Lachaud, one of the most famous defense barristers in France, is losing. For three-quarters of an hour he has aimed his most cunning arguments at one obstinate juryman — the seventh, first on the second bench — and the man will not move. The case is desperate.
Then, in the middle of a passionate demonstration, Lachaud stops. He turns to the president of the court and says: "Would you give instructions for the curtain there in front to be drawn? The seventh juryman is blinded by the sun." The juryman reddens, smiles, thanks him. And he is won over for the defense.1
No argument did that. A small human kindness did — a moment of being seen and cared for. Le Bon uses the scene to make his point about juries, which is the point he makes about all crowds, sharpened to a courtroom: a jury does not decide by weighing evidence. It decides by sentiment. Get the feeling right and the verdict follows; get the argument perfect and lose anyway. The man who understood this beat the man who had the better case.
Start with the fact that most unsettles the legal profession. For a long time, French administrators carefully picked jurors from "the enlightened classes" — professors, functionaries, men of letters. Later, juries were drawn mostly from small tradesmen and clerks. And the verdicts came out the same.2
A former President of the Court of Assizes, Bérard des Glajeux, admitted it in his memoirs: "whatever the composition of the jury has been, its decisions have been identical."3 Swap the professors for the grocers and nothing changes. Le Bon's reading is exactly his general law of crowds: the moment individuals form a deliberating body, "intelligence stands for nothing." A gathering of distinguished specialists returns judgments "not sensibly different" from a gathering of masons.4 The jury's output is a property of the crowd form, not of the people in it — which is why you can't improve a jury by improving its members.
The proof in practice: the great barrister Lachaud used to burn his objections striking every intelligent-looking person off the jury list. Experience eventually taught the whole Parisian bar that such objections were useless — public prosecutors and defenders simply stopped bothering, "and still the verdicts have not changed, they are neither better nor worse."5 You cannot engineer the outcome by selecting cleverer jurors, because cleverness was never the operative variable.
If not intelligence, what? Feeling. "Like all crowds, juries are very strongly impressed by sentimental considerations, and very slightly by argument."6
Le Bon quotes the practitioners without flinching. A barrister: jurors "cannot resist the sight of a mother giving its child the breast, or of orphans." Des Glajeux: "It is sufficient that a woman should be of agreeable appearance to win the benevolence of the jury."7 The nursing mother, the orphans, the pleasant face — these move verdicts. The chain of legal reasoning mostly doesn't. And juries are "profoundly impressed by prestige": very democratic in who sits on them, "they are very aristocratic in their likes and dislikes." Name, birth, wealth, celebrity, a famous counsel — "everything in the nature of distinction or that lends brilliancy to the accused, stands him in extremely good stead."8
So the good counsel's job, Le Bon says, is to "work upon the feelings of the jury, and, as with all crowds, to argue but little, or only to employ rudimentary modes of reasoning."9 The English barrister he quotes watches the jurors' faces phrase by phrase, reading the effect of each sentence, adjusting in real time — which is why a speech prepared in advance is nearly useless, because it can't respond to the feeling in the room. The lawyer isn't building a proof. He's managing an emotional state.
Here's the mechanism that turns this from observation into a usable tactic. A counsel doesn't need to convince the whole jury. He needs "only the leading spirits among it who will determine the general opinion."10
Le Bon's quoted barrister: "I have found by experience that one or two energetic men suffice to carry the rest of the jury with them." So the real target isn't twelve people — it's the two or three whose conviction the other nine will follow, exactly as any crowd follows its leaders. And the first move on those two or three isn't to convince them. It's to please them: "The man forming part of a crowd whom one has succeeded in pleasing is on the point of being convinced, and is quite disposed to accept as excellent any arguments that may be offered him."11
That's why the sunlight worked. Lachaud's obstinate seventh juryman was one of the leaders — the one the verdict hinged on — and the curtain didn't argue with him, it pleased him. Pleased, he became "disposed to accept as excellent" whatever came next. Find the two or three who lead, please them before you try to convince them, and the body swings.
Watch the sentiment-logic produce a consistent, almost coherent pattern — one that reveals what juries are actually doing under the feeling.12
Juries, Le Bon observes, are "very indulgent in the case of breaches of the law whose motive is passion." They're rarely severe on infanticide by a girl-mother, or on the young woman who throws vitriol at the man who seduced and abandoned her. But they're without pity for crimes "of which it appears possible they might themselves be the victims" — the crimes most dangerous to ordinary people.
Look at the line they're drawing. It isn't the law's line. The law, "equal for all," would treat the abandoned girl's crime and the burglar's murder by the same code. The jury instinctively splits crimes into dangerous to society and not dangerous to society, and forgives the second. The seduced girl who avenges herself, Le Bon notes, is even "rather useful than harmful, inasmuch as it frightens seducers in advance." The jury feels its way to a rough social calculus that the rigid law can't make.
This is the case that complicates the whole "crowds are stupid" reading. The jury reasons badly and decides by feeling — and its feeling encodes a genuine distinction the law is too blunt to draw. The sentiment isn't just error. It's a crude instrument sensing something real about danger and desert that the code, blind and equal, cannot sense.
You will face juries that aren't legal — hiring panels, review committees, boards, a jury of peers in any room where a few people decide together. The mechanics transfer directly.
You're presenting to a committee and you've built an airtight logical case. Le Bon warns the airtight case is aimed at the wrong faculty. Before the argument, get the feeling right: make them comfortable, make them feel respected and seen, land one concrete human image instead of ten data points. The proof matters far less than whether the room feels good about you and the thing you're proposing. Argue little; move the sentiment.
Find the two or three who'll carry the rest. Don't spend equal effort on everyone in the room. Every deciding body has a couple of leaders whose conviction the others will follow. Identify them early — who defers to whom, who speaks and gets nods — and concentrate there. And on those two or three, please before you persuade: the person you've pleased is already halfway to accepting your case.
You're the obstinate holdout being worked on. Now run it in reverse as a defense. If someone is being unusually attentive to your comfort, flattering your judgment, making you feel seen right before asking for your agreement — notice the sequence. Being pleased is the state in which you become "disposed to accept as excellent" whatever comes next. The sunlight is a real kindness and a softening. You can accept the kindness and still hold the argument to its own merits.
You want a group to reach a fair decision, not just a swayed one. Le Bon's mechanism is also a warning about your own panels. If verdicts don't depend on who's in the room but on who manages the room's feeling, then the "fairness" of a committee is more fragile than it looks. Building in independent written positions before discussion, or separating the deciders, blunts the leading-spirits capture — the same corrective the collective-hallucination problem needs.
Two misreadings.
Failure one — concluding juries are worthless. Le Bon's own conclusion is the opposite, and it's the next section. The jury decides by sentiment, yes — but that sentiment does real work the law can't, and the alternative is worse. Don't let "decides by feeling" collapse into "should be abolished"; Le Bon fought hard against exactly that inference.
Failure two — treating the tactics as courtroom-only. The sentiment-over-argument mechanism, the leading-spirits capture, the please-before-convince rule — these are properties of any small crowd that decides together. Fencing them into "how lawyers manipulate juries" misses that they run in every panel, board, and committee you'll ever face. The courtroom is just where they were studied most closely.
Then Le Bon turns, and defends the thing he just exposed. Reformers wanted to abolish juries and hand verdicts to professional magistrates. Le Bon says that would be a disaster — and here he introduces a second kind of crowd, more dangerous than the jury: the caste.13
The magistracy, he notes, is "the only administration whose acts are under no control." An examining magistrate fresh from university can jail a person of any standing on a mere supposition, hold them for a year, and free them without indemnity or apology — a power Le Bon compares to the old lettre de cachet. And the accused reaching a jury has already been judged guilty by several magistrates; "the errors of juries have always been first of all the errors of magistrates." So replacing the jury with judges wouldn't remove error — it would remove the accused's "only chance of being admitted innocent."
The load-bearing line, and the reason this page ends where it does: "The power of crowds is to be dreaded, but the power of certain castes is to be dreaded yet more. Crowds are open to conviction; castes never are."14 A jury can be moved — by sentiment, prestige, a nursing mother, a drawn curtain — and that very movability, which makes it manipulable, also makes it reachable, temperable, capable of mercy the rigid law forbids. A professional caste is closed: same faces, same assumptions, accountable to no one, incapable of being moved. Le Bon's paradox: the jury's weakness — that it can be swayed — is exactly its value, because a body that can be swayed can also be swayed toward mercy and truth, and a caste that cannot be swayed cannot be corrected at all.
Evidence Le Bon offers. Des Glajeux's memoir admission that verdicts stayed identical across radically different jury compositions; the Parisian bar's abandonment of juror objections as useless; the practitioner testimony on sentiment (nursing mothers, agreeable women, prestige); Lachaud's sunlight trick and the leading-jurymen tactic; the consistent indulgence pattern for crimes of passion; and the institutional comparison of jury vs. unaccountable magistrate. The composition-independence evidence is strongest — it's a near-natural-experiment.
Tensions. Le Bon holds two things that pull against each other: juries decide irrationally by feeling, and juries should be preserved because their feeling is wiser and more merciful than the law. He never fully reconciles "the sentiment is error" with "the sentiment encodes a real social calculus the code can't." The crimes-juries-forgive pattern suggests the sentiment is doing something more like rough justice than pure irrationality — which complicates his own broader "crowds lower intelligence" thesis. A second tension: the practitioner quotes reflect the gender assumptions of 1895 (the seduced girl, the agreeable woman); the mechanism transfers but the examples are period-bound.
Open question. If the jury's value is precisely that it can be moved — reachable, temperable, mercy-capable — where a closed caste cannot, does that generalize into a principle for institutional design: prefer movable bodies to closed ones for decisions that need mercy and correction, even at the cost of manipulability? Filed as an ongoing thread.
Set this beside Electoral Crowd and Committee Tyranny and the pair maps two applied crowds in the same civic machinery. The electoral crowd chooses by prestige and formula and is captured by committees; the jury decides by sentiment and prestige and is captured by two or three leading spirits. Same underlying law — a deciding body run by feeling, prestige, and a small controlling minority — deployed once at the ballot and once at the bench. But they split on Le Bon's verdict: the electoral committee is "the most redoubtable danger," while the jury is a thing to "cling to vigorously." The difference is the alternative. There's no less-crowd-like substitute for the electorate that Le Bon trusts, but the jury's rival — the magistrate caste — is worse, so the jury's crowd-nature becomes a feature. Reading them together shows Le Bon isn't anti-crowd on principle; he's comparing each crowd to its actual alternative.
Bring in The Leaders of Crowds. The leading-jurymen tactic is that page's law at small scale: a crowd cannot do without leaders, so twelve jurors reduce to two or three whose conviction the rest adopt. This page adds the courtroom-tested refinement — please the leader before you convince him — which sharpens the general leader-psychology into an operational sequence. The convergence: whether it's a nation or a jury of twelve, you move the body by moving the few, and you move the few by getting their feeling right first.
Psychology — Psychological Crowd and the Law of Mental Unity. The mental-unity law says twelve individuals become one deciding creature whose intelligence drops below any member's. The jury is that law's cleanest natural experiment: swap professors for grocers and the verdict is identical, because the deciding entity is the crowd-form, not the members. The insight neither produces alone: a jury's composition is nearly irrelevant to its verdict, which means every effort to improve justice by improving who sits is misdirected — the leverage is in the crowd-conditions (how feeling and leading spirits are managed), not in the roster. That reframes jury reform entirely: you don't fix the output by selecting better people, because the output was never a function of the people. It's visible only where crowd psychology meets courtroom practice.
History — Institutions as Effects, Not Causes. Le Bon's defense of the jury is an institutions-as-effects argument turned into a choice between two institutions: the jury and the magistrate-caste are both expressions of how a society distributes the power to judge, and Le Bon prefers the movable institution to the closed one precisely because a movable body can be corrected and a closed one can't. The insight where they meet: the decisive property of a judging institution isn't its accuracy but its reachability — whether it can be moved toward mercy and away from error at all. A caste optimized for consistency is optimized for being un-correctable, which is why Le Bon dreads it more than the sway-able crowd. Neither crowd psychology nor institutional history alone gets you to "prefer the manipulable body because manipulable means correctable"; that lives in the overlap.
The Sharpest Implication. The manipulability you'd want to design out of a jury is the exact property that makes it worth keeping. A body you can sway with a nursing mother and a drawn curtain is a body you can also sway toward mercy the rigid law forbids and toward doubt the prosecution's caste never entertains. Take this seriously and a deep intuition inverts: the "objective, professional, unswayable" decision-maker we tend to trust more is precisely the one that can never be reached, corrected, or moved toward mercy — closed, consistent, and accountable to no one. Sometimes the reachable, feeling-driven, manipulable crowd is the safer institution, not despite its softness but because of it.
Generative Questions