A young historian sits down across from you and asks: what are the core principles? MacCulloch's answer comes back in two beats. "Be skeptical and then be sympathetic. And the skepticism stops you being too sympathetic."1 The order matters. Sympathy without skepticism produces hagiography — the heroic biographer who falls in love with their subject and quietly stops noticing the inconvenient evidence. Skepticism without sympathy produces dismissiveness — the cynic who reduces every historical figure to their worst motive and misses what the person was actually trying to do. MacCulloch's principle is that the skepticism is what makes the sympathy honest. You distrust first, hard, and the residue that survives the distrust is what you can warm to without lying.
The skepticism isn't generic doubt. MacCulloch operationalizes it as a sentence: "Why is this person saying this lie to me?"1 That sentence does three things at once. It assumes there is a lie (or at least a slant). It asks for the motive (why). And it puts the question in the present tense, as if the source were sitting in the room — even when the source is a 16th-century document. The historian becomes, in that moment, an interrogator of motive. Not necessarily an aggressive one. But someone who cannot let the source's account stand on its own authority.
This is closer to the question a behavioral-mechanics practitioner runs than to the question a deferential reader runs. Hughes' behavioral profiler asks variants of the same thing: what does this person need from me, what are they trying to hide, what would have to be true for them to be saying this? The historian is doing motive-interrogation on the dead. The difference is that the historian has to do it without the live signals — no tone, no body, no reaction time — so the technique has to be more disciplined and less intuitive.
Once that interrogation has run, the sympathy can arrive on safe ground. The historian, MacCulloch says, is also alert to the fact that "they are basically interesting because they are human beings."1 After the skepticism has stripped the source's account of its pretensions, what remains is a human being — flawed, motivated, often confused about their own role — and that human is intrinsically interesting and worth understanding. The two-beat posture is skepticism in service of accurate sympathy, not skepticism instead of sympathy.
There's a structural claim hiding inside MacCulloch's principle that does most of the load-bearing work: history is about human beings. "Human beings create structures and that has a history in itself, but at the center are fallible human beings. So, all the time you're looking at people in the context of events which they usually don't control, even if they think they do."1
Three pieces stack here. First, the ontology is human-centered, not structure-centered. The structures matter, but they are made by humans and remade by humans, and the humans are the unit of analysis. Second, the humans are fallible — not just morally but cognitively. They make mistakes about their own situation. Third, and most important, they usually don't control the events they're inside, even when they think they do. The protagonist of the historical episode is also its captive.
This last move is what lets skepticism-then-sympathy work without collapsing into either cynicism or apology. The cynic says the historical actor was a fool or a knave; the apologist says they were doing their best under the circumstances; the MacCulloch stance says they were operating with limited information and limited control, often inside events much bigger than they could see, and they were lying to themselves about how much agency they had — and that's the normal condition of being a person in history, not a special pathology. Once that's the baseline, the skepticism doesn't need to be a moral indictment, and the sympathy doesn't need to be exculpatory. The historian is looking at human beings who didn't know what they were doing, doing it anyway, and constructing post-hoc accounts that papered over the not-knowing. The two-beat posture is exactly the apparatus that recovers what was really happening through what they said was happening.
In practice the historian runs both beats on every source. The operational version looks like this:
Skepticism beat. What is this source's interest in the story it's telling? Who benefits from the version on the page? What is being foregrounded and what is being suppressed? What does the author need to be true for their own self-conception to hold? What does the document not say? What can be checked against external evidence? When MacCulloch reads Bede on the Anglo-Saxon Christian mission, the skepticism beat asks: Bede is a monk writing inside a monastic culture; what would he need to suppress to keep the story heroic? When he reads Churchill on the Second World War, the same beat asks: Churchill was an essential player; what would he need to suppress to keep himself essential?
Sympathy beat. Once the skepticism has done its work, what's the actual person on the page? What were they trying to do? What did they think they were doing? What were they protecting? What were they unable to see? Bede was preserving the Anglo-Saxon Christian inheritance for a culture that might lose it; Churchill was writing a memoir that was also a contribution to public memory and a defense of his own decisions. The sympathy beat doesn't excuse the distortion the skepticism beat detected — but it lets the historian write the person as a person rather than as a foil.
The danger of skipping either beat: skip skepticism and you absorb the source's frame as your own; skip sympathy and you produce a flat dismissal that misses what the source was actually doing. MacCulloch's discipline is to never let one run without the other.
Read flat, this looks like common sense. But the specific operationalization ("Why is this person saying this lie to me?") is unusually sharp. It commits the historian to assuming a lie (or a slant) before reading — which is methodologically aggressive. The alternative principle, sometimes called the "principle of charity," runs the opposite order: assume the source is telling the truth as they see it until forced otherwise. MacCulloch's order is structurally inverted from that.
Tension: the inversion makes sense for sources where the writer has a stake (memoirs, official histories, propaganda). It works less well for routine administrative documents, archaeological evidence, or fragmentary primary sources where there isn't an obvious motive for distortion. MacCulloch doesn't address the second class directly, but the implication is that the principle is calibrated for witness sources — sources where someone is testifying about events they care about. The shift in source-type would shift the calibration.
A second tension: the "even if they think they do" clause about historical actors' lack of control is itself a strong claim that needs evidence. Some historical figures clearly did control the events around them at certain moments (a general winning a battle, a politician passing a law). MacCulloch's position would be that even these moments of control are smaller than the actor experienced them as — that there are always structural conditions making the controlled action possible, and the actor's sense of authorship is partly illusory. That's defensible but it's not obvious, and pushing too hard on it slides toward determinism.
Open questions:
Try running the two-beat with a live source in a room. You're a Hughes-trained interrogator across from a person you suspect is lying. You don't start with skepticism — you start warm. Mirror their posture. Match their breathing. Let them feel like the conversation is going to be okay. Only after the rapport has lowered their defenses do you start the actual probing. Skepticism arrives inside the conversational frame the sympathy created.
Now do MacCulloch's job with the same person — except they're dead, and what's in front of you is a single document they wrote. There's no rapport to establish. There's nothing for the sympathy to lower. The defenses you'd be trying to drop aren't in the source anymore; they're in you. So the order flips: skepticism first, hard, because the only thing skepticism can break here is your own credulity. Then sympathy, after the document has been pressed for what it isn't saying.
The order isn't arbitrary in either case. It's set by whether the source can still respond. Live sources can be worked with inside the conversation, so rapport-first makes the skepticism more productive. Dead sources can't, so skepticism-first is the only way to keep the historian from absorbing the source's slant as her own.
Now run Lieberman alongside both. His reader is the cop, the negotiator, the person who needs accurate motive-reads in real time. She can't afford the sympathy beat — there's no second meeting where she'll come back and try to understand the suspect as a person. MacCulloch can. The historian has the long time-horizon that makes the sympathy beat affordable. The cop doesn't. Both disciplines are correct for their conditions; what changes between them is what the situation permits.
The two-beat posture is most useful when held against the manipulation-and-deception literature on one side and the depth-psychology literature on the other. Skepticism-then-sympathy is structurally what a clinically-trained observer does, and it shows up in radically different domains under different names.
Behavioral Mechanics: Deception Detection — the operational form of motive-interrogation. MacCulloch's "why is this person saying this lie to me?" is the verbal version of the same question Hughes' DRS arithmetic or Lieberman's adjective-column profiling is running on live conversation. The structural parallel is the assumption that the source has a slant and that the slant is recoverable by patient interrogation. The split is that historical sources can't be re-questioned — the historian has only the single document. So the historian's deception-detection has to be more saturated (running on every line) and less conclusive (no follow-up question available). What the handshake produces is the recognition that historiographic skepticism is deception-detection on cold material, and that the techniques developed for live interrogation can be partially adapted to documentary sources, with appropriate adjustments for the absence of feedback.
Behavioral Mechanics: Elicitation Framework (Hughes 6MX) — Hughes' rapport-then-extraction is structurally inverted from MacCulloch's skepticism-then-sympathy. The inversion reveals that the order depends on whether the source is alive: live sources need rapport-first because skepticism can be deployed inside the conversation; dead sources need skepticism-first because the document has no defenses to lower. The handshake produces a meta-principle: the optimal sequencing of skepticism and sympathy depends on whether the inquiry is dialogical or documentary. The historian and the interrogator are doing structurally identical work but in inverse temporal orders, governed by what the source can and cannot still do.
Psychology: Evolutionary Psychology Hub — Trivers' work on self-deception and Buss's error-management theory both describe why human beings reliably distort their own accounts of their own behavior. The MacCulloch ontology ("fallible human beings... events they usually don't control, even if they think they do") is the historiographic version of the EP claim that humans evolved to misread their own agency. The handshake produces a useful frame: historiographic skepticism isn't moralism; it's adjusting for the standard cognitive distortions humans bring to their own self-accounts. The historian who runs the skepticism beat is calibrating for documented features of human cognition, not assuming bad faith. That reframes the principle as empirical rather than ethical.
Eastern Spirituality: Buddhist Philosophy Hub — the Buddhist analysis of self-grasping describes a structurally similar phenomenon: the conventional self is constantly constructing post-hoc narratives that overstate its own agency and coherence. The historian's "even if they think they do" clause about historical actors' lack of control parallels the Buddhist analysis of the agent-illusion in moment-to-moment cognition. The handshake produces a long-arc claim: the disciplined historian is doing, at the historiographic level, what the contemplative practitioner is doing at the cognitive level — recovering the actual operations from the self-narrating overlay. The further implication is that contemplative training and historiographic training have an underexplored cross-fertilization potential, with each tradition's apparatus refining the other's.
The Sharpest Implication
This isn't really about historians. If MacCulloch is right that human beings are reliably wrong about their own agency and their own motives, then your own retrospective account of your own past is exactly the kind of source the skepticism-then-sympathy beat is designed for. You are a witness about yourself. You have a stake in the version you give yourself. You suppress the parts that don't fit the story. You overstate your own control over events that mostly happened to you. The same two-beat posture MacCulloch runs on Cromwell or Churchill is the posture you'd need to run on your own memoir of last year — and you won't, because the skepticism beat hurts when the source is you. What this means is that personal self-knowledge is a historiographic problem, and the discipline that produces accurate history is structurally identical to the discipline that would produce accurate self-knowledge. Most people don't have either.
Generative Questions