SHTIKA · SILENCE
This essay puts the exceptions before the trait, because taught in the other order it has been used to keep people quiet about harm.
Every other essay in this collection introduces a trait and then qualifies it. This one is arranged the other way round, deliberately, because a restraint-in-speech teaching delivered in the usual order has a documented history of being used to stop people reporting harm.
So, before anything else: this trait does not apply to reporting harm, to warning someone about a person who could hurt them, to raising a safety problem, to making a complaint, or to telling the truth to somebody who needs it to make a decision. Not as a grudging exception — those are simply not what the trait is about.
And the tradition draws two distinct things here, which are worth separating because they are often run together. One is permission: to’elet, constructive purpose, under which otherwise forbidden speech becomes allowed. The Chofetz Chaim sets conditions on it — direct knowledge rather than hearsay, no exaggeration, the constructive end actually achievable, told only to those who need it, and no greater harm done to the subject than the case requires. The other is obligation: Leviticus 19:16 forbids standing idly by the blood of your neighbour, so where somebody is at risk, warning is not merely permitted but required. The difference is not that the conditions vanish — you still have to be accurate, and you still must not exaggerate, which are requirements of truthfulness rather than concessions to the prohibition. The difference is what the conditions are for. Permission weighs whether the speech is worth making at all. Obligation does not put that question, and it does not require you to satisfy an institution that your grounds are adequate before you may speak.
And this is not an outsider softening the tradition to make it usable. The tradition drew these exceptions itself, at length and in detail. The Chofetz Chaim’s 1873 treatment of harmful speech, which is the most developed account of the subject in any literature I have read for this library, devotes substantial space to the circumstances in which speech that would otherwise be forbidden is instead permitted, and in some cases obliged — where there is a constructive purpose, where someone is at risk, where a warning prevents harm. Those are two different things and the essay keeps them apart below.
The exceptions are in the source. What happens in practice is that the restraint gets taught and the exceptions do not, and the essay below is arranged to make that harder.
With that established: the domain of this trait is a remark about a person, made to someone who is not that person, which serves no purpose.
That is it. Not silence in general, not reticence as a personality, not keeping your views to yourself. One specific act, which most people perform several times a day without registering it.
The reason it gets its own trait is that it is the commonest form of harm that leaves no evidence and produces no obvious victim, which makes it almost impossible to notice yourself doing.
The saying is characteristically dry and it is talking about the body rather than the soul, which is a joke about consequences.
It sits in a collection of sayings that elsewhere insists on speaking up, which is the balance the practice needs and the balance the practice usually loses.
The category of lashon hara — harmful speech — is developed in this literature more finely than anywhere else I have found.
It distinguishes speech that is true from speech that is false, and forbids a great deal of the true kind. It distinguishes what is said, to whom, in what setting, with what effect. It considers the speaker’s motive separately from the content.
And in the same works, it sets out when the prohibition lifts. That structure — a detailed prohibition with a detailed set of conditions under which speech is instead obligatory — is the achievement, and it only works as a whole.
The half that travels is the prohibition. The half that does not travel is the exceptions, which is exactly the wrong half to lose.
Brad Bushman’s work dismantled the catharsis hypothesis.4 Venting anger — including verbally, including to a sympathetic third party — does not reduce it. In his experiments it reliably increased subsequent aggression compared with doing nothing at all.
That is a narrow finding and it is the relevant one. The usual defence of the remark about a person is that it discharges something. It does the opposite: it rehearses the grievance, sharpens it, and makes the next expression more likely.
It does not follow that all speech about difficulty is harmful, which would be an absurd generalization from these studies. Talking about a problem in order to work out what to do about it is a different act with a different evidence base. The finding covers expression of anger as an end in itself.
The objection here is not philosophical and it is the most serious in the collection.
The case against
This is not a hypothetical risk and it should be evidenced rather than gestured at. Australia’s Royal Commission into Institutional Responses to Child Sexual Abuse examined this directly and, in its 2017 final report, recommended that all Jewish institutions in the country state explicitly in their complaint-handling policies that the halachic concepts of mesirah, moser and lashon hara do not apply to reporting allegations of child sexual abuse to police. A government inquiry does not make a recommendation of that shape unless the misuse was found to be occurring.
Rabbi Mark Dratch’s paper “Let Them Talk: The Mitzvah to Speak Lashon Hara” makes the argument from inside the tradition: that the prohibition is routinely invoked to silence victims and their advocates, and that speaking is not merely permitted for them but obligatory.
The mechanism is simple and it does not require bad faith from most participants. Teach restraint thoroughly and the exceptions lightly, and a person with a genuine report will apply the restraint, because they are conscientious and because that is the half they were taught.
The guard is structural rather than personal: the exceptions get stated first, every time the trait is taught, and anyone teaching it who does not state them should be assumed to have got it wrong. That is why this essay is arranged the way it is.
Count for two days before changing anything. Remarks about a person, to someone who is not that person, serving no purpose. The number is the finding, and nobody guesses it correctly.
Before each one, one question: what is this for? Not whether it is true — accuracy is not a justification, and the tradition is emphatic about that. What it accomplishes.
If the answer is that somebody is at risk, this is not a question about the trait. It is the obligation not to stand idly by, and it goes to whoever can act — which, where a crime may have occurred, means the authorities. You still owe accuracy: say what you know and not more. What you do not owe is a prior judgment about whether the case is strong enough to be worth making, and you do not need anyone’s permission to make it.
If the answer is a lesser constructive purpose — a decision somebody has to make, a problem that needs solving — then the permission conditions apply: know it first-hand, do not exaggerate, tell only who needs it, and go no further than the purpose requires.
And keep the domain narrow. This trait is not about being less talkative, less direct, or less willing to disagree. Applied outside its one target it becomes a general timidity, which is not what anyone was arguing for.
The count comes first and it is uncomfortable.
The tradition that produced the most detailed account of harmful speech I have found also wrote the exceptions in detail, in the same works, for the same reason.
It only functions as a whole. Taught as a prohibition alone it becomes the instrument that every institution needing to conceal something has reached for.
So the trait is narrow, the exceptions come first, and the test for anybody teaching it — including this essay — is whether they stated the exceptions before they stated the rule.
Count for two days. Remarks about a person, to someone who is not that person.
Then one question before each: what is this for? If somebody is at risk, that is not this trait at all — it goes to whoever can act.
Hebrew terms are transliterated without diacritics. Classical sources are cited by work and location; where a formulation is modern it is named as modern. Modern research is cited by author and publication.