Shared Channels, Three Tasks: Dub and Spoken Broadcast in the Coordination of Work, the Signalling of Hazards and the Keeping of Memory in the Zion Cluster, 2045–2047

1. Introduction

The Zion cluster was founded by workers who declined to return when their contracts ended, and who stayed in orbit to build a community of their own. The cluster is a Rastafarian community, and dub music has been part of its life since the founding. Outside accounts describe Zion mostly as a settlement of tugs, hauling berths and growing rings, with a distinctive sound, and the community's own history is told elsewhere (McKenzie-Hart, 2042). Sound-system practice in orbital communities has been reviewed in general terms (Rahmani-Stone, 2039), but little has been written about what the sound does in work.

Dub is a music of the studio. Its producers strip a recorded rhythm to its bass and drums, add echo and reverb, and cut elements in and out, so that the track becomes an arrangement of presence and absence. In Zion the music is not confined to leisure. It runs on shared broadcast channels that reach every work space, and it is joined there by spoken announcements, greetings, requests and recitations. Work in orbit is coordinated across shifts, hazards are real, and a community that began as an act of refusal has a history to keep. Residents told us repeatedly, in the first weeks of fieldwork, that the channels do all of this together.

This study asks how. We examine three functions of the shared channels: the coordination of work shifts, the signalling of hazards, and the keeping of community memory. We approach them as an ethnography of practice, taking residents' own accounts seriously and treating the broadcast record as evidence of what is done, not as a transcript of what is meant. We write in 2048 and report fieldwork from 2045 to 2047. The community has been generous with its time, and the account that follows is bounded by what its members chose to share. Every practice, channel and informant described is drawn from this fieldwork, and we have changed or omitted details at informants' request.

2. Field Setting and Method

Setting. The cluster comprises several habitats and working modules connected by tethers and transit tubes. Three shared channels carry broadcasts to speakers in every module: the work channel, which runs throughout the working day; the yard channel, used in the repair yards and hauling berths; and the evening channel. Any resident may request time on a channel through the cluster's rota of selectors, the people who choose and play the music and read the announcements. A track in this paper means a recorded piece chosen for broadcast, and a version means an instrumental or stripped-down mix of a track, a form that the community regards as a creative genre in its own right.

Participant observation. We observed work at four sites: two hauling berths, a repair yard and a horticulture ring. We spent 212 hours with crews and shift workers across the three channels' working days, and attended 44 shift changes at which we noted what was broadcast. Observation was conducted openly. Both authors are staff of the Zion Cluster Technical College and are known to many residents, which shaped access and may have shaped behaviour. Our presence may have altered what was played.

Recorded broadcasts. We drew a stratified random sample of 96 hours from the recorded archive of the three channels over 2045 to 2047, stratified by channel and by time of day (Zion Oral History Collection, 2047). A segment is a continuous stretch of music or speech between transitions, and the sample yielded 1,184 segments. One coder assigned every segment a primary function and a lead mode: music-led, speech-led or mixed. A second coder independently coded a random 120 of the segments, and the kappa statistic reported here, κ = 0.90, is the agreement beyond chance between the two coders on the primary function of those 120; the lead mode was not double-coded. Disagreements were resolved by discussion. The five functions were defined as follows. Shift and time coordination covers signal tracks, time calls and announcements that mark a change of shift. Hazard signalling covers spoken hazard notices and the cut-and-hold, which coders identified from the audio as the removal of the vocal and upper rhythm with the bass line sustained. Community memory covers recitations of the founding story, naming of the dead, dedications to past crews and commemorative reminiscence. Social exchange covers greetings, requests, dedications to the living and general music for listening. Technical and handover covers equipment notices, maintenance instructions and the naming of tasks and stations at handover; the announcement of incoming crews is coded here when it names tasks and under shift and time when it marks the change itself. These counts describe the sampled hours and do not estimate rates for the whole archive.

Interviews. We held 27 interviews with 22 elders aged between 61 and 88, five of them interviewed twice. Elders were identified through the community's own networks, and each was told how the recordings would be used and could withdraw at any time. The study was reviewed by the research ethics committee of the Zion Cluster Technical College, and the council of elders agreed to the fieldwork on the terms described here. Because the community is small, we have generalised informants' roles in the text so that quoted informants cannot be identified from their roles. Interviews are deposited, with the consent of each elder and with restrictions that elders set, in the Zion Oral History Collection (2047). Some elders declined to explain matters that they regarded as the community's own, and we have not pursued them. In addition, we held 31 short conversations with crew members and selectors during observation, which are recorded in field notes and are not deposited. Informants are identified here by code only: codes beginning with E denote elders, those beginning with Y or B denote workers met during observation, and those beginning with S denote selectors. Quoted passages are paraphrases that informants reviewed.

3. Findings: Shift and Time

Shift and time coordination was the most common function after social exchange, accounting for 358 of the 1,184 coded segments (30%), and 73% of those segments were music-led (Table 1). The pattern we observed is that a particular track is played to mark a change of shift, and that its length matters. In 41 of the 44 shift changes we attended (93%), a signal track was played before the change, at a median of 10 minutes before it (IQR 7–12). In the other 3, the change was announced in speech alone.

Workers described the track as a clock. A worker at the repair yard (informant Y3) explained that a crew knows from the first bars which track is playing and therefore how long remains before handover, since the selectors always use the same recordings for the purpose and the lengths are known. A crew member at a hauling berth (informant B9) said that this was more reliable than a displayed time, because a worker with greasy hands, a helmet or a task in progress hears the track and does not need to look at anything. In the horticulture ring, shifts were organised around the watering cycle, and the selector chose tracks to match its stages (Chambers, 2043).

Spoken notice complements the track at handover. The incoming crew's names and the berths or stations they take over are read out, and the outgoing crew's last tasks are named. Elders saw this as continuing an older practice from the cluster's early years, when crews handed over at the berth in person and the channel confirmed it for everyone else (informant E12; Grant-Reid, 2040). The signal track and the spoken notice divide the work of coordination between them: the music marks time and the voice names people.

4. Findings: Hazard Signalling

Hazard signalling accounted for 97 segments (8%). It is the smallest of the three functions in count. We found two main forms. Of the 97 segments, 41 (42%) were speech-led notices that named a hazard and a place. The other 56 (58%) were music-led or mixed, and each showed what residents call a cut-and-hold, as the coders identified it from the audio: the selector removes the vocal and the upper rhythm from the track, leaves the bass line sounding and then follows with speech.

The logic of the cut-and-hold was explained to us in practical terms. Bass frequencies carry through the structure of the cluster's hulls and tubes and are felt as well as heard, so a worker in a noisy space or wearing protection receives the signal where speech would not carry (Livingstone-Cole, 2043; Brathwaite, 2041). Informants added a point about attention, which we report as their explanation and have not tested. A change in a familiar track draws notice in a way that a new sound does not, and workers who have listened to the channel all day learn what its normal shape is. A selector (informant S2) described the cut as a way of making the music say that something is wrong.

Residents differ in how much they rely on each form. Crews in the hauling berths, working near machinery and in suits, depended on the cut-and-hold. Staff in the horticulture ring, working in quieter spaces, preferred speech. We saw no case in which the cut-and-hold was used for anything but a hazard, and informants said that selectors who misused it for other purposes would lose their place on the rota. The convention is therefore protected by a social sanction, which is also a form of community knowledge: everyone is expected to know what the cut means.

5. Findings: Community Memory

Community memory accounted for 214 segments (18%), 56% of them speech-led. The content was of four kinds: the recitation of the founding story, the naming of those who have died, the dedication of tracks to past crews, and the broadcast of elders' reminiscences on commemorative days. The founding story, which concerns the workers who refused to return, is told by elders on the work channel on anniversaries. We have not recorded its content in detail, since the elders who tell it regard its wording as theirs to give.

Naming is the practice we are best placed to describe. When a death occurs, the elders arrange a broadcast on which the name is spoken over a version, and in later years the name is read on the anniversary. Informants spoke of this as keeping the person in the work of the cluster. A younger worker (informant Y8) said that she knew the names of crews she had never met because she had heard them every year, and that she knew which berths they had worked. In this sense the channel keeps an informal register of the cluster's dead that is also a register of its work.

Three of the 22 elders (informants E4, E7 and E12) placed all three functions within a single view, in which keeping time, warning of danger and remembering people are all ways of looking after the cluster's members. They said that the broadcasts could not be divided into work and non-work, and one said that the community would not accept a channel that did one without the others. The other elders did not address the point, so we cannot say how widely the view is held. We offer it as these elders' account and not as our analysis.

6. Discussion

The three functions share one medium and depend on different mixtures of music and speech. Time is marked chiefly with music, hazards by a combination of an altered track and speech, and memory chiefly with speech set over a version. The differences are not arbitrary. They follow the properties of each form: a track has a known length and a recognisable shape, bass travels through structure, and a spoken name carries content that a rhythm cannot.

Three elders' account of the three functions as one practice bears on how the findings should be read. A study that treated the channel as a work tool, a safety system and a cultural archive in turn would describe each accurately and might miss that these elders hear them as a single broadcast. Those elders also spoke of what separation would cost: if hazards were signalled on a dedicated alarm channel and memory moved to a recording archive, the community would, in their view, lose the daily habit of listening to the same channel that makes each form legible to all. We could not test this, and other residents may see it differently.

Unity of this kind suggests a dependence. All three functions run through shared channels, so a failure of the channels would plausibly affect work, safety and memory at once. We were told of earlier occasions on which channels failed and crews fell back on hand signals and runners. Residents also reported a gradual shift among younger workers to personal audio during shifts. If the shared channel loses its audience, the cut-and-hold loses its force, and the track that marks time loses its value as a common clock. We did not measure the shift to personal audio, which rests on residents' reports, and it should be a priority for later study.

7. Limitations

We observed four work sites in a cluster with more, and they were chosen with the help of residents who agreed to host us. Other sites may differ, and we would expect practice in the evening channel and in domestic spaces to differ from what we saw at work. The recorded sample of 96 hours is stratified by channel and time of day but covers only part of three years, and its counts describe those hours.

Our access came through institutions and relationships, and so did our silences. Elders chose what to tell us, and the practices they did not describe are absent. The authors are staff of a college that the community knows, and that position may have made informants more willing to speak about work and less willing to speak about matters they regard as the community's alone. We do not claim to have heard everything that matters.

Coding is interpretive. A segment's function was assigned from its content and context, and a segment may serve more than one function. Agreement between coders was high (κ = 0.90) but was measured on 120 segments only. The account of practice rests on informants' explanations, which we could not always check against the recordings, and informants may describe an ideal as well as a practice. Our description of the changes under way, including the spread of personal audio, rests on informants' reports and not on measurement.