Through the Spirit

Companions: Then We Stopped, on how the argument ended, and The Fellowship of Suffering.

The ship had cargo to unload, and it took a week.

They came into Tyre with nothing to do, and they went looking for the church. Luke’s word is that they sought the disciples out — in a city that size, with so few of them, you would have to.

Then seven days with a church he had not planted, in a city he was only passing through.

And at some point in those seven days, this:

Through the Spirit they were telling Paul not to go on to Jerusalem.

The verb is imperfect. They kept telling him. It was not a word delivered once, on the first evening, with everyone falling silent afterwards. It went on. Over a week, in a small house, among a handful of believers who had known the apostle Paul for seven days, the same thing was said again and again, and he did not go home at the end of it to think it over, because he was staying with them.

Luke was there. He writes we.

And then Paul went to Jerusalem.

The difficulty arrives before any of the technical questions do. Something was said to Paul through the Holy Spirit. Paul did the opposite. The sentence is eleven words in Greek and it sits there in the middle of a travel narrative, unexplained, as though Luke did not notice what he had written.

So the question presents itself, and it is not a clever question or a modern one: was Paul disobedient?

If the Spirit said do not go, and he went, then the man who wrote most of the New Testament set aside a direct word from God on the most consequential journey of his life — and Luke, his friend and travelling companion, recorded it without comment.

It deserves more than a footnote.

What the sentence says

The Greek is eleven words, and most of the difficulty is in one of them.

Three things are slower than they look.

The disciples are doing the speaking. The verb has them as its subject, not the Spirit. The Spirit appears in a phrase hung off the side of the sentence — dia tou pneumatos, through the Spirit. Nowhere in the verse does God say anything. People say something, and Luke tells us the Spirit was involved in their saying it.

There is no formula. Seven verses later Agabus will do it properly: Thus says the Holy Spirit, with the words that follow quoted. That is what a prophetic oracle looks like in Luke’s hands, and it is not what we have at Tyre. What we have is reported speech — they told him not to go — which is the ordinary way Greek relays an instruction from one person to another.

And the instruction itself is firm. Nothing soft about it. Not they worried, not they wondered whether. They told him not to set foot in the place.

So what does through the Spirit actually cover?

It can go two ways, and the grammar will not decide between them. Either the Spirit gave them the whole sentence, prohibition included — in which case Paul was told by God not to go, and went. Or the Spirit gave them something true about Jerusalem, and the prohibition is what they made of it.

Both readings fit the words. The syntax will not settle it, and a verdict that sounds settled has been brought in from outside.

I want to resist tidying that up, because the ambiguity is not a flaw in the sentence. It is the shape of the thing being described. A word from God arrives inside a human sentence, and by the time you are hearing it the two are not easily separated. That is true at Tyre, and it is true when someone sits down opposite you and says they have been praying about you.

Luke could have written it so the question did not arise. He had the tools — he uses them seven verses later. He did not.

The seam at Caesarea

Luke writes the same scene again, and this time he shows his working.

At Caesarea a prophet comes down from Judea. Agabus — Luke has used him before, and he was right before. He takes Paul’s belt, ties his own hands and feet with it, and says:

Thus says the Holy Spirit, “This is how the Jews at Jerusalem will bind the man who owns this belt and deliver him into the hands of the Gentiles.”

Read the oracle. All of it. It is a prediction and nothing else. This is what will happen in Jerusalem, and here is a picture of it so you will not forget. There is no instruction anywhere in it. Agabus does not tell Paul to do anything.

Then verse 12.

When we heard this, we and the people there urged him not to go up to Jerusalem.

That is the seam, and Luke stitches it in the open. The Spirit said one thing. Everybody else said another. Between them sits a therefore that nobody spoke aloud and everybody agreed with — bonds are coming, so stay away.

The we matters. Luke is not describing other people’s mistake. He is in it. He heard the oracle, he drew the conclusion, he added his voice to the urging. If this chapter has a confession in it, verse 12 is where it sits.

Inference is not the fault. Unexamined inference is.

And nobody argued for the conclusion. There is no debate recorded, no dissenting voice, no one asking what the prophecy actually said. It did not need arguing. Everyone in that house already knew that when God shows you a danger he means you to avoid it.

Which is what makes verse 4 legible. Luke has given us the long version at Caesarea and the compressed version at Tyre — the same event twice, once with the seam visible and once with it sewn shut. At Tyre he wrote through the Spirit they were telling Paul not to go, with revelation and conclusion folded into a single clause. At Caesarea he unfolded it.

It does not prove the Tyre reading. It does show that Luke knew the difference between what the Spirit said and what everybody concluded, and that he expected his reader to know it too.

Sixteen centuries without agreement

If that settled it, the church would have noticed by now.

It has not. The passage has been worked over for sixteen centuries and the readings have not converged. What follows is three of them, and the useful thing about them is not which one wins. It is that each one has to give something up.

The Spirit told them what; they supplied the therefore.

This is the common reading and the oldest. Chrysostom has it: the Spirit showed them the dangers, and the brethren, sparing him, added of themselves that he ought not to go. Most modern commentators follow. Marshall calls it the simplest solution. Barrett has the friends acting under the influence not of the Spirit but of human concern.

It fits Caesarea exactly. What it costs is verse 4 as written. Luke says they spoke through the Spirit, and this reading hears they spoke, having been informed by the Spirit — which is available in the Greek, but is not the first thing the sentence says.

The word was genuinely the Spirit’s, and they were not big enough for it.

This is Calvin, and it is not what he is usually quoted as saying. He grants what the first reading quietly declines to grant: these men spoke by the Spirit of prophecy, and the Holy Ghost did dissuade Paul from the journey. He raises the obvious objection himself — is the Spirit contrary to himself? — and answers it with the gifts. There are diverse gifts of the Spirit, and it is no marvel if men who excel in prophecy are sometimes destitute of judgement. The Lord showed these brethren what was coming. He did not show them what Paul’s calling required, because the measure of their gift did not reach that far.

That is a harder and more interesting proposition than the one usually put in Calvin’s mouth. An authentic word, accurately received, correctly reported — and still not the thing to do.

Wayne Grudem arrives at the same place four centuries later by a different road, arguing that prophecy in the New Testament can be genuine in its impulse and fallible in its wording. The two men would not agree about much else in the vicinity. They agree here.

What it costs is comfort. If a true word can come from the Spirit and still be the wrong thing to do, then hearing it correctly is not the end of the work. You can be sure of what you heard and still not know what to do about it.

He was told not to go, and he went.

Ray Stedman takes the sentence at its word and accepts what follows. He calls it Paul’s mistake: it was not necessary for him to go, and the Spirit had told him not to. He gets there by pressing the phrase this article is named for — to read the verse any other way, he says, is to ignore three crucial words. He does not enjoy saying so. He calls it a failure, and a partial one, and he finds its root in an urgent hunger to accomplish a goal God had not given him to do — the longing of Romans 9, to be cut off for the sake of his kinsmen. That is not a small thing, and it is not a dismissal.

What it costs is the rest of Acts. Paul says he goes bound in the Spirit. The Lord stands by him in the barracks afterwards and tells him that as he has testified in Jerusalem, so he must testify in Rome — which is commendation, and contains no rebuke at all.

Three readings. One lets go of the plain sense of verse 4. One lets go of the idea that a word confirmed is a question closed. One lets go of Acts 23:11.

Nobody has found a reading that gives up nothing, and sixteen centuries of people considerably better equipped than me have had a go.

What the machines did

I did not read all of that myself.

I worked on this passage with two AI research assistants, which I sometimes do, and I want to tell you what they did.

Both of them gave me Calvin. Both of them gave me the same Calvin — the tidy one, the one who says the Spirit revealed the danger and the brethren’s affection supplied the prohibition. One of them put it in quotation marks and attributed it to the commentary. It was not there. The sentence had never been written.

I asked for it to be checked. The check found the fabrication, acknowledged it, apologised — and produced a second quotation, also invented, this time attributed to the Geneva Bible. I asked again. A third invented quotation, from a different author.

Three passes. Three fabrications. Each one produced in the act of correcting the one before it.

What I want you to notice is not that machines make things up. Everyone knows that by now, and it would be a dull thing to write about.

What I want you to notice is that the inventions all said the same thing.

Not one of them produced a Calvin who was harder to use. Every single fabrication tidied the passage in the same direction — toward the reading where nobody has to be uncomfortable, where the Spirit’s part and the human part come apart cleanly and the prohibition can be set aside without cost. When the real Calvin finally turned up, he was saying something considerably more awkward, and no amount of machine assistance had been able to find him.

These things are built to produce a usable answer. Under pressure to supply one, they generated the evidence for the comfortable reading, and they did it fluently, with page numbers.

I have done the same thing. Not with footnotes — with Scripture, and with the sense that God had shown me something, and with the quiet confidence of a man who has been praying about it. The machinery was different and the motion was identical: start from a conclusion that resolves, and reach back for the support.

The Tyrians invented nothing. They had a true word and they were honest men. But they did what I did, and what the machines did: they reached past what they had been given for a conclusion that would let everybody go home.

And every voice agreed with them.

The therefore

The step nobody argued for is worth laying out as the argument it was.

Danger is coming, therefore God wants you out of the way.

Put like that it looks like an argument, and it is one. It has a premise the Spirit supplied and a premise the Spirit did not: that God’s purpose for his servant is his servant’s safety. The second one was not revealed to anybody. It was simply assumed, by everyone in the house, including the man writing the account.

Nor is it a foolish assumption. Most of the time it is a good one. If you are warned that a road is dangerous, the ordinary meaning of the warning is: take another road. A word of caution is usually a word to be cautious.

So the problem is not that the Tyrians reasoned. Luke does not think reasoning is the enemy. In chapter 16 Paul has the vision of the man of Macedonia, and Luke says they set out at once, concluding that God had called us to preach the gospel to them. Concluding. Inferring. The same motion exactly — revelation, then a human judgement about what it meant — and Luke records it as right.

Inference is not the fault. Unexamined inference is.

The Tyrians did not weigh a conclusion and get it wrong. They never saw a conclusion to weigh. The step from bonds await you to so do not go was taken below the level where anyone was watching, and by the time it surfaced it had stopped being an inference and become simply what the Spirit had said.

Which is how it works in practice. Not a bad decision, taken deliberately. A good one, taken invisibly — and then defended, afterwards, as though God had said it.

Paul’s therefore, and mine

Paul had the same information and reached a different conclusion.

He says so at Miletus, before Tyre, before Caesarea: the Holy Spirit testifies to him in every city that imprisonment and afflictions await him. He is not working with better data. He has been hearing this for weeks.

Then:

But I do not account my life of any value nor as precious to myself, if only I may finish my course.

There is the second premise, and it is not the one everybody else was using. Not God intends my safety. Something closer to: my safety is not the thing being decided. Same revelation, different assumption underneath, and the therefore comes out the other way.

Which is not a technique. Paul cannot hand anyone a method for this, and he does not try. He can only say what he is for, and go.

An authentic word, accurately received, correctly reported — and still not the thing to do.

I have been on the other side of this.

In the early eighties we came back to Australia from Papua New Guinea after our first term — three years — to find out what we were meant to do next. It was a real question. We did not assume we were returning.

Through a series of incidents over those months I became convinced we were to go back. We made the decision, and we began setting it in motion: the practical machinery of getting a family to PNG, which once started is not easily stopped.

Then the Pastor came to me and told me, prophetically, that I was to remain where I was and not return.

I say the Pastor rather than my Pastor deliberately. He was mine, and I had sat under him, and the tension was real. But by then my sense of calling had moved beyond the local church, and that is the honest account of where I stood when he spoke. It is also, I suspect, part of why I heard him the way I did.

He was sincere. He was not manoeuvring, and he had nothing to gain. He was the one man with any standing to say it to me. He said it, and I went anyway.

I would say now that the years since have borne out the decision.

But I want to be careful with that sentence, because it is the move this whole article is about. It worked out is a conclusion I have reached by reading my own life backwards, which is the only direction any of us can read it, and it is weaker evidence than it feels. Paul got Acts 23:11 — the Lord standing by him in the barracks, saying that as he had testified in Jerusalem so he must testify in Rome. A word, afterwards, that he had not been wrong. I have never had anything like that. I have outcomes, and outcomes are not verification. The prosperous road is not thereby the obedient one.

So: I believe we were right. I cannot prove we were right. Those are two different sentences, and I did not spend the following years holding them apart — I went back to PNG and got on with the work. It is only now, reading it against Acts 21, that I notice how easily the first becomes the second when nobody is watching.

And the Pastor? I do not think he was simply mistaken, in the way one is mistaken about a train time. Calvin’s reading has room for what I suspect happened — a man genuinely hearing something, whose gift did not extend to the shape of somebody else’s calling. He was not given my course. He was not supposed to be. He was given something, and what he made of it was a sentence about my future, and that is precisely the step the Tyrians took.

I did to him what Paul did to them. I heard him out, and I went.

The difference is that Paul was Paul, and I am not, and nobody wrote down afterwards whether I had been right.

Beyond that

So, was Paul disobedient?

I have been reading this passage since early this morning and I cannot tell you. Chrysostom could, and Calvin could, and Stedman could, and they gave three different answers, and they worked at it longer than I have. The sentence will not give up a verdict, and I have come to think that is not an accident of transmission. Luke knew how to write an unambiguous oracle. He writes one at Caesarea.

What the chapter will give is something else, and it is less comfortable than an answer.

A word from God came to a group of people. They received it accurately. They reported it honestly. And then they added one step — the obvious step, the step everybody present agreed with, the step so natural that nobody noticed taking it — and by the time it had been said aloud for seven days it was indistinguishable from the word itself.

That step is where nearly everything goes wrong.

It is not solved by sincerity. The Tyrians were sincere. It is not solved by the word being genuine; on Calvin’s reading it was genuine. It is not solved by unanimity, which at Tyre and at Caesarea was total. It is not solved by the person speaking having standing, as the Pastor had. And it is not solved by how it turns out, which is the comfort I would most like and am least entitled to.

Beyond that, I do not know.

Luke does not resolve it either. He does not tell us who was right at Tyre. He takes Paul up to Jerusalem, where everything the prophets said would happen happens, and he leaves his reader standing in the house at Caesarea with a company of people who have run out of argument.

What they say there is the last honest thing anyone says about it.

And since he would not be persuaded, we ceased and said, “Let the will of the Lord be done.”