), Kurunmi, the ruler of Ijaiye, used a “crook of brass” as an opa ase. This description does not seem to point to an ose Sango either. Yet, Kurunmi had in a way appropriated the cult of Sango, whose elder son and first representative was not the Oyo emperor anymore, but Kurunmi himself (Peel 2000: 83-84). One can consider that this ruler who considered himself as Sangos successor would have used an ose Sango as an opa ase if it had systematically been the case in Oyo.Here we have seen that the ose Sango, the staff used by the priests of the thunder deity, and a distinct staff, the opa ase, were both used in the same diplomatic context by Oyo messengers. This will be instrumental in determining the reason for the formal similarity between the sokpo and the wensagunkpo and their common reference as récades, despite their functional dissimilarities.The kingdom of Dahomey became tributary to Oyo under the reign of the Fon king Agaja (1718-1740). The first clash between the two states occurred in 1728 after the conquest by Dahomey of the small but rich coastal states of Aliada and Whydah (now Benin), which controlled the slave trade in the area from the ports of Jakin and Savi. Aliada and Whydah were themselves tributary to Oyo before their conquest by Dahomey. This conquest clashed with Oyo's interests in the area. Oyo invaded Dahomey several times until the latter was subjugated in 1740. Dahomey remained a vassal to Oyo until 1824, when King Gezo (1818-1858) finally liberated Dahomey after defeating an Oyo army.During its political domination by Oyo, Dahomey was deeply influenced by Oyo in many respects. Among them, Bay (1998) has noted a number of features found at the court of the Fon king Tegbesu (1740-1774) that were not mentioned at his predecessors’ court. Tegbesu is remembered by tradition as having been raised at the Oyo court as a hostage and to have brought several innovations from there to Dahomey's court.The first is the mysticization of the king, who only appeared in public once a year, through its reclusion. This feature is found in Oyo and widespread in the rest of Yorubaland, where it is associated to the ruler's sacredness.The second is the creation of the office of an important dignitary called sogan, the literal translation of olokun esiti (literally, “master of horses”), which was applied to an important dignitary in the Oyo empire. Bay noted that the reference to “master of horses” in Dahomey could not be explained if the title was indigenous to Dahomey. Indeed, horses were much less used and institutionalized in Dahomey than they were in Oyo, whose powerful cavalry distinguished it from the more southern forest states it was in contact with.The third is the use of royal messengers and tribute collectors with a hairstyle identical to that of the Oyo messengers. In Dahomey, they were called lati (a phonological adaptation of the Yoruba word ilari) and wensagun (the Fon word for “messenger”).Pruneau de Pommegorge, a former director of the French fort of Whydah in the 1760s, wrote in his Description de la Nigritie (1789: 168-70) of a “half-head” sent by Tegbesu on his way of delivering a message. The messenger is holding a wensagunkpo, there only described as a “wand” (“une canne”).Yet another feature is recorded by tradition to have been introduced in Dahomey's court during the eighteenth century. The oldest source recording this introduction specifically associates it with the reign of Tegbesu (Le Hérissé 1911: 108). Like the cult of Sango in Oyo, it was associated with thunder and lightning. It became the most important public deity of the kingdom, associated not only with with thunder, lightning, and justice, but also with royalty and with Dahomey's domination. Pares (2005) has noted how this cult in Dahomey was strikingly similar to that of Sango in Oyo.The timing of the elevation of this cult as a deity with functions similar to those of Sango in Oyo suggests that it indeed resulted from an Oyo influence. Despite having been inspired by Oyo, the thunder cult introduced in Dahomey was not that of Sango, but that of the indigenous Fon thunder deity So. One can wonder why Sango was not adopted as such by the kings of Dahomey. Sango was specifically associated with Oyo and its emperor, who was considered the deity's incarnation. I suggest that the kings of Dahomey could not associate themselves with Sango, the tutelary god of their rivals, the emperors of Oyo. They thus used their own thunder deity, So, in this respect.This kind of contestation of Oyo's domination can be found elsewhere in Dahomey royal art. One of them can be found on a bas-relief of the main palace of King Gezo in Abomey, the former capital of the kingdom (Fig. 4). It depicts a monkey eating an ear of corn and is said to be a reference to Oyo's greed in collecting expensive tribute from Dahomey. This caricature of the Oyo ruler seems to have been inspired by Yoruba sculpture, where the figure of the seated monkey eating an ear of corn is often found.3Similarly, Blier (2003: 138-39) has suggested that the thrones called gandeme in the Fon language were inspired by those used by the Aja and Ewe people, who like the Fon claim a common origin in the precolonial kingdom of Tado. She believes that the use of gandeme began under the reign of king Agonglo (1789— 1797), perhaps as a political reaction to the influence of Oyo, from whom Dahomey wanted emancipation.Like symbols of Sango in Oyo, So was frequently depicted in Dahomey's art. There he was represented as a ram holding in its mouth a stem topped by a crescent (Fig. 5). This symbol represented both fire and a stone celt. This depiction of the deity as a ram contrasted with Sangos iconography, where the ram was typically depicted in a submissive position, complementing Yoruba myths that describe this animal as Sangos antagonist or favorite sacrificial victim (Bascom 1969: 341-43; Verger 1957: 313; Sachnine 1996: 112-13).4The artifact most associated with Sango is the ose Sango. This is a wand shaped like a double axe. The design of an ose Sango is quite variable. Some of them also include figurative characters.The ose Sango appears in Yoruba art as an artifact and in other media as well (Lawal 1970; Folaranmi 2015). However, it never appears in Dahomey royal art. Instead, the sokpo is frequently found in this art, both as artifacts and as depictions in other media (Fig. 6). It seems that the functional influence of the Sango cult over that of So in Dahomey came alongside the replacement of ose Sango by the sokpo in similar contexts.As we have seen, the use of ilari in Dahomey is the result of influence from Oyo during the reign of Tegbesu. This influence was contemporary to the elevation of So as a royal deity, which results from the influence of Sango's status in Oyo.Since the ilari in Oyo were initiated to Sangos worship or accompanied by Sango priests in their diplomatic duties, one can expect this association to have been replicated in Dahomey, only with Sango replaced by So, royal messengers in Dahomey being initiated to So's cult or accompanied by So priests bringing their sokpo on their duty.This seems to be confirmed by the depiction, in a French journal, of a diplomatic mission sent by King Behanzin (1889-1894) of Dahomey in 1893 (Fig. 7). There, three bare-chested men appear sitting on their knees on the lower row of the picture. Described as “slaves, ” they are the only unnamed characters of the picture. The rightmost of these three characters holds a sokpo in his left hand. The stem and the points of crescent of the So symbol are linked by two rows of zigzags. As we have seen, this feature is is characteristic of high-ranking So priests’ wands.In this same picture, a character named Chedingen is shown in the middle row of the picture holding a wensagunkpo. The curved part of the staff is sculpted as what appears to be a shark, the symbol of King Behanzin. Chedingen is described as “the head of the mission.”5A sokpo and a wensagunkpo were thus used in this diplomatic mission. This situation parallels the one previously mentioned in Oyo, where ose Sango and opa ase were both brought on diplomatic missions.I have no direct evidence of both sokpo and wensagunkpo being brought on Dahomey diplomatic missions other than this depiction of Behanzin's mission in France. It is however interesting to note that this practice in Dahomey could be predicted from indirect evidence.I suggest that the practice of sculpting the blades of the wensa- gunkpo as royal emblems took place within this diplomatic context. There the sokpo, which had their blade already sculpted as the emblem of the Gbe-speaking thunder deity, influenced the form of the wensagunkpo. The origin of the functional use of the wensagunkpo in Dahomey is unclear, but the use of staffs with a similar function is widespreadv in the region, especially in Yorubaland, and very likely predated the rise of Dahomey in the early seventeenth century.6Since the use of ilari is only attested in Dahomey during Tegbesu's reign, one can possibly assume that Dahomey kings used such staffs at this time, although ilari were also attested in the kingdom of Whydah earlier in the eighteenth century, probably following their domination from Oyo.I have suggested a new hypothesis about the origin of the récade. In order to do so, I have questioned the word récade as a relevant appellation for all the artifacts it refers to. I have used the words wensagunkpo and sokpo, which are the words for two kinds of recades, namely the messenger staff and that of the thunder deity. I have suggested that the form of the latter influenced the form of the former in their use in the context of diplomacy. This use was inspired by the situation in Oyo, where functionally equivalent staffs were used in similar contexts. In order to express their antagonism to Oyo, the Fon elevated their own thunder deity, So, to a status similar to that of Sango, the thunder-linked imperial deity of Oyo. This antagonism was also manifested in art. The staff of Sango priests, which was used in diplomatic missions in Oyo, was replaced by sokpo in Dahomey. This context, where wensagunkpo were also used, spurred the influence of the latter on the former, resulting in their formal similarity and their collective reference as récades. My hypothesis of sokpo being brought on Dahomey diplomatic missions so far, however, only rests on one direct piece of evidence, which is the depiction of a whole Fon diplomatic mission from Dahomey, although this is also suggested by indirect evidence. I hope future research will shed light on this hypothesis.