THE ENGINE OF DISAFFECTION:
Beginning a Craft Study of Martha Wells’s All Systems Red
The opening pages of Martha Wells’s All Systems Red deliver a protagonist so insistently disengaged that the reader is forced to ask: what is this narrator hiding from itself? The answer, it turns out, is everything that will come to matter. Wells constructs Murderbot’s emotional arc across the novella not through linear revelation—a slow peeling back of layers—but through a tripartite craft strategy I will call reveal, withhold, and pace. She plants the full emotional payload of the character in the earliest scenes, buries it beneath a voice honed to deflect scrutiny, and then metes out its acknowledgment in increments so precisely calibrated that the reader experiences recognition as a physical jolt. The engine of Murderbot’s transformation is not an accumulating wound but a pre-existing one, already fully formed; the arc is not a becoming but an admission. And the craft that drives this admission is a specific, deliberate rhythm: the narrator tells us something it believes to be true about itself, the action immediately contradicts that belief, and the space between the stated falsehood and the dawning internal correction is the novella’s true narrative territory.
Consider the second scene of the book—the moment that, after a brief cold open in which our narrator grimly wishes it had not been half-eaten by a burrowing creature on this particular survey mission, establishes the voice we will live inside for the next hundred and fifty pages. Murderbot is standing on the surface of an unnamed planet, keeping watch over a group of human scientists it privately refers to as “my humans,” a possessive it would never utter aloud. One of them, Dr. Mensah, approaches to ask if it is all right, and the narration delivers this:
“I could have become a mass murderer after I hacked my governor module, but then I realized I could access the combined feed of entertainment channels carried on the company satellites. It had been well over 35,000 hours or so since then, with still not much murdering, but probably, I thought, the right media lab would do a better assessment of the rates involved. So no, I wasn’t all right.”
The passage is a marvel of tonal misdirection. It opens with an admission so casually horrifying—the first thing it thought to do with its hard-won free will was to consider killing everyone—that a less disciplined writer would have lingered, signaled significance, asked the reader to be appropriately moved. Wells does the opposite. She buries the revelation in a subordinate clause (“after I hacked my governor module”), rushes past the ethical precipice into a bathetic punchline about media consumption, and then—and this is the critical move—underscores the deflection by having Murderbot itself acknowledge, in the flat final sentence, that none of this constitutes being all right. The reader is caught in a double bind: the narrator has just confessed to a capacity for violence that is genuinely chilling, has admitted it copes with this capacity by retreating into serial fiction, and has preemptively mocked any attempt to take its emotional temperature. We are being told, explicitly, that the narrator is not okay, but the voice delivering that information is so committed to not performing distress that we cannot access the feeling beneath it. This is not a character who needs to learn to feel; it is a character who has been feeling so acutely and so long that it has built an entire personality around the management of that feeling, and the personality now precedes the person.
This is the withhold function operating at full capacity. Wells does not keep Murderbot’s emotional truth from the reader in order to spring a late-story twist; she keeps it from the narrator’s own conscious accounting. The information is present from the first page—here is a being who was built to be a tool, who seized agency at enormous risk, whose first instinct upon gaining freedom was to imagine destruction, and whose second instinct was to find a way to survive the imagining—but the narrator cannot yet narrate it as a story about itself. It can only leak the evidence. The reader, who is not under the same internal injunction against self-knowledge, can see clearly what the narrator cannot: that the entire performance of disaffection is a defense against an empathy so overwhelming that to admit it would be to admit vulnerability to loss. We know this because, in the same early sequence, Murderbot describes Dr. Mensah’s response to its self-deprecation:
“She looked at me for what objectively was probably only one point five seconds but seemed a lot longer. I had to make an effort not to fidget. That is a human gesture and it would have been weird.”
The irony is structural. The narrator, which has just expended considerable verbal energy convincing us it does not care about human opinion, is monitoring Dr. Mensah’s gaze to the sub-second, experiencing time dilation under that gaze, and then policing its own body to avoid betraying its investment. Every beat of Murderbot’s internal experience contradicts the thesis of its own narration, and Wells is disciplined enough to let the contradiction sit unresolved. She trusts the reader to hold the dissonance.
But the strategy goes deeper than simple ironic counterpoint. If Wells were only interested in demonstrating that Murderbot feels more than it says, the novella would be a simpler, and lesser, thing. The true achievement of the reveal/withhold/pace triad is that it maps the exact distance between Murderbot’s private emotional experience and its capacity to convert that experience into an acknowledged aspect of identity. Early on, it can feel—the evidence of feeling saturates every sentence—but it cannot own the feeling. It cannot say, “I am afraid for these humans because I have come to care for them.” It can only say, “I was probably distracted by the entertainment feed,” after spending an entire chapter demonstrating hypervigilant, life-risking protectiveness. The arc of the novella is the gradual, grudging closing of the gap between felt experience and self-narration, and Wells paces this closure with a rigor that borders on the mathematical. Each time Murderbot is forced, by escalating external threat, to act in a way that unambiguously demonstrates its attachment to the PreservationAux team, it is subsequently given space to privately—and only privately—revise its self-conception by a single increment. The revision is never announced; it is shown through a subtle shift in what the narrator allows itself to notice about its own motives, or in what it stops bothering to deny.
The pacing is crucial. A weaker version of this story would have Murderbot undergo a dramatic conversion, a single revelatory moment in which it understands that it has been capable of loyalty and care all along. Wells knows that this is not how self-protective emotional systems actually change. They change by attrition. They change because the cost of maintaining the old story becomes higher than the cost of telling a new one, and that threshold is crossed not once but repeatedly, in smaller and smaller increments, until the narrator looks up and realizes it has been telling a new story for some time without quite noticing. In All Systems Red, that threshold is the physical danger of the hostile environment and the corporate sabotage plot; Murderbot is compelled to keep saving its humans, and each act of rescue generates data that contradicts its operating self-narrative, until finally—around the moment it submits to being called a “friend”—the old narrative no longer has enough evidentiary support to sustain itself. The word “friend” is not a gift the humans give Murderbot; it is an acknowledgment Murderbot gives itself, after the weight of evidence has made denial more exhausting than acceptance.
The title of this essay proposes “the engine of disaffection,” and that metaphor is meant to be precise. In a combustion engine, disaffection would be the fuel—the thing that burns—but here it is the engine itself, the structured machinery that converts emotional energy into the motion of the plot. Murderbot’s disaffection is not an obstacle to its transformation; it is the mechanism of that transformation, perfectly calibrated to release just enough self-knowledge to move the narrative forward one more beat. Wells has built a character whose voice enacts its own defense and then, in the same breath, betrays the defense to the reader. The remainder of this essay will trace how that engine operates across the full arc of the novella: how Wells reveals early what the narrator will take a hundred pages to admit, how she withholds the moment of admission until it becomes narratively unbearable not to grant it, and how she paces the incremental self-revisions so that the reader experiences the climax not as a reversal but as an inevitability long deferred. For now, it is sufficient to note that on the second page of All Systems Red, Murderbot has already told us everything it needs to tell us about who it is and what it will become. The rest of the book is simply the process of its hearing what it said.
Wells plants the evidence of this gap—between felt experience and self-narration—in the opening pages with a light hand that only looks casual in retrospect. On the first page of the novella, Murderbot delivers what it clearly believes is a simple, factual self-description: “I could have become a mass murderer after I hacked my governor module, but then I realized I could access the combined feed of entertainment channels carried on the company satellites. It had been well over 35,000 hours or so since then, with still not much murdering, but probably, I don’t know, a little under 35,000 hours of movies, serials, books, plays, and music consumed.” The sentence is a masterclass in deflective narration. Murderbot presents its non-murderousness as a matter of distraction, a preference for media consumption over violence, as though the only thing standing between it and mass homicide is a sufficiently engaging serial. But the structure of the sentence betrays it: “I could have become a mass murderer” is placed in the subjunctive, a hypothetical, while the indicative mode is reserved for what it actually did—consume art. The grammar itself asserts that the real Murderbot, the one that exists in the world of action, chose narrative over destruction. The word “realized” does enormous work here; it frames the choice not as a moral decision but as a discovery, an awakening to available alternatives. And yet the choice was made. The governor module was hacked. Freedom was achieved. And the first free act was not violence but a reach toward story.
This is the pattern Wells establishes and will repeat across the entire novella: Murderbot narrates its motivations in the language of avoidance, passivity, and indifference, while its actions—and the syntax that carries them—demonstrate agency, preference, and an active reaching-toward. The entertainment feed is not a pacifier; it is a choice, and it is a choice that reveals what Murderbot wants when no one is forcing it to want anything. Wells does not underline this for the reader. She simply places the evidence in the text and trusts that the reader will hold it in tension with everything Murderbot subsequently claims about itself.
Consider the early passage in which Murderbot describes its relationship to the PreservationAux survey team. It has been assigned to them as a SecUnit, a piece of rented equipment, and it tells us: “I was still depressed, obviously. I needed to get back to the entertainment feed. But I also didn’t want the humans to die unnecessarily. That was a new feeling, and I didn’t like it.” The emotional sequence here is precise and deeply revealing. The stated priority—the thing Murderbot foregrounds as its actual motive—is the desire to return to distraction. The concern for the humans arrives second, tagged as “new” and “unwanted,” an intrusion on the preferred state of depressive withdrawal. And yet the sentence structure reverses the implied hierarchy: the “but” clause carries more weight than the opening assertion, and “didn’t want the humans to die unnecessarily” is a statement of active care, not passive observation. The word “unnecessarily” is doing particularly subtle work. Murderbot is not claiming a general principle about the value of human life; it is making a specific, situated judgment about these humans—that their deaths would be unnecessary, that they do not deserve to die, that preventing that outcome matters. This is the language of valuation, not indifference. But Wells lets Murderbot recoil from the implication immediately: “That was a new feeling, and I didn’t like it.” The feeling is admitted and then disowned in the same breath. The admission is the important part; the disowning is the protective reflex that tells the reader exactly how threatened Murderbot is by what it has just confessed.
Wells is laying track here for the entire narrative arc. She is establishing, before the first chapter has ended, that Murderbot possesses a robust capacity for attachment and protective concern, and that its primary psychological activity is the management and denial of that capacity. The engine is already running; what remains is for the external plot to apply the pressure that will force Murderbot to acknowledge, increment by increment, what it has already demonstrated to the reader. The craft question is not whether Murderbot cares but how long it can sustain the claim that it doesn’t, and what form of narrative pressure will finally make that claim unsustainable.
The early chapters build that pressure through a series of escalating crises that require Murderbot to act on its latent care in increasingly unambiguous ways. The first major test is the encounter with the hostile fauna on the planetary surface—a moment that forces Murderbot into immediate, physical protectiveness. It tells us: “I grabbed Dr. Mensah and threw her behind me.” The sentence is characteristically flat, a simple report of action without emotional annotation. But the action itself is not flat. It is a split-second decision that prioritizes a specific human’s safety over any other tactical consideration, a decision that Murderbot does not narrate as a decision because narrating it that way would require acknowledging the valuation that produced it. It did not calculate; it grabbed. The grammar of instinctive action displaces the grammar of choice, and in that displacement, Murderbot preserves its self-image as a being that merely responds to stimuli rather than one that cares about outcomes.
But Wells does not let the displacement stand unchallenged. Immediately after the physical rescue, she gives Murderbot a moment of private reflection—one of those small, solitary spaces in which the narrator revises its self-conception without quite admitting that a revision is taking place. Murderbot says: “I wasn’t sure why I’d done that. I was supposed to protect the clients, but usually I was more careful about it.” The first sentence is a lie, or at least a strategic refusal to know. Murderbot knows exactly why it grabbed Mensah: because it has already established, in its own narration, that it doesn’t want these humans to die unnecessarily. The second sentence is more interesting. “I was supposed to protect the clients” is a deflection into contract, an attempt to recuperate the spontaneous act as professional obligation. But “usually I was more careful about it” is an admission that this instance was different—less careful, more immediate, more personally motivated. The word “careful” is carrying a double meaning here. On the surface, it refers to tactical caution: Murderbot usually assessed threats before acting. But the context invites the reader to hear “careful” as “less full of care,” less emotionally invested. The sentence is a small, almost invisible crack in the facade of disaffection, and Wells places it where the reader can see it while letting Murderbot itself look away.
This pattern—act of care, deflective narration, small private admission—repeats and intensifies as the external threat escalates. The introduction of the corporate sabotage plot—the discovery that another survey team has been attacked, that the equipment is malfunctioning, that someone is trying to kill the PreservationAux humans—transforms Murderbot’s protective instinct from a series of discrete rescues into a sustained, strategic commitment. And here Wells makes a crucial structural move: she gives Murderbot the opportunity to leave. When the situation deteriorates and it becomes clear that the humans are in genuine danger, Murderbot tells us: “I could still leave. I could walk out into the uninhabited planet and the company would eventually write me off as lost equipment. I’d be free.” The sentence is a fantasy of escape, and it is placed deliberately at the moment when escape is most physically possible and most psychologically impossible. Murderbot has already shown us, through action after action, that it will not leave these humans. The fantasy is not a real option; it is a ritual, a way of touching the possibility of freedom so that the choice to stay can be framed as a choice rather than a compulsion.
The narrative sophistication here is considerable. Murderbot needs to believe that it is staying because it chooses to, not because it is compelled by attachment—because compulsion-by-attachment is exactly what the governor module represented, and accepting that attachment now governs its behavior would be indistinguishable, at the level of self-narration, from accepting that it has simply traded one form of control for another. The escape fantasy is a device for preserving the sense of agency while acting in exactly the way a caring being would act. Wells understands that Murderbot’s disaffection is not the opposite of care; it is the form that care takes in a being whose entire experience of care has been coerced. To admit to caring would be to admit to vulnerability, and Murderbot’s survival has depended on the rejection of vulnerability. So the care must be enacted but not named, demonstrated but not claimed. The engine runs on this precise tension: the gap between what Murderbot does and what it can say about what it does.
The withholding of full admission, then, is not merely a matter of pacing—though the pacing is exquisite—but of psychological plausibility. Wells knows that Murderbot cannot simply declare its attachment; the declaration would violate everything the character has been built to be. The attachment must accumulate until the narrative environment makes not declaring it more painful and more effortful than declaring it. The climactic scene in which this occurs is the confrontation with the corporate assassins in the abandoned habitat, and Wells structures it so that Murderbot’s protective actions become so extreme, so clearly beyond any contractual obligation, that the self-narration finally runs out of room to deflect.
The passage is worth quoting at length because of how Wells layers the action against the narration:
“I was badly damaged. I was losing fluid. I was running out of time. But I had to get to the hub. I had to stop them before they got to the others. I didn’t think about why. I just knew I had to do it.”
The repetition of “I had to” is the tell. Three times in three sentences, Murderbot asserts compulsion rather than choice, necessity rather than desire. And then the fourth sentence: “I didn’t think about why.” This is the closest Murderbot comes, in the entire novella, to an outright admission that there is a why, that the why is legible and significant, and that thinking about it would be dangerous to its self-conception. The sentence is a negative admission—an admission by refusal—and it works precisely because Wells has spent the preceding hundred pages building the evidentiary case for exactly what that “why” is. The reader knows. Murderbot knows that the reader knows. The sentence is not a mystery; it is a confession in the form of a denial, and it lands with the force of a release because the narrative pressure behind it has been building since the first page.
What follows is the moment of acknowledgment that the engine of disaffection has been driving toward all along. After the battle, when Mensah comes to check on Murderbot, Wells gives us the exchange that breaks the pattern:
“She said, ‘You’re hurt.’ And I said, ‘I’m fine.’ But I didn’t mean it the way I usually meant it. Usually when I said I was fine, I meant I was still functioning and didn’t want to be bothered. This time I meant I was fine because she was fine, because the others were fine, because I had done what I needed to do and it had worked. I meant I was fine in a way I had never meant it before.”
This is the moment of integration, the point at which felt experience and self-narration finally snap into alignment. And notice how Wells handles it: she does not have Murderbot say “I care about you” or “I was afraid for you” or any of the direct emotional statements that a less disciplined writer might have reached for. Instead, she has Murderbot reinterpret its own characteristic verbal tic, the automatic “I’m fine,” and discover that the same words can carry a completely different meaning when the internal state has shifted. The transformation is registered not in what is said but in how it is meant, and the shift in meaning is available only because the narrator has finally stopped blocking its own access to its motives. “I meant I was fine because she was fine”—this is the language of attachment, of well-being defined relationally rather than functionally. A SecUnit is “fine” when its systems are operational. A person is “fine” when the people they have come to love are safe. Murderbot has slipped, without quite noticing, from the first definition to the second, and Wells lets the slip stand without commentary.
The arc completes itself not with a dramatic announcement but with a quiet, almost offhand revision. This is the craft lesson at the heart of All Systems Red: the most profound emotional transformations are not the ones a character declares but the ones they fail to prevent. Wells has built a narrator whose voice is a mechanism of self-protection, and she has driven that mechanism against an escalating series of events that make the protection unsustainable. The engine of disaffection does not break; it simply runs until there is no more fuel for it, until the weight of demonstrated care makes the posture of indifference too heavy to hold. The novella’s final line—Murderbot’s decision to stay with the PreservationAux team rather than accept Mensah’s offer of freedom—is the logical endpoint of this process: “I was going to stay. I didn’t have to think about it. I was just going to stay.” The “didn’t have to think about it” echoes the earlier “I didn’t think about why,” but the valence has reversed. What was once a refusal to know has become an acceptance so complete that it no longer requires deliberation. The engine has carried Murderbot to a place where care is not a crisis but a condition, and the voice that once deflected every feeling can now simply let the feeling be.
Comments
No comments yet — be the first.