Umbanda: Africana or Esoteric?

Umbanda is a dynamic and varied Brazilian spirit-incorporation tradition first recorded in the early twentieth century. This article problematizes the ambiguity of categorizing Umbanda as an ‘Afro-Brazilian’ religion, given the acknowledged centrality of elements of Kardecist Spiritism. It makes a case that Umbanda is best categorized as a hybridizing Brazilian Spiritism. Though most Umbandists belong to groups with strong African influences alongside Kardecist elements, many belong to groups with few or no African elements, reflecting greater Kardecist influence. Kardecist elements are universal in Umbanda; Africana elements are not. Kardecism is a western esoteric tradition and a key factor in the emergence of a wide variety of Latin American Spiritisms, including Umbanda. Labelling Umbanda as ‘Afro-Brazilian’ fails to acknowledge the beliefs and practices of the many Brazilians who practice other forms of the tradition. This case study of the categorization of religions – looking at scholarly ideology, race, historical origins, doctrine and practice – emphasizes the need to look at specific contexts and to avoid broad generalizations.

Umbanda is a dynamic and varied Brazilian spirit-incorporation tradition first recorded in the early twentieth century. This article problematizes the ambiguity of categorizing Umbanda as an 'Afro-Brazilian' religion, given the acknowledged centrality of elements of Kardecist Spiritism. It makes a case that Umbanda is best categorized as a hybridizing Brazilian Spiritism. Though most Umbandists belong to groups with strong African influences alongside Kardecist elements, many belong to groups with few or no African elements, reflecting greater Kardecist influence. Kardecist elements are universal in Umbanda; Africana elements are not. Kardecism is a western esoteric tradition and a key factor in the emergence of a wide variety of Latin American Spiritisms, including Umbanda. Labelling Umbanda as 'Afro-Brazilian' fails to acknowledge the beliefs and practices of the many Brazilians who practice other forms of the tradition. This case study of the categorization of religions -looking at scholarly ideology, race, historical origins, doctrine and practice -emphasizes the need to look at specific contexts and to avoid broad generalizations.
important precursors, including the works of Andrew Jackson Davis. Today, it is most prominent in English-speaking countries -often in 'spiritualist churches ' -and in Iceland (Cox, 2003;Dempsey, 2016). Spiritualist groups are rare Latin America. In Brazil, for example, 61,739 people (0.03% of the total population) self-identified as Spiritualist on the 2010 census; the number of Kardecists was 3,848,876 (2.0%); the number of Umbandists was 407,331 (0.2%) (IBGE, n.d.).
'Kardecism' refers to French Spiritisme and its foreign offshoots. It was a historical offshoot of Spiritualism, sparked by French fascination with the 'turning tables' of séances, starting in 1854 (Aubrée and Laplantine, 1990;Sharp 2006;Monroe, 2008).
It remains rooted in and faithful to the nineteenth-century works of Allen Kardec and other French auteurs spirites. Kardec presented his books as a set of questions answered by highly evolved spirits who incorporated in mediums. Kardecism is found throughout Latin America, as well as in Europe and among Brazilian emigrant groups around the world. Latin American Kardecism differs from European forms in its greater emphasis on healing and this tendency to hybridize with other traditions (Engler and Isaia, 2016).
Key beliefs of Kardecism were influenced by esotericism, Christianity and perhaps Buddhism, along with nineteenth-century French Catholicism, liberalism and positivism (Aubrée and Laplantine, 1990;Engler and Isaia, 2016). God created all spirits as equals in an ignorant, confused state, spiritually distant from God. The right and natural trajectory of all spirits is to return to God by evolving spiritually over many lifetimes on this and other worlds. Charity is a key measure of spiritual evolution. Highly evolved spirits no longer have the need to perform expiation while incarnate. Sometimes these spirits are identified with historical figures, including well-known Kardecists of previous generations. These spirits work with mediums, often through automatic writing, in order to pass on spiritual teachings, helping their less advanced fellow spirits with their spiritual evolution. Jesus was a highly evolved spirit who incarnated voluntarily in order to pass on spiritual teachings. There are no demons or angels, no purely evil or good beings, only God and the hierarchy of spirits.
'Possession' by spirits is impossible, because each soul is unassailably in control of its Engler: Umbanda 5 body. Less evolved spirits can influence our thoughts through resonances between our spiritual or psychological weaknesses and theirs. These confounding spirits are not exorcized but reoriented and educated through rituals of ' disobsession.' Other rituals include the reception of messages from departed loved ones and the passe -a direct echo of Mesmerism and its ' animal magnetism' -in which an expert adjusts the 'magnetic fluids' of a patient using non-contact hand movements. Some Umbanda groups distinguish between the basic passe and the more powerful 'magnetic passe' (Engler, 2016: 210).
'Esotericism' refers to Western Esotericism, ranging from ancient Hermetism, Gnosticisms, Astrology and Alchemy, through Kabbalah, Renaissance ritual magic, Swedenborgianism, John Dee's Enochian magic, Rosicrucianism and High-Grade Freemasonry, to New Thought, Theosophy, UFO religions, New Age and modern ritual magic. Scholars argue over how to define and delimit the scope of 'Western Esotericism' (Faivre, 2012(Faivre, [1992; von Stuckrad, 2005;Hanegraaff, 2012). Faivre's list of six characteristics is useful here (2012 [1992]: 20-21, italics added): 1. 'universal correspondence … non-"causal" matches between all the levels of reality in the universe … relationships between the (macrocosm) and humanity (microcosm) …'; 2. 'Living Nature … [T]he whole of Nature, considered as a living organism … has a history, linked to that of humanity and the divine world'; 3. 'mediation and imagination … [R]ituals, symbols …, intermediary spirits, etc. … present themselves as so many mediations likely to ensure passages between the various levels of reality… [T]he ' active' imagination … exercised on those mediations, makes them a tool for knowledge …'; 4. 'The experience of … the transmutation of oneself …, and, as a corollary, of part of Nature (in many alchemical texts, for example)'; 5. 'concordance: common denominators between several different [esoteric] traditions are assumed a priori'; 6. 'the importance of "channels of transmission"; for example, master to disciple, from initiator to initiate'.

Engler: Umbanda 6
This is a list of family resemblance. Few types of western esotericism have all six characteristics, but each has several. By this measure, Kardecism is an esoteric tradition. Kardecism has a developed and coherent esoteric doctrine. The mesmeric doctrine of magnetic fluids explains the resonances between spirits and with God (a limited form of correspondence). The charitable interventions of highly evolved spirits assist in this passage, over many lifetimes, between levels of reality (mediation). This ascension takes place as we lower, incarnate spirits evolve spiritually (transmutation). The truths of Kardecism stem from the teachings of highly evolved spirits (a view of esoteric transmission similar to John Dee's earlier work with angel magic and Theosophy's later doctrine of ascended masters). Esoteric transmission overlaps with historical transmission in Kardecism's view that the works of the nineteenth-century French masters is supplemented by the teachings of newer generations of Kardecist leaders, who send their texts from beyond the grave via automatic writing.
In addition to this strong fit with Faivre's typology, Kardecism has direct historical relations to other esoteric traditions. Beyond its foundational relation to Spiritualism, it was greatly influenced by Mesmerism in its emphasis on the relation between health and the balance of magnetic fluids, in its ritual manipulation of these fluids, and in its emphasis on alternative personalities during trance states, prominent in the work of Anton Mesmer's follower, the Marquis de Puységur. Puységur's work was developed by Swedenborgians, who held that spirits of the dead spoke through this trance mediumship (Goodrick-Clarke, 2008: 178). Kardecism, in turn, influenced later esoteric traditions, including Theosophy, modern ritual magic, UFO religions and the New Age movement. Kardecism is a western esoteric tradition in both its beliefs and practices (Bergé, 2005).
'Esoteric Umbanda' points to a limited subset of Umbandist groups. As discussed further below, this label is best used for groups that include esoteric elements other than Kardecist ones, for example, New Age healing rituals.
'Spiritism' is used in a range of ways by both insiders and scholars. I use it as an umbrella term for spirit-incorporation traditions with western esoteric elements. Not Engler: Umbanda 7 all spirit-incorporation traditions are Spiritisms, such as Africana religions with no esoteric elements. There are Spiritisms outside Europe and the Americas, such as Caodai, which emerged in early-twentieth-century Vietnam, shaped by Kardecism, Buddhism, Taoism, Confucianism, Christianity and Islam (Oliver, 1976: 29-32). The limitation to western esotericism recognizes contingent historical relations that are central to the colonial (and decolonial) contexts of the Americas. This take on 'Spiritism' has no whiff of essentialism. In general terms, no concept or word has a single, correct, definitive meaning: each means what it means by virtue of its connections to other words and concepts, and these webs of semantic relations vary historically, culturally, institutionally and even between individuals (Gardiner and Engler, 2016). People come to agree -as we interpret each other -and the meaning of a given term is no more than the contingent result of those acts of interpretation. Meanings are not occult objects to be uncovered out there in the world, leading to fights about the one, true meaning. There are no such things as Candomblé, Spiritualism, Kardecism, esotericism and Spiritism, just as there is no such thing as Christianity. In each case there is a diffuse set of doctrinal, ritual and institutional groupings that coheres well enough that we can make sense to each other when we talk about them as a group.
I use 'Spiritism' to point to the fact that esoteric elements -Spiritualist, Kardecist, Theosophical, New Age, and so on -constitute a web of overlapping links between dozens, if not hundreds, of spirit-incorporation traditions, especially in the Americas. These are usually treated as belonging to separate categories. Almost no work recognizes commonalities between Spiritisms in the Americas. The rare work tending in this direction trips over the lack of an effective comparative framework.
(See, for example, the confusing discussion of Kardecism in June Macklins ' [1974] study of 'New England Spiritualism' and 'Mexican-American Spiritualism'.) My proposed delimitation of 'Spiritism' picks out a set of traditions that are related historically, doctrinally and ritually and that scholarship has failed to address as a group, in part because a common category was lacking. There is something here worth talking about. This article is meant as a contribution to that conversation.

Africana Umbanda
Umbanda's rituals involve the incorporation of spirits who offer religious services to largely non-Umbandist (primarily Catholic and Kardecist) ' clients,' often collectively known as 'the assistance.' Clients come for advice, medical and spiritual diagnoses, healing rituals, purifying and protecting blessings, and so on. The two most important types of spirits are caboclos (usually indigenous spirits) and pretos velhos (' old black' spirits, usually identifying themselves as former slaves). Other spirit types include boiadeiros (cowboys), crianças (children), marinheiros (sailors), malandros (rogues), ciganos (gypsies, known for their work with crystals in more esoteric groups) and sereias (mermaids) (Barros, 2011). Umbanda varies regionally within Brazil and also in terms of three additional dimensions: from relatively Afro-Brazilian terreiros (grounds) to relatively Kardecist centros (centers) of 'white Umbanda'; from terreiros/centros focused primarily on ritual consultation with spirits to those working with 'spiritual surgery' techniques, the latter groups overlapping with 'Esoteric Umbanda', which offers various healing rituals drawn from New Age traditions; and from traditional to Neo-Umbanda, the latter characterized by more reflexive and dynamic innovation in basic doctrine (Hale, 2009a;Engler, 2016;Engler, forthcoming-b;Espírito Santo, 2017). Umbanda, Candomblé and other Afro-Brazilian ritual traditions' (FBU, n.d.). If we take the majority as defining Umbanda, it is Afro-Brazilian, but many groups are not.
Insisting that the former represent the only ' authentic' Umbanda is a vicious circle.
Many important Brazilian scholars do not assert that Umbanda is Afro-Brazilian.
Duglas Teixeira Monteiro wrote that 'Umbanda derives from Kardecism, but it adopted numerous rites from other religions already existing in Brazil ' (1977: 67).
Lísias Nogueira Negrão acknowledged the centrality of Kardecism when noting that some groups are ' especially influenced by Kardecism, though it continues to influence the rest, which are closer to the Afro-Brazilian religions ' (1996: 367). For Renato Ortiz, 'Umbanda is not a black religion; … it is opposed to Candomblé'; 'If 'Candomblé' and 'Macumba' are African religions, the Spiritism of Umbanda is, on the contrary, a-I would say the-national religion of Brazil' (Ortiz, 1977: 43;emphasis in original;see Concone, 1987). Paula Montero's ethnography of healing and Yvonne Maggie's ethnography of authority in Umbanda also offer more nuanced views (Montero, 1985: 194;Maggie, 2001Maggie, [1977). Carlos Rodrigues Brandão underlines the attributive nature of the label when he classes Umbanda among 'ritual traditions of real or supposed Afro-Brazilian origin ' (1987: 44). There are, of course, various other Brazilian scholars who do not use this categorization.
One reason that Umbanda is categorized as Afro-Brazilian is that most fieldwork on Umbanda has been done in peripheral, lower-middle-class areas of São Paulo and, to a lesser extent, Rio de Janeiro. These particular groups are characterized by Africana elements (Engler, 2012: 26;see Hale, 2009a for an important exception).
This Africana sub-type of Umbanda dominates the literature. There are few studies of 'white' and ' esoteric' Umbandas and of the rich variety of regional variants.
Another reason for the categorization of Umbanda as Africana is ideological. In the 1930s, Umbanda expanded and consolidated during precisely the period when Brazilian intellectuals rejected previous negative assessments of racial mixing and 'miscegenation' and elaborated Brazil's 'myth of three races' with its ideology of 'racial democracy' (DaMatta, 1987: 58-85;Fernandes, 2007Fernandes, [1972Magnoli, 2009 : 143-162). Beginning in the 1960s, 'racial democracy' was criticized as racist, because the myth of a harmonious status quo obviously masked ongoing stark inequalities (Magnoli, 2009: 157-162). Candomblé, with its unifying emphasis on Africana identity and 'pure' African roots, became an important site of contestation, as examinations of race in Brazil became more critical in the 1960s and 1970s (Agier, 1992: 63, 76). Candomblé has served for decades as a rallying point for struggles over Afro-Brazilian culture and identity and against racism (Selka, 2007).
In this ideological context, there is a bias toward an emphasis on African elements within Umbanda's range of forms. Seeing Umbanda as African and not European is a valuable corrective in a colonial nation that brought in more African slaves than any other in the world (Slenes, 2010), one where racism continues to impact tens of millions of Afro-descendent Brazilians, in conditions that are too often read through the lens of U.S. society (Guimarães, 2001;Sansone, 2003;Pinho, 2005;Nascimento, 2006). These are pragmatic reasons to adopt the view that Umbanda is essentially Africana, independent of a fuller look at Brazil's many Umbandas. It is strategically effective if we assume that scholarship should or 'pardo' (black or brown) members is higher in Candomblé than the average in the national population of Brazil: that is, in racial terms, Candomblé is 'blacker' than Brazil as a whole. The proportion of 'white' members in Umbanda is higher than in the population as a whole: in racial terms, Umbanda is 'whiter' than Brazil, though not as 'white' as Kardecism. Catholicism has a very slightly lower number of negro or pardo members than the national population: it is significantly less 'white' than Umbanda. If we were to take the racial identity of self-identifying members as the sole criterion of Afro-Brazilianness, then Brazilian Catholicism is more Africana than Umbanda is. Clearly, questions of race cannot be resolved by census categories, but, equally clearly, Umbanda cannot be equated with Candomblé in terms of the racial identities of its members.
A fuller assessment of the nature of Umbanda is valuable because it underlines the creative elaboration and appropriation of religious resources by Brazilians. This insistence on people's agency shapes a different sort of advocacy. For example, Cristina Borges' work on Umbanda sertaneja in the north of the state of Minas Gerais -a regional tradition closely related to Quimbanda -reveals a powerful regional discourse of religious and cultural identity (2011). Categorizing all Umbanda as Africana obscures the religious creativity of regional popular traditions. Concluding that Umbanda is esoteric, to whatever extent, does not make it more European; it makes it more Brazilian.

Origins
The issue of Umbanda's origin might seem more useful for assessing the extent to which Umbanda is Africana. Umbandists and scholars tell many conflicting stories of Umbanda's origins. 2 This section reviews three Umbandist views and three scholarly views. In the end, the evidence is mixed. The issue of origins not particularly helpful for attributing any original or unitary nature to Umbanda.
A first insider origin story points to Africa. This is sometimes presented by Umbandists in mythological terms, for example by invoking an enslaved Angolan sorcerer who planted a powerful root, from which Umbanda practices at some specific terreiro later grew (Hale, 2009b: 228). These stories are suggestive, but there is no way to assess their relevance, in historical, ideological or mythical terms.
The most common insider origin story is Kardecist (Brown, 1994: 38-41;Hale, 2009b: 227). In 1908, in a sick boy, seventeen-year-old Zélio Fernandino de Moraes, was mysteriously cured of a serious illness. Kardecists explained that the cure resulted from the charitable intervention of a highly evolved spirit. His parents took him to a Kardecist ritual the next day, and a powerful spirit incorporated in Zélio The third insider origin story is esoteric in a broader sense. As Umbanda developed in the 1940s and 1950s, many Umbandists claimed a wider range of origins for their religion: African but also Kardecist, Indigenous (Guarani), Vedic, Egyptian, Lemurian, extraterrestrial, and so on (see Bastide, 1995: 445-47;Oliveira, 2008: 114-19;Cumino, 2008: 33-79, 204-07;Giumbelli, 2010). Some contemporary Umbandists continue to present their tradition as having extraterrestrial or Atlantian origins (Hale, 2009b: 228). Presenters at the First Brazilian Conference of Spiritism and Umbanda in 1941 pointed far beyond Africa: Umbanda has been ' on Earth for more than one-hundred centuries, with roots lost in the unfathomable past of the  Novo (1937Novo ( -1945. During this period, Afro-Brazilian religions were persecuted but Kardecism, given its upper-class connections, was not; some Umbandists saw an opportunity to escape persecution by downplaying their religion's relations to the former and emphasizing its relations to the latter (Oliveira, 2007;Engler and Isaia 2016: 194-97).
These interpretations beg the question of whether Umbanda is more Africana than esoteric. As we see there is no clear evidence for a purely Africana origin of Umbanda; most accounts of its origins place equal or stronger emphasis on Kardecist roots, and Kardecism is esoteric. Arguably, the admixture of additional esoteric elements in the 1940s and 1950s was as much an elaboration of pre-existing Kardecist roots as it was a denial of African roots. Both factors may well have played a role. We simply don't know. Scholars who insist on the rejection-of-Africa thesis appear to do so on the basis of a presupposition that Umbanda is Africana.
Two scholarly views suggest that Umbanda is African in origin. First, Umbanda is said to have emerged from Macumba. Arthur Ramos, writing in 1934, described Umbanda as a Central African transplant, equating it with 'Macumba' and pointed Engler: Umbanda 14 to the cabula, a Central African ritual for communing with departed ancestors (2001: 97-98, 103, 99). Bastide later echoed the view that Umbanda emerged from Macumba (1995: 447). 'Macumba' refers not to a specific religion but to a range of popular Afro-Brazilian rituals (often labelled 'black magic') that aim at healing and worldly benefits. From the nineteenth century, it was equated with 'low spiritism' in contrast to upper-class Kardecism or 'high spiritism.' Some groups, primarily in Rio de Janeiro, do use the term to describe themselves, but scholars suggest that 'from obscure origins the term "macumba" came to designate that set of spirits, practices, and religious goals classified as illegitimate by a diverse set of actors in the struggle to assert the legitimacy of their own set of spirits, practices, and religious goals' (Hayes, 2007: 287;also see Brown, 1994: 25-36). To say that Umbanda emerged from Macumba points to an Africana origin with no further specification. The evidence is similarity in vocabulary and ritual form, but this could stand as evidence for Umbanda's hybridizing tendency, independent of the question of origin.
A second scholarly account of Umbanda's African origin points to an earlier African tradition in Brazil, the eighteenth-century Calundu-angola (Rohde, 2009;Mott, 1994;Malandrino, 2010: 173, 223-230). This argument is also weak. It is based on suggestive ritual parallels between earlier practices and Umbanda, with no evidence of intervening influence. If we presuppose that Umbanda is Africana, it follows that there must be some historical continuity with earlier African practices transplanted to Brazil. On this view, the calundus seem a likely candidate. But if that assumption is set aside, hard evidence of some historical relation is required, and none has been produced. Historian Laura de Mello e Souza -who made this same case in her earlier work -concludes, 'I no longer believe that the end of the line explains the genesis of the process, i.e., that there is a coherent nexus between … Umbanda and Calundu-angola' (Mello e Souza, 2002; see also 1986: 355).
Both these scholarly accounts of Umbanda's African origin take for granted that the religion is Africana, and then cast about for the best candidate for an origin story among earlier African traditions and rituals. If we bracket the presupposition that Umbanda is Africana, the similarities that ground these arguments could be seen as answering a different question: what Africana elements did Umbanda pick up as it Engler: Umbanda 15 moved from its origins in Kardecism to a Brazilian hybrid esoteric tradition? This is not to assert this alternative view; I simply note that it makes as much sense, based on the evidence.
Most scholars prefer a third account, pointing to the large and growing cities of Brazil's south east and south (São Paulo, Rio de Janeiro and Porto Alegre) as the site of Umbanda's emergence in the 1920s and 1930s (Ortiz, 1999(Ortiz, [1978Brown, 1994: 37-46;Negrão, 1996: 65, 67). On this view, Umbanda originated under conditions of urbanization: 'the birth of the Umbandist religion coincides precisely with the consolidation of an urban-industrial class society'; 'Umbanda is pre-eminently a religion of urbanites' (Ortiz, 1999(Ortiz, [1978Brown, 1994: 132).
Three racialized processes are held to have been central to Umbanda's origin: the blackening (empretecimento) of Kardecism as its members searched for more highenergy, stimulating rituals; the whitening (embranquecimento) of Candomblé due to the expectations of an increasing number of white members, often new immigrants, along with the formation of a 'low Spiritism' among the lower classes; and the reception and rejection of indigenous caboclo spirits in Kardecist groups (Camargo, 1961: 34-35;Ortiz, 1999Ortiz, [1978Brown, 1994: 38-48). On this view, Umbanda has Kardecist, Candomblecist and Indigenous roots, but it does not help us decide which of these roots was more important.
This dominant scholarly account has problems. The details of this mixing of Kardecism, Candomblé and indigenous elements in the 1920s and 1930 remain unclear, given the scant, scattered and partial nature of sources. It is one thing to suggest that a given religion -Kardecism or Candomblé -went through a process of schism or hybridization. It is something else to suggest that two different religions experienced symmetrically opposed pressures -blackening and whitening -at precisely the moment when a third racially inclusive development (caboclos) appeared on the scene, and, moreover, that this all led to the formation of a third religion.
The logic of this account is suspicious for three reasons beyond its inherent improbability. First, there is little or no historical evidence to prove that any particular Kardecists or Candomblecists were actually motivated by the desire for more or less Third, this view that Umbanda originated in the happy convergence of three simultaneous processes of racial inclusion is suspiciously aligned with the ideology of racial democracy. As noted above, Umbanda grew and became more organized during the same decades that scholars and politicians began to celebrate Brazil's 'myth of three races.' Umbanda became a self-conscious poster child for racial and cultural mixing. Strong evidence would be required to support this view, but the explanation is more speculative than solid. Umbanda's decline in numbers since the 1960s and 1970s -the period when Candomblé became a symbol of resistance to racism -may reflect its inability to respond consistently to a more critical politics of race, given its bewildering variations of doctrine, rituals and groups, with competing conceptions of its racially identified spirits pulling in different directions (Silva, 2018). For example, most preto velho spirits are mild-mannered caricatures of elderly house slaves, bent with age and offering calm, sage advice, free of any social critique.
In a very few terreiros the pretos velhos are strong, young, vibrant critics of racial injustice (Hale, 1997). The latter exception underlines the fact that Umbanda has been far less a site of anti-racist critique and activism than Candomblé. Umbanda is less Africana in this sense as well.
To sum up, investigating the origin(s) of Umbanda highlights both Africana and esoteric roots, depending on the account. Furthermore, the search for origins presumes that finding a single, clear source would reveal the religion's essence.
Leaving aside the issue of whether this commits the genealogical fallacy -confusing issues of origin and current state -the historical evidence for Umbanda's roots is simply unclear. Umbanda may or may not have begun as a schismatic movement

White and Esoteric Umbandas
In order to address our guiding question -is Umbanda Africana or esoteric? -it makes sense to move past the issue of origins and focus on Umbanda's doctrinal and ritual landscape. There is a tendency to see 'white' and ' esoteric' forms of Umbanda as minor offshoots of an Afro-Brazilian mainstream: as exceptions that prove the rule. This makes sense only if we assume from the start that Umbanda is an Africana religion. What do things look like if we set aside that assumption and start by asking what forms of Umbanda are actually practiced in Brazil today?
White and esoteric variants are less common in Brazil's large southern cities, hence in academic studies, but they are the only type of Umbanda that is found in smaller cities in many regions of the country. With their lower populations and inherently more conservative religious tendencies -where even historical Protestantism and Kardecism tend to be seen as somewhat outré by the Catholic majority -these small cities are a tricky market for religious services. Types of Umbanda that stray too far from Kardecist or Catholic influences are unlikely to gain a foothold.
It is in this context, in these types of Umbanda, that Christian and esoteric elements come to the fore. Some centros -who would never, of course, use the Afro-Brazilian term terreiroshave nothing recognizably African in their doctrine, rituals or iconography, apart from the preto velho spirits: no drums, no food offerings to the entities, no use of West African terms, no mention of orixás (which incorporate in very few Umbandist groups). Lindsay Hale's study of centros in Rio de Janeiro led him to conclude that, in 'White Umbanda … the orixás, … have not only been divested of their identity as African deities but, as with all things that evoke Africa, pushed to the margins, silenced, washed white' (Hale, 2009a: 125). Hale also found that, in some centros of White Umbanda, caboclos are European-explicitly non-indigenous-spirits (2009a: 94-111). Of course, Hales' claim that that Umbanda was ' divested' of African elements reveals a presupposition that Africana identity is primary (Engler, 2015). Like most scholars, he avoids the question of the centrality of Kardecist elements.
One of the reasons that so many scholars reject esoteric elements as late superficial additions to an Africana substrate is that they know so little about esotericism. With rare exceptions in my experience -over fifteen years of fieldwork and two-and-a-half years as a visiting professor in São Paulo -Brazilians acknowledge that Kardecism has always been a part of Umbanda, but they do not recognize that Kardecism is esoteric. This is clear in a recent debate among Brazilian scholars. Silas Guerriero recognizes that Umbanda 'inherited from [Kardecist] Spiritism a moralizing tendency,' but he also argues that: Age' elements. None of the participants take into account the fact that Kardecism itself is esoteric. This is correlated with these authors' shared presumption that Umbanda is Africana. Second, the two sides use the term 'Esoteric Umbanda' in different ways. Oliveira and Guerriero refer to groups that include elements of non-Kardecist western esoteric beliefs or practices (for example, New Age elements).
Carneiro and Carneiro refer to a particular school of Umbanda, which is represented by a series of publications by two key authors (W.W. Matta e Silva and Francisco Rivas Neto), and by two centros and a teaching institute in São Paulo. All of these scholars agree that Umbanda is in essence Afro-Brazilian and that it sometimes incorporates non-Kardecist esoteric elements. In doing so, they all miss the point: if Kardecism is essential to Umbanda, then Umbanda is essentially esoteric.
To sort this out, the Carneiros' limitation of 'Esoteric Umbanda' to the Matta e Silva/Rivas Neto school is unhelpful to scholarly discussions and should be set aside as internal religious-market positioning. It makes more sense to define 'Esoteric Umbanda' as groups or tendencies that include esoteric elements other than Kardecist ones. That is, 'Esoteric Umbanda' sensu stricto is a small institutionalized group in São Paulo; 'Esoteric Umbanda' sensu lato is a tendency, in and beyond White Umbanda groups, to incorporate other esoteric elements, beyond Kardecism.
This allows us to capture the recent emergence and growth of these groups. So, for example, groups of White Umbanda where I conduct fieldwork practice healing with aromatherapy, crystal therapy acupuncture, realignment of chakras, and so on.
One centro appropriates its basic conception of the 'seven lines' and the entities of Umbanda -with their relation to days of the week, colors, types of therapy, and so on-from the publications of Matta e Silva, but with other esoteric elements including astrology and palm reading. (This group is a case of Esoteric Umbanda both sensu stricto and sensu lato.) In sum, it might be the case that all Umbanda is esoteric (Kardecist), but only a small minority of groups are 'Esoteric' (incorporating New Age healing rituals, for example). To the extent that Umbanda has esoteric elements in either of these senses, it is a Spiritism, in the sense that I propose.

Kardecist Umbanda
A key question is whether Umbanda is more Kardecist or Africana with respect to doctrine and ritual. In asking this, we should look at all current forms of the religion that self-identify as Umbanda. 3 To exclude centros of white and esoteric Umbanda would create a vicious circularity: Umbanda is obviously Africana if only Afro-Brazilian terreiros are accepted as authentic. A better approach is to ask whether Candomblé-like ritual and doctrinal elements are found in all types of Umbanda and Kardecist elements less widely represented. If so, Umbanda is primarily Africana.
If, on the other hand, Umbanda is primarily esoteric, we should find the reverse: universal Kardecist elements but only limited Africana elements.
The latter is the case. Umbanda's theology (or spiritology) is Kardecist across the board. In doctrinal terms, the presence of 'orixás' in some Umbanda discourse is the main marker of Candomblé's influence on Umbanda. In Umbanda, however, the orixás are not the divinities of Candomblé. They are spirits like all others but with two important differences: they are so highly evolved that they have never been incarnate, and they do not incorporate mediums. (As always in the panoply of Umbandas, there are exceptions to both these points.) In White Umbanda, the orixás are seldom if ever mentioned.
The incorporation of spirits in Umbanda does not mark a distinction between this world and a transcendent realm. It underlines the continuity between this world and 'higher levels,' through its emphasis (though less pronounced than in Kardecism) on multiple reincarnations and spiritual evolution. Umbanda's spirits are sufficiently ' evolved' that they no longer incarnate in this world of spiritual expiation and growth. By incorporating in mediums in order to help along their less advanced 3 There is a tendency for members of White Umbanda centros to call their practice 'mesa branca' (a synonym for Kardecism) and for members of more esoteric centros to call themselves 'spiritualist,' but, in interviews, many members of these groups also label their tradition as 'Umbanda branca.' My fieldwork suggests that the tendency to see Umbanda as normatively Afro is responsible in part for this tendency among some less Afro-Brazilian groups to resist the term 'Umbanda.' Véronique Boyer suggests, 'Candomblé and Umbanda form poles, tendencies that organize the religious universe with opposing and irreconcilable currents ' (1996: 18).
Camargo's view is misleading because Umbanda overlaps with Kardecism at one end of a spectrum, in ritual terms, and Candomblé at the other. Camargo's fieldwork was conducted largely in the city of São Paulo, neglecting Umbanda branca which is more prominent in the interior of that state. This led him to see Umbanda as Afro-Brazilian.
Umbanda itself is more spectrum than pole. Boyer is right to point to contrasts between Umbanda and Candomblé, but there is almost no spectrum of ritual forms between them. Rather, Umbanda contains that spectrum: from Candomblé-like to Kardecism-like groups. I have argued elsewhere that the various spectra internal to Umbanda itself are more prominent and revealing, and that they have been ignored, in part because of the tendency to project a single Africana essence on the tradition (Engler 2009a;2009b;2012;2016 roots, which is reflected in the theorization of diaspora in terms of politics and identity (Caroso and Bacelar, 2006;Gordon and Anderson, 1999). Africana or Afro- This parallel -between Kardecism's tendency to hybridize throughout Latin America, forming a range of Spiritisms, and Umbanda's tendency to do the same in Brazil -underlines the similarity between these two forms of Spiritism. Afro-Brazilian traditions manifest some degree of hybridization -frequently in the form of Umbandist influences -but to a lesser extent. Placing Umbanda in the context of Latin American Spiritisms further suggests the value of seeing Umbanda as neither Africana nor esoteric, but as a hybridizing Brazilian Spiritism.

Conclusion
Umbanda is not a unified religion. As Lindsay Hale discovered during his fieldwork in Rio de Janeiro, 'there was no singular Umbanda, there were different Umbandas' -'Umbanda takes many forms; referring to it in the singular masks differences as striking as day and night' (Hale, 2009a: xiv, 159, ix-x). Labeling Umbanda as Africana or Afro-Brazilian ignores the centrality of esoteric (Kardecist) elements in Umbanda's many hybrid forms.
There are two relationships between Umbanda and esotericism: a foundational relation to Kardecism; and an ad hoc appropriation of non-Kardecist esoteric ideas and practices, especially New Age healing rituals. To the extent that Umbanda is

Engler: Umbanda 26
Kardecist, it is esoteric. A review of accounts of Umbanda's origin led to two conclusions: Kardecism is more prominent than Africana traditions if we look at those stories as a group; but the lack of evidence leaves only one thing clear, that Umbanda was a plural, fragmented, hybridizing tradition from its earliest recorded days. A review of doctrine and practices shows that Kardecist ideas and practices are found throughout the entire range of Umbanda's many variants, and that Africana elements are more limited, even though they are found in the majority of groups.
Umbanda is at least as esoteric as it is African. Some readers will see a political downside to the suggestion that Umbanda is not as Africana as it is held to be. This more nuanced account presented here potentially weakens the impact of this example of the centrality of Africanness in Brazilian society and culture. I recognize and honour the Africana character of the majority of Umbandist groups in Brazil. The recognition that Umbanda is more than Afro-Brazilian denies nothing of the vibrant and dynamic legacy of the cultures of African slaves, as preserved and elaborated by Afro-descendant Brazilians and their allies.
Umbanda, however categorized, continues to contribute to a vital stream of social memory and to serve as a site of resistance to Brazil's endemic racism and inequality.
The ideological stance that Umbanda was always and will always be essentially Africana has pragmatic value. But it should be recognized for what it is: a normative strategy within complex struggles of identity and power. Scholars in the study of religion/s treat not just religions but discourses about them as social and cultural phenomena. Essentialisms are left at the door. From this perspective, 'Umbanda' overlaps in important and valuable ways with both categories, 'Africana' and ' esoteric.' There are compensating advantages to the suggestion that Umbanda is a hybridizing Brazilian Spiritism, as opposed to an essentially African tradition.
Defending the essential Africanness of Umbanda is an ideological stance that portrays Umbandists as static receivers of essentialized tradition, or as constrained by Afro-descendent identities in the ongoing re-invention of their tradition. This strategy is self-defeating. It ignores demographic, doctrinal and ritual facts, and it ignores Umbanda's relation to its Latin American context. More importantly, it places ideological limits on recognizing the degree of grassroots religious creativity of Umbandists. Underlining the dynamic, hybridizing nature of Umbanda honours the full extent of this creative agency. It further underlines Umbanda's exceptional role as an influence on a wide variety of Spiritisms, as it hybridizes with Africana, esoteric, Indigenous and popular Catholic traditions. The is an aspect of what I have called Umbanda's 'hybridity of refraction,' its remarkable ability to refract aspects of its social, cultural, historical and regional contexts (Engler, 2009a).
Labelling Umbanda as Africana or Afro-Brazilian delegitimizes the many Brazilians who practice other forms of the tradition. The case of Umbanda helps us make sense of Spiritism as a broad category of religions. This is an area that calls for further work. Almost no scholarship looks at commonalities between Spiritualism, Kardecism, Umbanda and related Africana and esoteric traditions throughout the Americas. Highlighting the prominence of Kardecism opens the door to fuller engagement with Umbanda's hybridizing relation with other Spiritisms. Above all, these issues emphasize the value of paying attention to local contexts and avoiding broad generalizations.