Uncited Press › Star Trek Open the interactive journal →
Trek · Political Economy & Social Structures

What the Society Became: Findings of Prime Directive Violation and the Rank of the Officer in 72 Judge Advocate General Inquiries, 2344–2375

Dr. Mariam Haddadin1
1 Starfleet Judge Advocate General's Office, San Francisco
Received 7 Sep 2026 · Revised 13 Oct 2026 · Accepted 24 Oct 2026

Abstract

General Order 1, the Prime Directive, forbids interference in the development of pre-warp societies, and the Judge Advocate General's Office decides whether an officer has broken it. This study asks which of two things better predicts a finding of violation: what became of the society, or the rank of the officer. I coded 72 inquiries opened between 2344 and 2375, drawn from the 104 in the Office's register, for the officer's act, for whether the society was altered in a lasting way, for the officer's rank grade and for the finding. The Office found a violation in 39 inquiries (54.2%). Findings were made in 30 of 36 inquiries in which the society was altered (83.3%) and in 9 of 36 in which it was not (25.0%), an odds ratio of 15.0 (95% CI 4.72 to 47.7; p < .001). Findings were made against 20 of 34 senior officers (58.8%) and 19 of 38 junior officers (50.0%), an odds ratio of 1.43 (95% CI 0.56 to 3.63; p = .45). In a model that held both predictors, rank added nothing detectable (odds ratio 2.16, 95% CI 0.65 to 7.19; p = .21). The hypothesis that rank predicts the finding is not supported. The record of each society's fate is thin and can be shaped by the finding itself, so the association describes the register and does not show how the Office reasons.

1. Introduction

In Starfleet usage a directive can be a command to act, a standing limit on action or a statement of purpose, and General Order 1 has at different times been treated as each of these. The order, which was in force in Starfleet by the 2260s at the latest and which cadets meet in their first term (Corrin, 2372), forbids interference in the natural development of a society that has not yet achieved warp flight. The Judge Advocate General's Office has held inquiries into alleged violations since it opened a register for them in 2344 (Starfleet Judge Advocate General's Office, 2376).

Earlier reviewers have worked on the register, each with a different aim. The first review covered the inquiries of its opening seventeen years and described the language of the findings, but it did not record what became of the societies concerned (Ostrowicz, 2361). A later study separated the three senses of the Directive and showed that panels use them unevenly (Valdemar-Nkemelu, 2367). A third examined the censure that follows a finding (Penhallow-Ades, 2369). None of them asked what a finding depends on.

Familiar episodes show the kinds of act at issue in the order. The exposure of a concealed observation post, and then of the ship and its captain, to the Mintakans in 2366, the handling of a society on the edge of warp flight, as with the Malcorians in 2367, the correspondence with a child on a pre-warp world in 2365 and the removal of a small population to another planet in 2370 all fall within it. I use them only as touchstones for the categories below. They are not cases in the register and play no part in the counts.

Two features of an inquiry might explain its finding. Is it what became of the society after the officer acted? Or is it who the officer was, and in particular how senior? The second possibility has a respectable basis, since a senior officer can be held to a higher standard, or may be shielded by standing. I took as the claim under test that rank predicts the finding, and I set it against the competing claim that the consequence for the society does.

The review is written from the vantage of 2376, and every inquiry and source it uses was closed before the end of 2375.

2. Methods

The source is the Office's register of inquiries into alleged violations of General Order 1, together with the case files it indexes and the post-contact status reports that the Contact Records Office keeps for the societies concerned (Starfleet Contact Records Office, 2376). The register lists 104 inquiries opened between 2344 and 2375. Seventeen files are sealed and were not available to me. In a further 15 inquiries no status report on the society had been filed by the end of 2375, so that its fate could not be recorded in the data. That left 72 inquiries, each with a finding, an officer and a recorded outcome for the society. How was each case coded? Four variables were taken from each file in turn. The finding was binary, a violation found or not found. The act was one of four types: exposure of the officer's presence, provision of knowledge or technology, medical or rescue intervention, and removal or relocation of persons. The society was coded as altered if the status report recorded a lasting change in its institutions, beliefs, technology or population that the report attributed to the act, and as not altered otherwise. Rank was coded as senior for commander and above and as junior for lieutenant commander and below, using the rank held on the date of the act.

What counted as a finding? Only a ruling in the closing statement that the officer had violated General Order 1, whether or not a sanction followed. What counted as a sanction? A formal reprimand, relief from command or post, or both. Who decided the codes? I coded every file, and a second reviewer from the Office coded all 72 independently. The two of us agreed on 69 of 72 for the society, on 66 of 72 for the act and on 72 of 72 for rank and the finding, and disagreements were settled from the text of the file.

The finding is binary, so the analysis uses tests for proportions and logistic regression. For each predictor alone I report the proportion with a score interval, the difference in proportions with a normal interval, the odds ratio with a Wald interval and a Pearson chi-squared test without continuity correction. For the sanction, where cells are small, I used a two-sided exact test for a two-by-two table. I then fitted logistic models for the finding with the society alone, with rank alone and with both, and compared them by the likelihood ratio, by Akaike's criterion and by the area under the receiver operating characteristic curve (Lysander-Obi, 2375). With 39 findings and at most two predictors the models are not overfitted. No other variable was added, because the act types are too sparse to enter without inflating the intervals.

All tests are two-sided, and the register is treated as a complete census of the inquiries that met the inclusion rule and not as a sample, so the intervals describe the uncertainty of the coding and of the process that produced the files and nothing wider.

3. Results

Across the register the Office found a violation in 39 of the 72 inquiries (54.2%) and found none in 33. Table 1 gives the findings by the type of act. Findings were most frequent after the removal or relocation of persons (10 of 14, 71.4%) and after the provision of knowledge or technology (12 of 18, 66.7%), and least frequent after exposure of presence (9 of 24, 37.5%). The difference among the four types is not significant (chi-squared 5.61 on 3 degrees of freedom, p = .13).

In the register the society was altered in 36 inquiries and not altered in 36. The even split is how the codes fell; no quota or matching was used. A violation was found in 30 of the 36 in which it was altered (83.3%; 95% CI 68.1 to 92.1) and in 9 of the 36 in which it was not (25.0%; 95% CI 13.8 to 41.1). The difference is 58.3 percentage points (95% CI 39.7 to 77.0). The odds ratio is 15.0 (95% CI 4.72 to 47.7; chi-squared 24.7 on 1 degree of freedom, p < .001).

By rank, the officer was senior in 34 inquiries and junior in 38. In all, a violation was found against 20 of the 34 senior officers (58.8%; 95% CI 42.2 to 73.6) and against 19 of the 38 junior officers (50.0%; 95% CI 34.8 to 65.2). The difference is 8.8 percentage points (95% CI −14.1 to 31.8), and the odds ratio is 1.43 (95% CI 0.56 to 3.63; chi-squared 0.56 on 1 degree of freedom, p = .45). Rank does not predict the finding in these data.

Cross-tabulating the two predictors (Table 2), an altered society was recorded in 16 of the 34 senior inquiries (47.1%) and in 20 of the 38 junior ones (52.6%), so rank and the fate of the society are close to unrelated in the register, and rank is unlikely to confound the result. Within each rank the pattern is the same as in the whole. Among senior officers a violation was found in 14 of 16 inquiries with an altered society and 6 of 18 without one. Among junior officers the figures were 16 of 20 and 3 of 18. The rank estimate rises when the society is held (Table 3) because an odds ratio grows when a strong predictor is added and because altered societies were a little more common in junior inquiries. It does not reach significance either way.

Models for the finding come next (Table 3). The society alone separates the findings well (area under the curve 0.79; likelihood-ratio chi-squared 26.4 on 1 degree of freedom, p < .001). Rank alone separates them hardly better than chance (area 0.54; likelihood-ratio chi-squared 0.56, p = .45). When both enter, the society keeps an odds ratio of 17.0 (95% CI 5.05 to 57.5; p < .001) and rank has an odds ratio of 2.16 (95% CI 0.65 to 7.19; p = .21). Adding rank to the society model lowers the deviance by 1.64 on one degree of freedom (p = .20) and does not lower the Akaike criterion (76.9 for the society alone, 77.3 for both). Adding the society to the rank model lowers it by 27.5 (p < .001).

What happened to the officer after a finding? Of the 39 findings, 22 ended in a formal reprimand, 12 in relief from command or post and 5 in no further action. Relief followed 7 of the 20 findings against senior officers (35.0%) and 5 of the 19 against junior officers (26.3%); the exact test gives p = .73. The sanction does not depend on rank in this register either.

4. Discussion

Rank does not predict the finding, and the claim is not supported. Senior and junior officers were found in violation at similar rates.

The odds ratio for rank has an interval that spans one, and rank adds nothing to a model that already holds the fate of the society. The same holds for the sanction after a finding. These are the results, and I draw nothing further from the positive estimates for senior rank, 1.43 alone and 2.16 with the society held, since both intervals include one.

A different matter is the fate of the society. Where the status report recorded a lasting change, the Office found a violation in five inquiries of six, and where it did not, in one of four. Why should findings track consequences so closely? The Directive is worded as a bar on interference, and an inquiry in principle finds a violation whether or not the interference left a mark. The register shows that in practice the Office usually did not.

That pattern is what a panel would produce if it treated the Directive as a principle about outcomes. The data do not show that the Office reasons in this way. They record findings and statuses, and they do not record the reasons that panels gave to themselves. The act types do not settle it either: findings were more frequent after removal and after the provision of knowledge than after exposure alone, but the differences among the types are not significant, and I did not tabulate the act against the fate of the society.

The finding bears on advice to officers only with care. If the status reports were written apart from the panels, which Section 5 doubts, counsel asked what the Office is likely to decide would do better to ask what became of the society than how senior the officer is. That rests on one register and describes past practice.

5. Limitations

Is the fate of the society recorded independently of the finding? Not always. Status reports are filed by the Contact Records Office, but several were written or revised after the inquiry closed, and a report can describe a society as altered because the panel had already found that the officer acted wrongly. If so, part of the association is built in by the record. I could not separate the two, and the odds ratio of 15.0 should be read with that in mind.

The 72 inquiries are 69.2% of the 104 in the register. The 17 sealed files can differ from the rest in ways I cannot know. The 15 inquiries without a status report are more likely to concern societies left undisturbed, since such a society is less likely to have a report filed at all, so the set probably holds too few of the not-altered kind. That changes the share of altered societies but need not change the odds ratio, unless filing also depended on the finding. The dichotomy of rank hides the grades within each group. The ns are modest. The interval for the rank odds ratio runs from 0.56 to 3.63 alone and from 0.65 to 7.19 with the society held, so a large effect of rank cannot be excluded, though none was found. Several officers appear in more than one inquiry, and the intervals treat the inquiries as independent.

This review extends the register work begun by Ostrowicz (2361) and adds the variable which that work lacked. Three priorities follow. First, the Office should require a status report on every society a fixed period after an inquiry opens, written by someone who has not sat on the panel. Second, each finding should be published with the status report the panel had before it, in one indexed series (Federation Council Legal Research Office, 2374), so that a later reviewer can tell which came first. Third, the register should be coded again each year with the same four variables. The sealed files should be coded first.

Prime DirectiveGeneral Order 1Judge Advocate General inquiriesStarfleet military lawnon-interferencelogistic regressionpre-warp societies

References

  1. Starfleet Judge Advocate General's Office (2376). Register of inquiries into alleged violations of General Order 1, 2344–2375. Starfleet Incident Records, JAG inquiry series GO1, files 1–104.
  2. Starfleet Contact Records Office (2376). Post-contact status reports on pre-warp societies subject to inquiry, 2344–2375. Starfleet Contact Records, SCR-75 status series.
  3. Ostrowicz, T. (2361). A first review of inquiries under General Order 1, 2344–2360. Journal of Interstellar Law, 23(2), 77–109.
  4. Valdemar-Nkemelu, S. (2367). Command, limit and purpose, the senses of the Prime Directive in Starfleet inquiries. Journal of Interstellar Law, 29(1), 12–47.
  5. Penhallow-Ades, R. (2369). Censure and relief from post after a finding under General Order 1. Journal of Interstellar Law, 31(3), 150–183.
  6. Corrin, H. (2372). Instruction in General Order 1 in the first term at the Academy. Starfleet Academy Board of Instruction Reports, report 201.
  7. Federation Council Legal Research Office (2374). Publication and indexing of Starfleet findings on non-interference. Federation Council Legal Research Office Reports, LRO-74-09.
  8. Lysander-Obi, E. (2375). Logistic models for binary findings in small archival series of legal files. Proceedings of Applied Speculative Statistics, 23(1), 18–39.
Read this article inside the full journal experience — browse by faculty, search across universes, and filter by discipline.
Open in Uncited Press →