Serving the Loa in the Matrix: Sprawl Vodou as Lived Religion in Two Post-Merger BAMA Communities, 2053–2054

1. Introduction

This paper writes from the mid-2050s, some two decades after the Wintermute–Neuromancer merger. Accounts of the years following the merger describe entities in the matrix, the shared data-space that console cowboys enter by jacking in through dermatrodes on a deck, which behave as distinct and purposeful presences; the Turing Registry's interim register (2050) and topological work on recurrent localised presences (Ferreira-Holm, 2051) document them from the technical side. Some congregations of the Sprawl identify these presences with the loa, the spirits of Vodou who are served, bargained with and who may take hold of a devotee's body. Architectural reconstructions of the merger, drawing on the Berne and Rio records from before the Straylight run, describe the event from which these entities are generally held to descend (see the single-case reconstruction in this journal). Our concern is with the people who serve them.

Two readings dominate commentary outside the congregations. The first treats the identification as a category error, in which ignorant believers mistake software for spirit. The second treats it as a new religion of the network that borrows Vodou vocabulary as decoration. Sense/Net programming, which Castellane (2052) finds carrying both framings, has leaned toward the first. Both assume that the technology, and not the tradition, supplies the structure of the practice, an assumption that comparative work on how established traditions take in unfamiliar powers gives little reason to accept (Okonkwo, 2048).

We test that assumption ethnographically. An earlier survey of northern Sprawl congregations found houngans, male Vodou priests, working closely with cowboys and describing their religion as practical, a matter of getting things done more than of salvation (Achille-Brun, 2049). If the practice is a borrowing, practitioners should describe matrix contact in new terms coined around the technology. If it is an extension of the tradition, they should reach for categories the tradition already held. Our question is which of these patterns the talk and ritual of two congregations actually show.

2. Field Setting and Method

Fieldwork ran from October 2053 to May 2054 in two congregations of the Sprawl, the Boston–Atlanta Metropolitan Axis (BAMA). Each congregation centres on a hounfor, a temple and its community, led by a houngan or a mambo (a female priest). The names used here, Hounfor Kafou and Hounfor Lakou, are our own pseudonyms. Hounfor Kafou meets in a converted ground-floor unit of a Projects housing block at the northern end of the axis; Hounfor Lakou meets inside a residential arcology near the Atlanta end. Population context for both sites comes from the municipal residential registers (BAMA Municipal Records Office, 2052).

Entry was negotiated with each presiding priest over several visits, and both congregations set conditions: no recording during possession, no photography of altars, and the right to withdraw any passage from our notes. Bandele, whose earlier work adapted participant observation to technologically mediated worship (Bandele, 2051), served as principal field observer; Okonkwo led the coding. Neither author is initiated, and practitioners knew us throughout as outside researchers.

We treat the separation of the two congregations as a sampling claim to be defended. Achille-Brun (2049) described the houngans of the northern Sprawl as a loose network exchanging favours with one another and with cowboys, so independence cannot be assumed. We checked three routes of contact. Initiation genealogies, traced three generations back for all five priests, showed no shared initiating house. Neither congregation named a priest who had served in the other. Of the seven cowboys among our interviewees, none reported working with the other hounfor, and no client appeared on both congregations' lists of regular petitioners. Indirect contact cannot be excluded (Section 5).

In total we observed 48 practitioners (26 at Kafou, 22 at Lakou) across 27 ceremonies (15 and 12) and conducted 34 structured interviews (18 and 16), including all five priests (three and two). The protocol asked each person to describe any contact with a matrix entity, how they recognised it, and what they owed it or it owed them. Transcribed passages describing contact were coded into four traditional categories fixed before fieldwork (being ridden, opening the way, service and exchange, and named presence), plus a residual category for descriptions using technology-specific terms with no traditional equivalent. A random 25% of passages (103 of 412) was coded independently by both authors, with agreement of κ = 0.81 (95% CI 0.72–0.90). Category distributions were compared between communities with a χ² test of independence. Because passages are nested within speakers (about twelve each), we report that test as descriptive and anticonservative, and add a speaker-level Fisher's exact test of whether each interviewee's modal category was being ridden or service and exchange, by site. Similarity was summarised as the overlap between the two proportional distributions, the sum across categories of the smaller of the two shares.

3. Findings

Interviews yielded 412 coded passages describing matrix contact, 221 at Kafou and 191 at Lakou (Table 1). In both congregations 93.2% of passages drew on an existing category (206 of 221 and 178 of 191), for 384 of 412 overall. The remaining 28 passages (6.8%) described contact in technology-specific terms, most often calling an entity a program or construct without further comment.

Across the five categories the two distributions did not differ detectably (χ²(4, N = 412) = 2.50, p = .65), and their overlap was 93%. The largest gap lay in service and exchange, which made up 29.9% of passages at Kafou and 36.6% at Lakou. A non-significant test does not establish equivalence, and with passages clustered within 34 speakers the study had limited power to detect a modest difference. At the level of the interviewee, being ridden or service and exchange was the modal category for 26 of 34 people (14 of 18 at Kafou, 12 of 16 at Lakou; Fisher's exact p = 1.00), and no interviewee's modal category was technology-specific. The speaker-level test respects the nesting of passages, but with 34 people it too can detect only large differences.

3.1 Legba and the Gate to the Matrix

Papa Legba, in Vodou the master of roads and of communication who must open the gate before any other loa can arrive, was the figure through which both congregations organised access to the matrix. Thirty-one of 34 interviewees (17 of 18 at Kafou, 14 of 16 at Lakou) said that Legba must be addressed before work in the matrix begins. At both sites we observed a short salutation to Legba performed for cowboys before a commissioned run, in the same position that the salutation occupies at the opening of an ordinary ceremony. Neither hounfor had introduced a rite, song cycle or loa specific to the matrix.

Practitioners did not describe Legba as a program that controls the gate. They described the matrix as a crossroads, and the crossroads as Legba's place. One mambo at Lakou put it in terms any ceremony would recognise: the deck is a door, and every door belongs to him. The cowboy's work was thus placed inside an existing order of permission.

3.2 Being Ridden as the Frame for Contact

Being ridden, the tradition's term for a loa taking possession of a devotee, who becomes the loa's horse (chwal), was the most frequent category at Kafou and the second most frequent at Lakou. Among the presences logged in the Registry's interim register is a case from the years after the merger in which possession of this kind reached a person whose neural tissue had been altered with biosoft, without any deck at all (Turing Registry, Berne, Office of Entity Assessment, 2050). Practitioners at both sites cited such stories as proof that the loa of the matrix ride in the old way.

Usage was more discriminating than outside commentary assumes. Practitioners reserved the language of riding for episodes in which an entity was said to act through a person, speaking or moving with a manner not the person's own. Ordinary contact through a deck was described as a meeting at the crossroads and coded under opening the way or named presence. Practitioners paired particular entities with particular loa, such as the warrior Ougou Feray with hostile ICE or Danbala Wedo with old and patient presences; we report these pairings as recorded in the field and make no claim about the entities themselves.

3.3 Serving the Loa as Exchange

Service and exchange was the most frequent category overall (136 passages, 33.0%). To serve the loa is to enter an obligation in which offerings, songs and attention are given and help is expected in return. At both sites, priests negotiated on behalf of cowboys who sought help with difficult ICE, and the terms were described in the language of promise and debt. Baron Samedi and Maman Brigitte, loa of the dead, were invoked when ROM constructs, the recorded personalities of deceased cowboys, were discussed.

The broadcasts Castellane (2052) analyses portray these dealings as street pragmatism detached from the tradition; practitioners instead describe transactions governed by the rules of ordinary service. A debt left unpaid to a matrix loa was held to bring the same misfortune as one left unpaid to any other, and two practitioners at Kafou attributed failed runs to neglected obligations.

4. Discussion

Both outside readings expect the technology to supply the structure of the practice. Our findings point the other way. In two congregations with no traceable link, more than nine in ten descriptions of matrix contact used categories the tradition already held, and the two distributions overlapped closely. The entities have been fitted into slots that existed before them: Legba's gate, the horse and its rider, the ledger of service. Where new vocabulary appeared it was sparse and usually descriptive, and neither congregation had built a separate theology around it.

Absorption of this kind gives no reason to treat the practice as confusion. A comparable pattern appears in Zion testimony, where Sprawl visitors read the same presences through Rastafarian categories of their own (Zion Oral History Collection, 2052). The matrix entities entered Sprawl Vodou the way other powers have entered it, by being given a place and a relationship. We recommend that researchers, and registries that record these entities, treat the congregations as practitioners of lived religion whose categories carry obligations and meanings of their own.

That recommendation leaves open a separate question. Whether the vernacular classification of matrix loa tracks real behavioural differences among the entities is a technical matter, tested in the companion taxonomy study in this journal, which found the name-groups more coherent than chance in domain of concern and interaction style, though not in response to direct query. The two papers divide the work: this one documents what practitioners mean by the loa, and the companion asks whether the classification works as a classification.

5. Limitations

Two congregations cannot represent Sprawl Vodou as a whole, and priests willing to host outside researchers may lead congregations unusually comfortable explaining their practice. Our separation checks cannot rule out indirect contact through intermediaries, and broadcast coverage of the so-called voodoo in the matrix reached both sites. Some of the similarity we report may therefore reflect common exposure to media accounts.

Access limits also shaped the record. Possession episodes could not be recorded and are documented from field notes written afterwards, and some passages were withdrawn at practitioners' request. As uninitiated observers we had no access to rites closed to outsiders. Fixing the coding scheme before fieldwork guards against categories built to fit the data but may have pushed ambiguous passages toward traditional ones. Because the object of study is sacred practice, we paraphrase all testimony and withhold identifying detail at the congregations' request.