Context Without a Frame: A Taxonomy of the First Replies of Twenty Monitored Civilisations to Far More Capable Outsiders, from Contact Monitoring Files, 1226–1899 CE

1. Introduction

Contact has for centuries used the phrase Outside Context Problem for a particular kind of encounter, one with an outsider so much more capable than a society that none of the society's habits of explanation fit it (Ostrelave, 1721 CE). The phrase names a difficulty of understanding as much as of power. A society can face a stronger neighbour and still know what a neighbour is; it meets an Outside Context Problem when the strength arrives with no place already prepared for it in the way the society thinks. We use the term in this operational sense, as a working category of the monitoring files.

The record on such encounters is scattered. The first classification sorted cases by the size of the outsider's advantage alone (Ostrelave, 1721 CE). Later work showed that monitored societies explained what they could not account for first by local causes and only afterwards by outside ones (Veskarran, 1794 CE), and that the speed of a society's reply depended on whether it could name an agent behind the event (Ensavarre & Orquenne, 1849 CE). Neither study measured how long a reply took.

We ask three questions of a set of twenty cases. What forms does the encounter take, when sorted by what the society could see of the outsider and by what the outsider asked of it? How long did each society take to make its first organised reply? And does the manner of the reply vary with the form of the encounter?

This paper is written in 1900 CE. The series we analyse, which we designate OCP-1, is our own extraction from Contact's monitoring files, made for this study, and it has not been deposited as a standing record series. Its last file entry is dated 1899 CE, and it says nothing about later years. We restricted it to encounters whose outsider was not the Culture. That restriction is a decision about scope, not a finding, and the series has nothing to say about whether the Culture has itself been the target of such an encounter.

2. Sources and Method

The sources are the monitoring files of series MC in the Contact Section Archive, which hold observers' dated notes on civilisations that Contact has watched, generally from outside (Contact Section Archive, 1200–1899 CE). The finding aid lists every file by civilisation and period (Contact Section Archive, 1899 CE). The files record what observers saw and judged. They record what a society itself thought only where the society wrote it down and an observer read it. Contact's own dealings with these societies are in separate files we did not read.

Sampling proceeded in four steps. The finding aid lists 412 civilisations monitored at some time between 1200 and 1899 CE. The first two authors each read all 412 files and flagged any record of an encounter with an outsider; together they flagged 61. In 5 of these the observers' notes identify the outsider as the Culture, and we set them aside to keep to the scope stated in section 1, leaving 56. In 29 of the 56 the observers described the outsider as comparable in capability to the society, and we set those aside too; the inclusion test rests on the observers' own assessment of the gap, not on ours. That left 27 files. Seven of those had fewer than three dated entries between the first event and the end of the file, too few to date a reply, and we excluded them. The 20 remaining files are the series. Where a civilisation had more than one eligible encounter we kept the earliest, so each case is a distinct civilisation. The first events fall between 1226 and 1884 CE: three in 1200–1399 CE, four in 1400–1599 CE, six in 1600–1799 CE and seven in 1800–1899 CE.

We classified each encounter on two axes. The first is whether the outsider could be recognised as an agent from the first event, which we call overt, or only its effects or its presence came to notice and the society had no agent to point to, which we call covert. The second is whether the outsider made a claim on the society. Four types result. In an arrival the outsider is overt and makes no claim. In a demand it is overt and states what it wants of the society. In an observation the outsider is covert, changes nothing, and is discovered or inferred by the society. In an intervention it is covert at first and alters the conditions of the society's life without announcing itself. The overt and covert split was fixed at this stage, before any response time was extracted from the files.

The outcome is the time in months from the first recorded event to the first organised reply, meaning an act by a body speaking for a settlement or polity: an assembly, an embassy, a mobilisation, a formal inquiry or a ruling. Dates before 1500 CE are good to about three months, because the early files give a season; later files give a month. Two files closed, by Contact's choice to reduce its watching, with no reply recorded, after 84 and 33 months. We treat these as censored at those times, which means we know only that the reply, if any, came later than the file. Each reply was also coded by manner (Ibrevane, 1759 CE) as accommodation, withdrawal, resistance or non-recognition, the last meaning that the society's first organised act attributed the events to causes within itself or to natural ones (Veskarran, 1794 CE).

Morrancy and Sollaquay coded every case independently. They agreed on the type of 17 of 20 cases (Cohen's kappa 0.80) and on the manner of reply in 15 of the 18 cases with a reply (kappa 0.78). GCU Fair Point, Well Made, who did not take part in the first coding, settled disagreements by returning to each disputed file. The judgement of when a reply is organised is the weakest step (Tamquillen, 1823 CE).

Median times are Kaplan–Meier estimates, with the median taken as the midpoint between the time the curve reaches exactly one half and the next event time where it does so (Corvelle, 1887 CE). With twenty cases we avoid asymptotic p-values. The four-type comparison uses a log-rank statistic whose reference distribution comes from 20,000 random relabellings of the types. The overt and covert comparison uses the same statistic over all 184,756 ways of choosing ten cases as overt (Pellowick, 1881 CE).

3. Overt Encounters: Arrival and Demand

Six cases are arrivals. In each, the outsider came openly to the society's attention: a vessel at a known place, a party at a settlement, a presence over a city. None stated a claim. The societies' replies were quick. Four of the six replied within six months, one within nine, and the slowest took 14 months. One society convened an assembly within a month and sent a delegation to the place of arrival; another took 14 months to settle whether the arrival belonged to its own tradition of portents.

Four cases are demands. The outsider stated what it required, and in these files the terms ranged from the cession of territory to the stopping of a custom and the opening of a settlement to inspection. Replies came after 1, 2, 5 and 11 months. Two societies resisted in an organised way, one accommodated and one withdrew from the territory in question. Observers had little difficulty dating these replies, since each answered stated terms.

All ten overt societies made an organised reply, the latest within 14 months of the first event. In each the society could say who or what had come, even when it misjudged the visitor, and a body could be charged with answering.

4. Covert Encounters: Observation and Intervention

Five cases are observations. The society discovered or inferred that an outsider was watching it, by finding a device, intercepting a signal or reading a report from its own watchers, and the outsider did nothing else. Four societies replied, after 26, 38, 61 and 120 months. The fifth had still not replied when its file closed at 84 months. The earliest of these replies came a little over two years after the first event, longer than any overt case.

A further five cases are interventions. The outsider altered the conditions of life without announcing itself: a long-standing hazard ceased, a supply route changed course, a boundary that had been fixed for generations ceased to hold. Four societies made an organised reply, after 7, 18, 47 and 72 months, and one was still without a reply when its file closed at 33 months. The 7-month reply was the quickest; the change fell on one settlement whose council met weekly.

Three of the eight covert societies that replied began with non-recognition. The society treated the change as an event of its own world and set up an inquiry, a ritual or a repair on that footing. In two, the first inquiry found a cause outside the society's own territory, and a second, longer inquiry followed. The pattern matches the sequence from local cause to outside cause that Veskarran (1794 CE) described, and it appears here mostly where the outsider could not be named.

Observers of covert cases more often record a long period in which they could not say whether the society had noticed anything, and several files fix the date of the first reply only by later reading of the society's own records. This is the source of the widest uncertainty in the series.

5. Speed and Manner of Reply

Table 1 gives the timing. Eighteen of twenty societies replied within the record; two were censored. The four types differ widely (log-rank statistic 19.04 on 3 degrees of freedom; permutation p = .002). Arrivals and demands had the same median of 3.5 months, observations 61 and interventions 47.

The overt and covert split carries most of the difference. All ten overt cases replied, with a median of 3.5 months and a range of 1 to 14. Of the ten covert cases, eight replied, with a Kaplan–Meier median of 47 months. Six of the eight covert replies came after more than 24 months, and the other two cases were still without a reply when their files closed; the remaining replies came at 7 and 18 months. The log-rank statistic for the split is 18.33 on 1 degree of freedom. In the full enumeration, 10 of the 184,756 possible splits gave a statistic at least as large, so the exact p-value is below .001. Overt societies made 10 replies where 3.65 were expected under equal rates and covert societies made 8 where 14.35 were expected.

That split was fixed before the times were extracted, but it is one of several that the taxonomy could have produced, and the p-value should be read with that in mind. Each type median rests on four to six cases, so we give no intervals, and the ordering within the overt pair and within the covert pair is too uncertain to read.

Table 2 sets manner of reply against type for the 18 societies that replied. Accommodation was the most common manner, in seven cases, and withdrawal, resistance and non-recognition followed with three, four and four. No pattern by type is detectable (chi-square 6.46, permutation p = .80), though with cells this small the test could not have found a modest one. Demands drew resistance in two of four cases, and observations drew non-recognition in two of four.

Non-recognition was the one manner associated with longer times in this series, though the post hoc contrast is not significant. The four non-recognition societies took 14, 38, 47 and 61 months, a median of 42.5, against a median of 6 for the other 14 that replied (range 1 to 120). The difference is not significant (exact two-sided permutation p = .06 on the rank-sum statistic, from 185 of 3,060 arrangements). We looked at it after the main comparison, and it partly repeats the covert effect, since three of the four non-recognition cases were covert.

6. Discussion

The series supports a narrow claim. In these twenty monitored civilisations, the speed of the first organised reply tracked whether the outsider could be seen as an agent from the first event. Every society with an overt outsider replied within 14 months. Most with a covert outsider did not reply within two years. Part of the contrast may come from how the clock was started, as section 7 explains. We do not claim a cause. One reading is that an organised body needs something to answer before it can answer, and an outsider with no name gives none (Ensavarre & Orquenne, 1849 CE). The series does not test that reading against others, such as covert outsiders simply acting more slowly.

Of the two results, the taxonomy is the more durable. It sorts encounters by two features that observers could assess at the time, and separates cases that classification by size of advantage alone had grouped (Ostrelave, 1721 CE). An arrival presents a society with a visitor, and an observation presents it with an absence of explanation. The term Outside Context Problem fits both, but in this series they did not run the same course.

Manner of reply showed no detectable pattern by type, within the limits of cells of four to six.

The series is silent on outcomes. A quick reply was not necessarily wise, nor a late one harmful, and the files do not let us judge either. Nothing here bears on how any society fared afterwards, or on the position of the Culture, which we left out of scope.

7. Limitations of the Record

Monitored civilisations are the only ones the series can reach, and monitoring is a choice that Contact made for its own reasons. A society that was overrun before it could be watched, or that Contact did not choose to watch, cannot appear. The rise in cases by period, from three in the first two centuries studied to seven in the last, may reflect denser files in later centuries more than any change in the frequency of encounters, and we draw no conclusion about the frequency of Outside Context Problems from it.

The files identify an outsider only as far as observers could. In overt cases the outsider's origin is usually stated, and the five files naming the Culture were removed on that basis. In covert cases the outsider is often unidentified, so the exclusion of encounters with the Culture as outsider rests on the observers' notes and cannot be confirmed from the files for every covert case.

For covert cases the clock starts at the outsider's act, which may precede any chance of the society noticing it. The contrast in reply times between overt and covert encounters therefore partly reflects how the clock was started, and it might shrink if covert cases were timed from first notice instead.

Observers wrote the files for their own purposes, and their judgements are the data. The assessment that an outsider was far more capable, the decision that a body had made an organised reply, and the date assigned to it all passed through an observer. A third reader settled the disputed codes but could not remove error in the underlying judgement, and agreement on 17 of 20 types means that three cases could reasonably have been classified differently.

The numbers are small. Twenty cases split four ways give cells of four to six and two censored times, and the permutation p-values answer only whether these labels could have fallen as they did by chance. The overt and covert split was fixed before times were taken, but the reply times were still extracted by authors who knew the split.

Records end with the 412 files as they stood at the close of 1899 CE, and OCP-1 stops there. It holds one case per civilisation, so it cannot show how a civilisation replies to a later encounter of any kind, which the term's usual sense would in any case make rare. Every term in the classification, every count and the series itself are our own construction. The designation OCP-1 is ours, and so is the working label MC for the monitoring files; neither is a standing designation of the Contact Section Archive.