24. How to talk with a man who doubts — and how to talk with a man who believes

A declaration before the chapter: this is a practical chapter, and its first half assumes one religion in particular, unlike the twenty-second chapter. So it is no part of the balance, and it is not to be read as a verdict on anyone. And whoever is on the other side should read the section set opposite it — that one was written for him.

And a second declaration, and it is a structural objection put to us: a doubting reader said: 'Publishing this chapter in the book that is given to me undoes it. If I have read the rules, then when my brother begins with "when did you start to doubt?" I shall not hear a question, I shall hear rule one.' And this is true, and it is intended. The alternative is that the rules be written for one side and hidden from the other, and that makes them a tactic and not a manner. And a rule that breaks once the other party knows it is a rule of manipulation and deserves to break. And what is left of them after the announcement — that you not show alarm, and that you say 'I do not know', and that you not debate him in front of people — works while it is known, and indeed works better. And we have added the last section as a like for like: how to talk with a man who believes, by the same rules reversed.

The question: your brother or your son or your friend has come to you with a doubt. What do you do?

The answer in two lines: six rules, and all of them branch off one — that you be prepared for him to stay where he is. And if you are not prepared, he will catch it off you in the first sentence, and the conversation is over before it begins.

1. Begin with the question and not with the answer

'When did you start to doubt?' opens. 'What is your evidence?' closes.

The reason: the second question puts him in the place of the defendant, so that he is taken up with protecting his position instead of examining it. The first asks for an account, and an account brings out what interrogation does not.

The example: the good doctor begins with 'what is troubling you and when did it start?' and not with 'prove to me that you are ill'. And the difference is not a matter of kindness; it is that the first answer carries diagnostic information and the second carries a defence.

And if he knows that it is a rule, say to him: yes, it is a rule written in a book you have read — and I am asking it because I want to know, not because it is number one. Honesty in this is heard, and pretence is smelt out.

2. Do not show alarm

Showing panic tells him: you are lost. And his position turns from a question he is examining into an identity he is defending. And an identity is not discussed; a question is discussed.

The example: the doctor opens the test results and his face goes pale. The patient has not read the sheet yet, and he has read the face. And after that he will not hear a single word as it was said.

3. Say 'I do not know' when you do not know

The reason is that a single manufactured answer is seen through, and all your sound answers fall with it in his eyes. And admitting the limit of your knowledge buys a trust you can spend where you do know.

The example: a mechanic guessed at the fault and replaced a part for no reason. The car broke down again, and you came to doubt everything he had told you before it — including what was true in it.

4. Separate God from the man who harmed him in His name — with no cheap acquittal

The bad form of it: 'religion is innocent, and what your father did does not represent it'. It is heard as a defence of an institution at the expense of the man in pain, and it calls him a liar about the plainest thing he knows — for the connection stands in his experience even if it does not stand in the ruling.

And what is of more use is three steps: that you call the act by its name (injustice, or cruelty); then that you grant that it did in fact take place in the name of religion; then that you put the separation to him as a possibility that belongs to him, not a duty imposed on him and not a condition of his being accepted.

And the psychological reason: much of the anger is directed at a father or a teacher or a community, and is then turned towards God because the man cannot turn it towards the one it belongs to. And naming the anger is half of its cure.

*And where it breaks down*: his anger may be at the ruling itself and not at the man who applied it, so that there is no father and no teacher in the matter at all. So do not assume the wound and do not go looking for it; whoever assumed it where it was not has said to the man: your objection is not an objection — and that is heavier than saying it outright.

5. Do not debate him in front of people

An audience makes stepping back a social loss, so that the saving of face wins out over looking into the matter.

The example: the man who was asked to apologise in a gathering of twenty men apologised with his tongue and grew more dug in within himself. And a word on the way home does what an hour on a platform does not.

6. Give him time

What was built over years is not pulled down in one sitting. And what has reached us from the accounts of those who have been through this — and it is a description that recurs in what has been passed on to us, not the result of a study we can refer you to, and we do not put it forward as a measured report — is that the road is long: first doubts, then a search for alternatives, then a turning point, then the building of a new life. And most of those you talk with have decided nothing yet.

The example: the man who poured the concrete and wanted to build on it the next day brought down the building and the concrete together.

And four sentences that are not said

And in the third row there stood 'and it is untrue in most cases', and a doubting reader said: 'most' means that it is true in some of them, so you did not void the charge but advised leaving it off because it is of no use. And the man across from me will sit down with a considerable probability in his mind that I am a liar — and you yourselves said that people smell that out. And he is right, and the fault runs deeper than the proportion: we possess no instrument by which we may know anyone's intention, so the proportion itself — be it small or large — is a claim with no ground with us.

And do not carry a broken instrument in your hand

Three things are put about in these gatherings and do not hold up:

1. 'Analytic thinking weakens faith'. The well-known experiment in which this was said (2012) failed to replicate in a direct, pre-registered replication on a larger sample (2017), and the title of the replication itself says: 'No evidence that analytic thinking decreases religious belief'. And take note: the replication does not establish the opposite — it does not say that analysis increases faith; it says that the original claim is without ground. So do not carry it and do not carry its reverse.

2. 'Science has proved that religion is an illusion of the brain'. The evolutionary account of the origin of belief in Pascal Boyer has the origin for its subject and not the settling of truth — and this is a claim about the book's subject to be seen in how it is built, not a declaration of neutrality on his part; indeed he holds religious claims not to be falsifiable at all. And the move from 'how it arose' to 'therefore it is false' is the genetic fallacy — and that is a rule of ours and not something transmitted from him. And its counterpart on our own side: whoever argues from the agency-detection device to 'the original disposition' is using a theory whose author — a committed Christian — built his book on its not overturning the belief and not establishing it either.

3. 'No morality and no society without religion'. Do not say it. Ara Norenzayan, whose thesis concerns the part religion plays in cooperation, devotes a chapter to secular societies that reached the same cooperation by way of alternative institutions. And whoever says this sentence in front of a man who knows his honest atheist neighbour has lost the whole conversation in a line.

And if you are the one in doubt: how to talk with a man who believes

And the rules are the same rules reversed, and we write them because a book that trains one side only is a book of mobilisation and not a book of inquiry:

1. Begin with the question and not with the answer. 'What was it that set your own mind at rest?' opens. 'Let me have your evidence' closes. And you will find that most of those you ask have never once been asked this question, and that his answer is not the argument you are expecting.

2. Do not alarm him. His alarm will make of your question a threat to his home and his family and his standing among people, and not a question in knowledge. And every door will shut in two minutes.

3. And accept 'I do not know' from him. If he says it he has given you the most truthful thing he has, and whoever punishes the man across from him for his truthfulness has taught him to lie. And this is the heaviest thing in the section for you: you want an answer, and he has given you the limit of his knowledge.

4. And separate religion from those who do well and do ill in its name — by the same rule you asked for yourself. For whoever rejects a claim by the acts of its followers has weighed by the criterion we ourselves declared to be the weakest thing in the balance, and it is the weaker on both sides.

5. And do not debate him in front of his family. The audience works on you as it works on him, and in his gathering you are the more exposed.

6. And give him time, and do not measure the conversation by whether he changed his mind in one sitting. And whoever measures it by that has used it as an instrument for winning — and that is the very thing we find fault with in our own people at the start of the chapter.

And four sentences that are not said on this side either: 'this is a desert superstition' (it cancels his person), and 'if you thought, you would leave' (it makes his remaining a verdict on his mind, and it is the sister of 'if you were sincere you would have been guided', letter for letter), and 'you are afraid of death, not convinced' (a verdict on an intention that cannot be seen, and it is the answering charge exactly), and 'science has settled this' (a promise that is not kept, and the first place that goes against it brings down all the rest).

And the limit of this chapter: these rules do not produce conviction and do not promise it. Their aim is to keep the conversation possible. And whoever measures them by whether he convinced or did not convince has used them as an instrument for winning, and people smell that out faster than they examine the argument.