Paper Two of the Series “The Ledger and the Covenant”
Abstract
Eva Stratt saves the human race and ends the novel in a courtroom. Most readings of Project Hail Mary treat this as an ethical puzzle to be settled — a trolley problem with billions on one track and one unwilling man on the other — and divide accordingly between those who think she was justified and those who think the book lets her off too easily. This paper argues that the ethical framing, while not wrong, is the less interesting of the two questions the novel raises, and that it obscures the more revealing one. Stratt’s method and the method Grace and Rocky use at Tau Ceti are not merely morally different; they are operationally different, and the novel quietly demonstrates that the second is capable of things the first is not. The problem Stratt could not have solved was solved by faculties her regime systematically suppressed: consent, transparency about ignorance, cheap and reversible error, and time spent on a relationship with no immediate payoff. Her imprisonment is then examined not as a verdict on her deserts but as a structural outcome that the literature on national deliverers would have predicted regardless of desert. The paper concludes that the anti-hero problem is better stated as a design problem — how a polity obtains Stratt’s speed without Stratt’s impunity — and offers four provisions toward an answer, together with a candid statement of the residue that no provision reaches.
I. The Puzzle
Stratt is right about nearly everything. She is right about the timeline, right that the astrophage problem admits of no gradual response, right that ordinary procurement and ordinary deliberation cannot produce a starship in a few years, right that Grace’s knowledge is needed aboard the Hail Mary, and right that no one else will carry the weight if she sets it down. She is also, by the novel’s own accounting, honest about what she is doing. She does not claim virtue, does not construct an ideology, does not develop a taste for her own authority, and does not build a following. She has no personal life in the text and no evident appetite beyond the mission’s success.
And she ends up before a tribunal.
The common reading resolves this as an ethics question. One widely circulated treatment states it as the ultimate version of the trolley problem, with the “one” being a man taken against his will and the “many” being the entire species, and concludes that Weir deliberately declines to supply an answer. A theologically inflected treatment sorts the novel’s characters into consequentialist, deontological, and virtue-ethical positions and locates Stratt firmly in the first. A sharper minority reading faults the novel for a lack of political interest altogether, reading her regime through James Scott’s account of high-modernist schemes that sweep aside local knowledge and traditional interests, and objecting that the book grants her the promises of centralized authority without depicting its characteristic failures. A widely shared fan observation notes what may be the most important textual datum of all: Grace never forgives her, and the novel does not require him to.
Each of these is a reading of Stratt against a moral standard. What none of them does is compare her method to the other method the novel puts on the page. That comparison is available, the novel structures itself around it, and it yields a result that the ethics framing cannot reach: that the coercive method has a capability ceiling, and that the book’s climax takes place above it.
II. Two Methods, Itemized
A. Stratt’s method
1. Unilateral authority, explicitly outside law. Her grant of power is not an office within a legal order but a suspension of legal orders. It comes with no term, no appeal, no review body, no succession plan, and no defined scope. Whatever else may be said for it, it is designed to be unreviewable, and the design is the point.
2. Speed as the supreme value. Deliberation is treated as waste. Consultation is treated as delay. The premise — that the timeline permits no argument — is true, and it converts every procedural safeguard into an obstacle by definition.
3. Persons as resources. Scientists are pulled from their careers and their countries. Grace is conscripted, drugged, and loaded aboard a ship after an explicit refusal. Others are removed from the project when they cease to be useful. The consistency here matters: this is not cruelty but a settled way of reasoning about people, the same way of reasoning that produced the coma decision analyzed in Paper One.
4. Information control. Access to what is known is allocated by the mission’s needs rather than by any claim the informed might have to know.
5. Irreversibility. The nuclear detonations in the Antarctic ice and the conversion of the Sahara are irreversible acts undertaken on a compressed timeline against models that could not have been adequately tested. Whatever their necessity, they possess a common feature: if the reasoning behind them had been wrong, nothing could have been undone.
6. Accountability deferred. She proceeds as though the question of judgment belongs to a future that either will not arrive or will not be hers to face. When it arrives, she does not appear surprised.
B. The Tau Ceti method
1. Consent at every step, with real refusals exercised. Neither Grace nor Rocky can compel the other, and both decline things. The refusals are not obstacles to the work; they are the mechanism by which the work is kept honest, because a party who can refuse is a party whose agreement carries information.
2. Transparency about ignorance. Both are able to say that they do not know. This phrase is close to unavailable inside Stratt’s regime, where the leader’s certainty is the organizing resource and admitted uncertainty is a solvent. Grace and Rocky’s willingness to state ignorance is what makes their joint model of astrophage improve rather than harden.
3. Iteration designed so that being wrong is survivable. The Taumoeba program is the novel’s clearest demonstration. They breed, test, select for ammonia resistance, discover the failure mode, and correct it — and they discover it because the design allowed for discovery. Small trials, reversible commitments, results checked before scaling. Compare the Antarctic detonations, which admitted no trial stage at all.
4. Redundancy through a partner rather than through control. Stratt achieves reliability by removing everyone’s discretion but her own. Grace and Rocky achieve it by each holding independent competence and each checking the other. These produce very different failure profiles: the first fails all at once at a single point, the second fails locally and gets caught.
5. Error-correction because two are watching. The Taumoeba resistance failure is caught because two independently trained minds with different backgrounds are looking at the same organism. Neither would have caught it alone. This is not a moral claim; it is an epistemic one, and it is the mechanism on which both worlds’ survival turns.
6. Time given to the relationship before the crisis. The hours spent building language, playing, and learning each other’s habits have no immediate mission payoff and are the reason the pair can act instantly when payoff is required. Stratt’s regime cannot make this investment, because from inside it every hour without output is waste.
III. Paired Test Cases
The novel supplies four matched pairs, and each rewards direct comparison.
1. Recruitment. Stratt drugs Grace and loads him aboard after he has explicitly refused, having established that no persuasion will work. Set beside this Rocky’s decision to continue after his own crew is dead — a voluntary continuation, made alone, with no authority over him and no one to know if he stopped. The two ships arrive at Tau Ceti carrying, respectively, a man who was taken and a being who came. Both are indispensable. Only one of them is reliable in the sense that matters, and the novel’s plot depends on Grace becoming the second kind before the end.
2. Handling a lethal unknown. The Earth-side handling of astrophage proceeds under conditions of forced speed, compartmentalized information, and irreversible large-scale intervention. The Tau Ceti handling of Taumoeba proceeds by controlled breeding, deliberate selection, staged testing, and the assumption that the first answer is probably wrong. The second procedure catches a fatal flaw; the first procedure has no stage at which such a flaw could have been caught.
3. Handling a partner’s failure. Within Stratt’s regime, failure is grounds for removal, and removal is the mechanism by which quality is maintained. Between Grace and Rocky, failure is absorbed and corrected, and neither party’s standing is at risk. The consequence is that Grace and Rocky report their errors immediately and Stratt’s subordinates have every reason not to. A regime in which admitting error is dangerous is a regime that will learn of its errors late.
4. The expenditure of a life. This is the hinge of the paper. Stratt spends Grace’s life. Grace, at the climax, spends Grace’s life — turning the Hail Mary around to save Rocky, forgoing home and near-certain survival. The arithmetic is identical: one life against a large good, weighed and expended. The difference is proprietary. Stratt disposes of what is not hers; Grace lays down what is. Every moral tradition worth consulting marks this distinction, and the novel marks it structurally by making the second act the one that redeems its perpetrator and the first the one that lands its perpetrator in court.
IV. Why Anti-Hero Rather Than Villain or Hero
The category matters because it determines what the novel is arguing.
She is not a villain. She has no self-interest in the outcome, no appetite for domination, no ideology, and no personal beneficiaries. One perceptive discussion of the film notes precisely this: what distinguishes her from the ordinary authoritarian is that her end is not her own power. She also bears a real weight that no one volunteers to share, and the burden is not performed for anyone’s benefit.
She is not a hero either, and the reason is not squeamishness about her methods. It is that she makes herself the sole exception to every rule she enforces. She requisitions others’ autonomy and retains her own. She subjects everyone to a calculus and exempts herself from being an input to it. She treats consent as a peacetime luxury while continuing to act as though her own judgment requires no one’s ratification. A person who spends others under emergency conditions and never once spends herself has not made a sacrifice; she has made an allocation.
The anti-hero category fits because the novel needs her, uses her, credits her, and refuses her. And the refusal is rendered structurally rather than rhetorically. Weir does not argue against her. He simply arranges the plot so that the decisive problem — first contact, the construction of a shared language, the joint biology that produces Taumoeba, the recognition that the first Taumoeba strain would fail — is solved by exactly the faculties her method suppresses. Coercion cannot produce Rocky. There is no requisition order that reaches a being outside human jurisdiction, no leverage to apply, no threat available, nothing to confiscate. At the point of contact, every instrument in Stratt’s toolkit is inoperative, and the man she conscripted has to fall back on the schoolteacher’s instruments: patience, explanation, demonstrated reliability, and gift.
This is the paper’s central claim, and it is an operational claim rather than a moral one. The coercive method has a ceiling. It can build a ship in four years. It cannot make a friend, and the mission required one.
V. The Imprisonment
The novel’s disposition of Stratt is usually read as its verdict on her. I want to propose that this reading, though natural, is probably mistaken, and that the imprisonment is better understood as a structural event with a well-documented history.
Reading A: ingratitude. Once safety is restored, scruples become affordable again, and the polity punishes the person who did what it needed done and could not authorize. On this reading the trial is a scapegoating ritual by which a community discharges its own complicity onto the agent it empowered.
Reading B: restoration. The emergency was real, and the suspension of law was still lawless. That she was necessary does not make her lawful, and a legal order that could not eventually reach her would not be a legal order. On this reading the trial is the system reasserting the proposition that no grant of authority is unreviewable, which is precisely what her grant purported to be.
Reading C: the structural reading, which this paper prefers. Both of the above are true, and the outcome would have followed in large part regardless of which was true.
The recurring pattern is documented across a wide historical range: the figure invested with extraordinary authority during an existential threat, whose success generates a form of standing that cannot be converted into any ordinary office, and who is subsequently destroyed by legal means. The essential features are consistent. First, non-convertibility: emergency authority has no peacetime equivalent, and there is nowhere to put the person afterward. Second, prestige as rival legitimacy: a figure credited with the survival of the community is a standing alternative to every legitimate authority she overrode, whether or not she has any wish to be. Third, asymmetry of the reckoning: she acted against many jurisdictions, each of which retains a claim, while she retains no protector. Fourth, and diagnostically, legal form: the destruction proceeds by trial rather than by assassination or coup, because legal form is what converts an unbearable rival into a criminal without the community having to admit what it is doing.
Stratt satisfies every condition. The trial is therefore weak evidence about her deserts and strong evidence about her position. Her successors, if humanity is ever in this situation again, will be destroyed too, whether they behave better or worse. This matters practically: if the reckoning is structural, then designing for it in advance is possible in a way that arguing about desert is not — which is the burden of Section VII.
One further textual point deserves emphasis. Grace never forgives her, and the novel neither requires him to nor treats his refusal as a defect. This is the correct outcome on the account offered here. Forgiveness is owed by the person wronged, and it is not the community’s to award on his behalf in exchange for what it received. A polity that needs its conscripts to forgive is a polity trying to purchase moral closure with someone else’s property — which is, once more, Stratt’s own characteristic move.
VI. Scriptural and Moral Frame
Romans 3:8 states the principle at issue with unusual bluntness: the rejection of “let us do evil, that good may come,” together with the observation that those who reported the apostles as teaching it were slandering them. The verse is not a counsel of prudence, and it does not contain an exception for sufficiently large goods. Its force in the present context is that it denies the very move Stratt’s defenders must make — the move by which the magnitude of the threatened harm converts an impermissible act into a permissible one.
Authority as office rather than possession. The scriptural treatment of rulers consistently frames authority as delegated and accountable: the ruler is a minister, holding power for a purpose and answerable for it, not owning it. Stratt’s grant inverts this. It confers power without any of the accountability structures that make power an office rather than a possession, and it does so deliberately, on the theory that accountability structures are what cost time. The theological objection is not that she used the authority badly. It is that authority of that shape should not exist, because it is authority no one holds under anything.
Saul at Gilgal and at Amalek (1 Samuel 13; 15). The archetype is exact and worth developing at length. Saul faces a genuine emergency: the Philistines are massing, his people are scattering, and Samuel has not come at the appointed time. His response is to perform the sacrifice himself, and his defense is the defense of every emergency actor — the situation was real, the delay was intolerable, the people were leaving, and he forced himself to do what had to be done. The narrative does not dispute the facts of his account. It disputes the inference. Urgency did not license what was not his to do. At Amalek the pattern repeats with a partial obedience justified by an apparently better use of the spared property, and the answer given is that obedience is better than sacrifice — that is, that the substance of the outcome does not retroactively authorize the manner of its achievement. Saul is the biblical Stratt in every particular except that his emergency was smaller and his motives less pure, and the difference in scale is precisely what the passages deny is relevant.
Uzzah and the ark (2 Samuel 6:6–7). The steadying hand, extended to prevent a real disaster, by a man with no standing to touch what he touched. The episode is difficult and is often read as disproportionate, but its structure is the one under discussion: the impulse to prevent a harm does not by itself confer the authority to act, and the sincerity of the impulse does not alter the question.
Whose life may be spent. The distinction identified in Section III has its clearest expression in John 15:13, where the greatest love is the laying down of one’s own life for one’s friends. The construction is reflexive and the emphasis is not incidental. Set against it stands the older pattern in which national security is purchased by passing other people’s children through the fire — a practice the prophets treat not as a permissible emergency measure but as the defining abomination. The distance between these two is the distance between Grace’s climactic decision and Stratt’s founding one, and it is the reason the novel can redeem the first while leaving the second in court.
Grace’s redemption, correctly stated. He does not become brave. Several reviewers have noticed that meeting Rocky does not transform him into a hero and that he remains the man who did not want to go. That is right, and it is the more interesting account. What changes is not his courage but his ownership of the ledger. The cowardice was never chiefly a deficiency of nerve; it was the assumption that his life was his to preserve and other lives were the appropriate currency for saving the world — which is Stratt’s premise with the sign reversed, and which is why he and Stratt understand each other so well and like each other so little. At the climax he spends what is his. That is not bravery arriving late. It is the correction of an accounting error.
VII. The Anti-Hero Problem Restated as a Design Problem
If Section V is right that the reckoning is structural, then the productive question is not whether Stratt was justified but how a polity might obtain her speed without her impunity. Four provisions follow from the analysis, and none is adequate alone.
1. Pre-committed sunset with a defined terminal date and a named successor regime. The chief defect of the grant is that it is open at both ends. An authority that expires on a date certain, with the return of jurisdiction specified in advance, is capable of most of what Stratt did while remaining a thing held under something. The cost is real: an agent who knows she will be reviewed will decline some acts that ought to be performed. The alternative cost is a person who expects no review ever.
2. Contemporaneous recording for later audit, sealed during the emergency. The regime’s information control and its irreversible acts share a common problem: no contemporaneous record exists of what was known when. A sealed-record requirement imposes almost no delay, does not compromise operational secrecy, and converts the eventual reckoning from a contest of memories into a review of documents. It also has a preventive effect on the actor, which is both its purpose and its cost.
3. A separate custodian of consent, with no operational role and no authority to stop anything. The distinctive gap in Stratt’s regime is that no one anywhere in it is charged with representing the interests of the people being spent. Such an officer should not have a veto — a veto would simply reproduce the delay the emergency cannot afford — but should have standing to record objections, mandatory access, and a guaranteed place in the eventual review. The function is documentary and it is the cheapest of the four.
4. Judicial rather than automatic indemnity. The novel’s regime offers immunity in advance; the world subsequently withdraws it. Both moves are bad. Indemnity determined afterward by a court, against a standard fixed beforehand, gives the agent something better than a promise that will not be kept and gives the polity something better than a scapegoat.
The residue. Honesty requires stating what these do not reach. Some acts will only ever be performed by a person who expects not to be judged, and every provision above reduces, at the margin, the population of people willing to serve and the range of acts they will undertake. The design problem admits of improvement but not of solution. The most that can be claimed is that the improvement is worth its cost, and that a polity which has thought about the question in advance will produce fewer Stratts and treat the ones it produces less shabbily.
VIII. Conclusion
The ethical reading of Eva Stratt is not wrong, but it is a smaller question than the one the novel actually poses. Stated as ethics, the matter ends in an unresolvable exchange between those who weigh outcomes and those who weigh acts, and Weir declines to arbitrate. Stated as method, it resolves. The coercive procedure builds the ship and cannot solve the problem the ship encounters. The consensual procedure could not have built the ship and is the only thing that works once it arrives. Both were needed, which is the uncomfortable finding, and the novel arranges its plot so that the second gets the last word.
Paper One argued that the partnership between Grace and Rocky is covenantal — mutual self-binding without enforcement — and that the coma decision revealed a mission philosophy that treated persons as payload. This paper has argued that the same philosophy governed the whole enterprise, that it has a capability ceiling the novel deliberately locates, and that its chief practitioner was undone by a mechanism that had nothing to do with her deserts. Paper Three will take up the remaining question: what could possibly ground a covenant between species that share no world and no ancestry, and whether the novel’s stated answer — a common biological seed carried between star systems — can carry the moral weight the story assigns it.
Notes
1. The exact disposition of Stratt at the novel’s end requires verification before publication, and the paper’s language should be adjusted to match. At least one commentator describes a trial scene occurring in the book prior to the mission’s conclusion rather than a straightforward imprisonment afterward; my Section V speaks of “the trial” and “the tribunal” in terms that will need tightening once the text is checked. If she is tried but not convicted, Reading C survives intact — the structural claim concerns the reckoning, not the sentence — but Reading A weakens considerably and should be restated.
2. The Sahara conversion and the Antarctic detonations need precise citation, including what each was intended to accomplish and on what modeling basis. My use of them in Section II.A.5 rests on their irreversibility rather than their wisdom, and the argument should not overreach into claiming they were ill-judged, which the text does not support.
3. Section IV asserts that no instrument in Stratt’s toolkit is operative at the point of first contact. This should be tested against the possibility that Grace’s training under her regime supplied capabilities he could not otherwise have had — the astrophage expertise above all. The honest version of the claim is that her regime supplied the technical preparation and could not supply the relational competence, and that the relational competence came from the classroom she was pulled out of. Paper Seven takes this up directly; a sentence acknowledging it belongs here.
4. On the structural account of the deliverer’s fate in Section V: this is compressed from my larger series on persecuted national military deliverers, and the four conditions are stated there at length with the historical cases. For a standalone publication of this paper, either the conditions need supporting cases in the text or the section needs an explicit cross-reference to the series. The compressed version as it stands asserts a pattern without demonstrating it, which is the section’s chief vulnerability.
5. The Saul material in Section VI is the strongest scriptural parallel in the suite and is currently underdeveloped relative to its importance. Consider expanding it into its own section, or into a short companion paper, working through 1 Samuel 13:8–14 and 15:13–23 in full. The exegetical point that carries the weight is that the narrative concedes the facts of Saul’s emergency and rejects the inference anyway.
6. I have deliberately not litigated whether the utilitarian calculus is correct on its own terms. The trolley framing is well represented in the existing commentary and there is little to add to it. If a reviewer objects that the paper dodges the central ethical question, the answer is that the paper’s claim is that the ethical question is not central — but that answer should be stated in the introduction rather than implied.
7. The observation that Grace and Stratt share a premise with the sign reversed (Section VI, final subsection) arrived late in drafting and may be the most useful thing in the paper. It should probably be moved forward into Section IV and given room, since it explains something the reception literature notices without accounting for: why their scenes together have the particular friction they do.
8. The film’s softening of Stratt is noted in the reception literature and is not treated here. It belongs to the adaptation paper, if that paper is written. Do not build any claim in this paper on the film’s portrayal; the argument is about the novel.
9. APA formatting note: the online commentary cited below is uneven in quality and is used here as evidence of the shape of the reception rather than as scholarly authority. The distinction should be made explicit in a prefatory note if the paper is submitted anywhere that would otherwise treat these as sources of record.
References
Arendt, H. (1963). Eichmann in Jerusalem: A report on the banality of evil. Viking Press.
Buber, M. (1970). I and thou (W. Kaufmann, Trans.). Charles Scribner’s Sons. (Original work published 1923)
Greer, J. (2026, June 12). The ethical discourse in Project Hail Mary. Christian Letters. https://jacksongreer.substack.com/p/the-ethical-discourse-in-project
Kant, I. (1998). Groundwork of the metaphysics of morals (M. Gregor, Ed. & Trans.). Cambridge University Press. (Original work published 1785)
Eva Stratt character analysis in Project Hail Mary. (2025). LitCharts. https://www.litcharts.com/lit/project-hail-mary/characters/eva-stratt
Lord, P., & Miller, C. (Directors). (2026). Project Hail Mary [Film]. Amazon MGM Studios.
Machiavelli, N. (1996). Discourses on Livy (H. C. Mansfield & N. Tarcov, Trans.). University of Chicago Press. (Original work published 1531)
Ostrom, E. (1990). Governing the commons: The evolution of institutions for collective action. Cambridge University Press.
Rossiter, C. L. (1948). Constitutional dictatorship: Crisis government in the modern democracies. Princeton University Press.
Scott, J. C. (1998). Seeing like a state: How certain schemes to improve the human condition have failed. Yale University Press.
The illiberalism of Project Hail Mary. (2025, April 27). Tragedy and Farce. https://tragedyandfarce.blog/2025/04/27/the-illiberalism-of-project-hail-mary/
TruStory FM. (2026, April 22). Project Hail Mary: Ethics in a crisis [Audio podcast episode]. Superhero Ethics. https://trustory.fm/superheroethics/project-hail-mary-ethics-in-a-crisis/
Walzer, M. (1973). Political action: The problem of dirty hands. Philosophy & Public Affairs, 2(2), 160–180.
Weekly, R. (2025, December 31). Project Hail Mary. https://rebeccaweekly.com/2025/12/31/project-hail-mary/
Weir, A. (2021). Project Hail Mary. Ballantine Books.
The Holy Bible: King James Version. (n.d.).
